I: The Text Behind the Text of the Hebrew Bible

012013

012

This is Part I of a two-part article; the second part will appear in the next issue of Bible Review. Part 2 will discuss the recovery of a missing passage in the Book of Samuel, as well as new developments in our understanding of late biblical religion.—Ed.

Nearly 40 years have passed since that fateful spring day in 1947 when a young Bedouin shepherd threw a stone into a cave in the cliffside on the northwestern shore of the Dead Sea and heard the sound of pottery shattering inside. When he and a companion later gathered nerve to crawl into the cave (now known as QumraÆn Cave 1), they found seven decaying rolls of leather. These were the original “Dead Sea Scrolls.”

William Foxwell Albright, the most distinguished Near Eastern archaeologist and Hebrew epigraphist of his generation, immediately hailed the finds as the greatest manuscript discovery of modern times.

In the years that followed, both archaeologists and Bedouin have explored and dug in hundreds of caves in the great waÆdis that, like the WaÆdi QumraÆn, cut through the towering cliffs that mark the Jordan Rift. In the competition between clandestine Bedouin diggers and archaeologists, it must be confessed that laurels have gone more frequently to the intrepid discover and patient shepherds. In any case, eventually ten additional caves with leather and papyrus manuscripts were found in the vicinity of Khirbet QumraÆn, the ruins of a community of Essenes—Jewish sectaries—to whose library the documents once belonged. So we now have manuscripts from QumraÆn caves numbered 1 through 11. From Cave 11 came the great Temple Scroll acquired by the late Yigael Yadin in 1967.^a1 More manuscripts and papyri were discovered in the large caves in the waÆdis south of QumraÆn: the WaÆdi Murabba‘aÆt, the Nahal Hever and the Nahal Se’eliÆm. More recently, in 1962, the oldest group of documents from the Jordan Rift was found in the WaÆdi ed-DaÆliyeh, north of Jericho. These are the Samaria legal papyri from the fourth century B.C.^b

Most recently of all, in 1963–1964, manuscripts were uncovered in Yadin’s excavations of the ruins of Herod’s fortress, Masada, atop a diamond-shaped mountain ovelooking the Dead Sea.

In any other generation each of these finds would have been regarded as nothing short of sensational. Altogether they have been overwhelming—in two senses. First, the magnitude of these discoveries can hardly be comprehended. Second, the labors of piecing together hundreds of thousands of fragments, editing, interpreting and assimilating these manuscripts have often overwhelmed the scholarly community with a responsibility both glorious and oppressive. Nearly 40 years of discovery and research are now past. I suspect that another 40 years will pass before the first exploratory investigation of these “treasures of darkness” will be completed. Almost each year a large new volume of unpublished material comes into print, and this will be so for many years to come. I am myself in the process of completing three volumes of unpublished manuscripts and papyri. The impact of all these discoveries and of all this research will be enormous.

I should like to explore here several important areas of historical study in which new insights and conclusions are emerging.

First, I shall discuss the bearing of new studies 013upon our understanding of the history of the biblical text. From the Dead Sea Scrolls, we have learned a great deal about the early transmission of biblical books, the fixation of the text of biblical books, and even the procedure by which the canon of the Hebrew Bible came into being. In short, we now know in some detail what the biblical materials were like before they became “biblical,” as well as the process by which the texts became fixed and chosen as “biblical.”

To place the new evidence in context, it will be useful to review briefly the status of the study of the history of the text of the Hebrew Bible prior to the discovery of the manuscripts on the shore of the Dead Sea.

The Bible survives in many Hebrew manuscripts and in several ancient versions translated from the Hebrew. In the medieval Hebrew manuscripts there are hundreds, even thousands of differences, mostly minor, rarely major. In the old versions, especially in the Old Greek version (which was written beginning in the third century B.C. and is commonly called the Septuagint) there are thousands of variants, many minor, but also many major. Even before the discovery of biblical manuscripts in the caves of QumraÆn and elsewhere in the Jordan Rift, these manuscripts and versions provided a rich body of resources for the textual critic’s attempts to reconstruct the history of the biblical text. At the same time, the history of the text of the Hebrew Bible has been confused and obscured by an assumption, or rather a dogma, on the part of the ancients—Rabbis and Church Fathers alike—that the Hebrew text was unchanged and unchanging, unaltered by the usual scribal realities that produce families of texts and different recensions^c in works that have survived over long 014periods of transmission.

This dogma of the “Hebraica veritas” may be found as early as the late first century of the Common Era, when Josephus, the Jewish historian, wrote:

“We have given practical proof of our reverence for our scriptures. For, although such long ages have now passed, no one has ventured to add, or to remove, or to alter a syllable; and it is an instinct with every Jew, from the day of his birth, to regard them as decrees of God, to abide by them, and if need be, cheerfully to die for them.”²

Josephus evidently regarded the Hebrew Bible as having, in theory at least, an immutable text.³

Origen, the Church Father, ordinarily used the Old Greek version of the Bible. But he, too, apparently assumed that his Greek Bible was translated from a Hebrew textual base that was the same as the Rabbinical Hebrew text in current use in his day. 016Hence, in his monumental Hexapla,^d he carefully corrected his Greek manuscripts to the “Hebraica veritas ”—incidentally, with catastrophic results for the subsequent transmission of the Greek Bible.

Jerome, writing in the fourth century, applied the principle of “correcting to the Hebrew” to the Latin Bible, displacing earlier Latin translations (based on the Old Greek Bible) with a new Latin translation that has come to be called the Vulgate, a Latin version translated from the standard Rabbinic Recension of the Hebrew Bible in use in Jerome’s time.

The search for the early stages in the history of the text of the Hebrew Bible began to be pursued scientifically in the late 18th century, but the extant manuscripts were all of medieval date, and the results were disappointing for those who hoped to find traces of archaic forms of the text. The sifting of the medieval manuscripts yielded, in its mass of variant readings, no evidence of alternate textual families or text types. The variants were secondary^e and of late date, the slips and errors of medieval scribes. Indeed, it could be argued that the theory of a fixed and unchanging Hebrew text was given added support by the evidence from the collections of medieval manuscripts.

Some of the more astute textual scholars, however, argued that all medieval Hebrew manuscripts derived from a single recension fixed early in the Christian era, and that this recension alone survived in the Jewish communities. Direct access to the early development of the text of the Hebrew Bible (prior to the recensions) was thus effectively blocked.⁴ Accordingly, the sea of variants in the great collections of manuscripts was of little or no help in the endeavor to recover ancient readings standing behind corruptions in the textus receptus.^f It could be, 019and was, argued that the medieval text stemmed from a single archetype, or from single manuscripts of each biblical work, which already possessed the pattern of errors held in common by the medieval text.⁵

The fact is, however, that in the 19th century, there was little hard evidence to determine precisely the procedure by which the Rabbinic Recension, found in all medieval manuscripts, came into being and was promulgated. In the end, the vigorous scholarly debates of the 19th century subsided, and while much research and theorizing continued, no major advances were made until the discovery of Hebrew and Greek manuscripts in the Wilderness of Judah—the Dead Sea Scrolls.

The discovery of ancient manuscripts in the 11 caves of QumraÆn provided the first unambiguous witnesses to an ancient stage of the Hebrew text of the Bible.⁶ These caves have yielded some 170 manuscripts of biblical books, most of them in a highly fragmentary state, and their publication is still in progress.⁷

Although all the evidence is not yet published, we can compare these QumraÆn manuscripts with a dozen or so biblical manuscripts, again fragmentary and some still unpublished, recovered from the Nahal Hever, the WaÆdi Murabba‘aÆt, and Herod’s fortress at Masada. The two groups of manuscripts—the QumraÆn manuscripts, on the one hand, and the manuscripts found in the southern caves and at Masada—vary in two critical respects. The manuscripts of the QumraÆn group are earlier, varying in date between about 250 B.C. to 68 A.D., at which, time the Essene community at QumraÆn was destroyed by the Romans as part of their suppression of the First Jewish Revolt (66–70 A.D.). On palaeographical grounds, we can date most of these biblical manuscripts no later than the first half of the first century of the Common Era, and most are earlier. The second “southern group”—from the caves of the WaÆdi Murabba‘aÆt, from the Nahal Hever, and from Masada—date as a group from a later period. Most important of the manuscripts of the southern group are the great Hebrew Minor Prophets^g Scroll from a cave in the WaÆdi Murabba‘aÆt and the Greek Minor Prophets Scroll from the Nahal Hever.

The Minor Prophets Scroll from Murabba‘aÆt can be dated palaeographically to the second half of the first century of the Common Era, and the biblical fragments from Masada to no later than 73 A.D., when the Romans stormed the bastion and destroyed its fortifications. A number of the biblical fragments from the southern caves date to the interval between the First and Second Jewish Revolts, that is, between 70 and 135 A.D., and once belonged to followers of Bar Kokhba, the messianic leader of the Second Revolt (132–135 A.D.).

The two groups of biblical manuscripts differ not only in date. The southern (later) group reveals a text that shows no significant deviation from the archetypal Rabbinic Recension—that is, the recension that is ancestral to the Masoretic Text,^h our traditional Hebrew Bible.⁸ This is in marked contrast to the QumraÆn group of documents, which reveals other text types.

The data drawn from the southern manuscripts enable us to conclude that before the end of the first century of the Common Era, a recension of the text of the Hebrew Bible had been promulgated that had overwhelming authority, at least in Pharisaic circles, and that came to dominate the Jewish community in the interval between the fall of Jerusalem to the Romans in 70 A.D. and the Roman suppression of the Second Jewish Revolt in 135 A.D.

The textual situation at QumraÆn differed totally. The QumraÆn manuscripts show no influences that we can detect of the standardization that marks the Rabbinic Recension. At QumraÆn we find evidence of discrete and, indeed, recognizable families of textual tradition, including text types that are different from the Rabbinic Recension. These variant streams of tradition have been called “recensions” or “families” or “local texts.”⁹

Sometimes one of these text types differs strikingly in detail from the traditional text that has come down to us. In extreme instances we discover that a textual tradition is preserved in a manuscript that stems not merely from textual changes in individual readings; it derives from an edition of a biblical work different broadly in content and length from the edition used in the Rabbinic Recension. For example, there are two editions of Jeremiah represented in manuscripts from QumraÆn: a long edition known from our traditional Bible and a short edition that also differs in the order of the prophetic oracles. There are two editions (or collections) of the Psalter, one Persian in date, one Hellenistic. There is a whole Daniel literature, of which the Book of Daniel is only a single part. Instances of different editions of biblical books, however, are relatively rare. For the most part the textual families reflected in extant biblical 020manuscripts are marked by variants in individual readings: grammatical changes, alternate vocabulary, omissions or additions of words, phrases, and even, on occasion, paragraphs.

The different text types of most biblical books appear to be the product of natural growth, or of local development, in the process of scribal transmission, not of a controlled or systematic recension, revision, or collation at a given place or time. At the same time, the different texts possess traits, some more or less systematic, that permit them to be classified in different families. The common traits of textual family include, for example, their “bad genes,” an inherited group of mistakes or secondary readings perpetuated by copyists generation after generation. Other distinguishing traits may be a particular orthographic (spelling) style, the type of script utilized, the repeated appearance of a peculiar chronology or numeral calculation (arising often in attempts to resolve apparent or real errors in traditional numbers), the systematic introduction into the text of parallel readings (especially in legal sections with parallel sections in other books), and repeated use of archaizing or “modernized” grammatical and lexical features.

The QumraÆn manuscripts not only provide evidence of early textual traditions; perhaps even more important, the data drawn from the QumraÆn discoveries enable us to identify and delineate other textual traditions that survive from times before the Common Era—including the Hebrew textual base of the Old Greek translation, the textual background of the Samaritan Recension of the Pentateuch, and the text type that was utilized in the Rabbinic Recension. In this great complex of textual materials, as many as three textual families have been identified in certain biblical books (the Pentateuchal books and the books of Samuel), two textual families in other books, notably in Jeremiah and Job, and in many books only one textual tradition is reflected in extant data (for example, Isaiah and Ezekiel). The textual critic is thus confronted with the task of organizing this evidence: the existence of a plurality of textual types in the early era, the limited number of distinct textual families, and the relative homogeneity of the variant textual traditions over several centuries of time.

I have proposed a theory of “local texts” to satisfy the requirements of this data. As applied to those books where three textua families exist, namely in the Pentateuch and Samuel, this theory may be sketched as follows: Three forms of the text appear to have developed slowly between the fifth century B.C. and the first century B.C., in the Jewish communities in Palestine, in Egypt, and in Babylon, respectively. The Palestinian text is the dominant family in the QumraÆn manuscripts. Its earliest witness is found in the Chronicler’s citations of the Pentateuch and Samuel. This Palestinian text was also used in the Samaritan recension of the Pentateuch. At least in its late form, the Palestinian text can be characterized as expansionistic, a full text marked frequently by conflation,ⁱ glosses,^j synoptic additions (that is, the insertion of readings from parallel passages in other sources), and other evidence of intense scribal activity. Omissions owing to scribal lapses are relatively infrequent. To this family belong the Pentateuchal manuscripts inscibed in the Palaeoa-Hebrew script, a derivative of the old national script of pre-Exilic Israel.^k

The second textual family, which we label Egyptian, is found in the Old Greek (“Septuagint”) translation of the Pentateuch and Reigns (the Greek version of Samuel and Kings), and in the short edition of Jeremiah found in one Hebrew manuscript at QumraÆn. In some respects the Egyptian family resembles the Palestinian text, especially the earliest of the Palestinian witnesses, and may be regarded as a branch of the Old Palestinian family.

The third family we designate “Babylonian,” although we are, in fact, uncertain of the locale of its origin. As we shall see, the intellectual influence of the powerful Babylonian community was to exercise a decisive role in the emergence of the authoritative Rabbinic Recension. This third text type, known thus far only, in the Pentateuch and Samuel, forms the base of the Rabbinic Recension. In the Pentateuch, it is a conservative, often pristine text, which shows little expansion, and relatively few traces of revision and modernizing.

Thus at QumraÆn and in traditions of the biblical text that broke off from the main Jewish stream before the turn of the Common Era, we find several textual families. None, including the text type ancestral to the Rabbinic Recension, shows evidence of a systematic recension or stabilization.

In the southern caves and at Masada, however, we find only a single text type, one that shows every evidence of the external controls that fixed the text we call the Rabbinic Recension. The southern group 021of manuscripts stands very close to the archetype this recension. We are led, therefore, to the conclusion that the Rabbinic Recension of the Hebrew— Bible—what we may also call the authoritative Pharisaic text—was fixed by the time of the Roman destruction of Jerusalem in 70 A.D. This recension became regnant only in the interval between the two Jewish Revolts when the Pharisaic party came wholly to dominate the surviving Jewish community and rival parties diminished and disappeared. Sects like the Christians and Samaritans continued to exist but only as separate communities, isolated from Pharisaic influence. Rabbinic Judaism survived and with it the Rabbic Recension.

The Rabbinic Recension was promulgated as a response and solution to a textual crisis that developed in late Hellenistic and early Roman times. Maccabean Revolt, initiated in 167 B.C., ultimately reestablished an independent Jewish state, which had not existed since the time the Babylonians destroyed Jerusalem and the First Temple in 587 B.C. In the wake of Maccabean victories that led eventually to the full independence of Judea under the rule of Simon the Maccabee (140–134 B.C.), Zionist revival was fueled, augmented by Parthian expulsions of the Jews. A flood of Jews returned Jerusalem from Babylon, Syria and Egypt.¹⁰ By the first century before and the first century after Common Era, competing local texts and editions had found their way to Judah, causing considerable confusion, as reflected in the library at QumraÆn. Moreover, the uncontrolled development of the of individual textual families became intolerable precipitated a textual crisis when the urgent need precise doctrinal and legal (halakic) exegesis arose Hellenistic Judaism. Party strife began in earnest the mid-second century B.C. with the emergence the Sadducean, Pharisaic, and Essene parties, the subsequent religious disputes between the parties increased the need for a fixed, authoritative text. By the beginning of the first century of the Common Era, there was further splintering into sectarian groups, and there is evidence of intense intra-party as well as inter-party and sectarian dispute and contention.

These data provide the general time and historical under context for the creation of the Rabbinic Recension. Other hints, limiting the time frame in which we must place the promulgation of the Rabbinic Recension, are found in the history of the Greek recensions.¹¹ The Rabbinic Recension was promulgated in the first half of the first century of the present era. In editions these same days we witness also the fixing of considerable hermeneutical rules,^l as well as read reports of Pharisaic discussions of the schools of Hillel and Shammai, which presume a more or less fixed text. I think it is not too much to go even further and to attribute the initiation of the recensional labors that fixed the text of the Pharisaic Bible to the great sage Hillel himself (early first century A.D.)— or at least to the school of Rabbinic scholars he inspired.

Hillel, it should be remembered, came to Palestine 022from Babylon and became the dominant and mostcreative spirit of his day; he was a giant whose impress on Pharisaism cannot be exaggerated, and his direct descendants were the principal leaders in the normative Jewish community for many generations.

The fact that Hillel (and his circle) were responsible for the selection of the proto-Rabbinic manuscripts that stood behind the Rabbinic Recension would explain a number of its peculiarities. For example, the texts of the Pentateuch and Samuel that were chosen for the Rabbinic Recension appear to be of Babylonian origin rather than the prevailing late Palestinian texts that were available.¹² In their recensional activities, the Rabbis also rejected the Palaeo-Hebrew script and its orthographic style, which was used in the most elegant Pentateuchal manuscripts inscribed in Palestine, choosing instead the common Jewish script in broad use in Palestine and throughout the diaspora Jewish communities. The choice of the common Jewish script is particularly striking in view of the official use of the old national script by the ruling high priests for temple inscriptions and for their coinage.

The fixation of the text by Pharisaic scholars followed a pattern unusual in the textual history of ancient literary documents. The Pharisaic scholars did not produce an eclectic text by choosing preferred readings and rejecting obvious glosses or additions. This was the procedure followed by Greek scholars in Alexandria in establishing a short, if artificial, recension of the text of Homer. Nor did the Rabbis combine variant readings from different textual traditions, a recensional technique that produced conflate recensions of the Septuagint and the New Testament. Instead, the Rabbis selected a single textual tradition, which we term the proto-Rabbinic text, a text that had been in existence in individual manuscripts for some time.¹³ In a given biblical book of the Hebrew Bible the Rabbis chose exemplars of one textual family or even a single manuscript as a base. They did not collate all the wide variety of text types available; on the contrary, they firmly rejected in some instances a dominant late Palestinian text. It should be noted, however, that they did not select, in the case of every book, texts having a common origin or local background. In the Pentateuch they chose a short, relatively unconflated text—a superb text from the point of view of the modern critic—which we believe derived from a conservative Babylonian textual tradition. In the Major Prophets, on the other hand, they chose the relatively late and full Palestinian text of Isaiah, Ezekiel, and Jeremiah. In Jeremiah, in fact, they selected the long edition of Jeremiah in preference to the shorter, and in some ways, superior edition.

The choice of a non-Palestinian text of the Pentateuch is of particular interest. The books of the 023Torah (the Pentateuch) held central authority for all the Jewish parties. Indeed, the Sadducees and the Zadokite priests of the separatist Samaritan community regarded the Pentateuch alone as the basis of religious doctrine and practice. The Samaritans, in contrast to the Rabbis, chose for their sectarian recension of the Pentateuch a late Palestinian text inscribed in Palaeo-Hebrew, also known from the finds at QumraÆn.

We may speculate that Hillel’s personal preference was responsible for the surprising choice of the Babylonian textual base for the Pharisaic Pentateuch. In this case, the conservative Torah scrolls that he knew and to which he was accustomed became, under his urging, the basis of the new Rabbinic Recension. It is quite possible that an old saying embedded in the Babylonian Talmud preserves a memory of Hillel’s role in the events leading to the fixation of the Hebrew text and canon:

“When Israel forgot the Torah, Ezra came up from Babylon and reestablished it; and when Israel once again forgot the Torah, Hillel the Babylonian came up and reestablished it …”¹⁴

This much is certain. The vigorous religious community in Babylon repeatedly in Jewish history developed spiritual and intellectual leaders who reshaped the direction of Palestinian Judaism and defined its norms. Such was the case in the Restoration after the Exile, again in the person of Hillel, and finally in the rise of the Babylonian Talmud.

In the lines above, we have written almost exclusively about the fixation of the text as opposed to the stabilization of the canon. In the remarks that follow we shall focus on the latter, and specifically on the fixation of the Pharisaic canon of the Hebrew Bible. We shall use the term “canon” in its strict sense: a fixed list of books of scripture that was deemed unvariable, not to be added to or subtracted from. In origin, the term canon meant a rule, and concretely in the usage of the Church Fathers, a closed list of books defined as authoritative for religious faith and practice.

The earliest clear definition of the “closed” Hebrew canon is found in Josephus in his apologetic work, Contra Apionem, written in Rome in the last decade of the first century of the Common Era. He asserted that there was a fixed and immutable number of “justly accredited” books, 22 in number.^m Their authority was founded on their derivation from a period of uncontested prophetic inspiration beginning with Moses and ending in the era of Nehemiah.

[W]e do not possess myriads of inconsistent books, conflicting with each other. Our books, those which are justly accredited, are but two and twenty, and contain the record of all time. Of these, five are the books of Moses, comprising the laws and the traditional history from the birth of man down to the death of the lawgiver. … From the death of Moses until Artaxerxes, who succeeded Xerxes as king of Persia, the prophets subsequent to Moses wrote the history of the events of their own times in thirteen books. The remaining four books contain hymns to God and precepts for the conduct of human life. From Artaxerxes to our own time the complete history has been written, but has not been deemed worthy of equal credit with the earlier records, because of the failure in the exact uccession of prophets.”¹⁵

Josephus’s “canon” specifically excludes works of Hellenistic date, and by implication works attributed to pre-Mosaic patriarchs. In the paragraph subsequent to the one cited, he adds that the precise text of the 22 books was fixed to the syllable.

Where are we to seek the origin of Josephus’s doctrine of a fixed text and a fixed canon? Josephus was a Pharisee, and I believe that he is here drawing upon his Pharisaic tradition and ultimately the work and teachings of Hillel.

There is no evidence in non-Pharisaic Jewish circles before 70 A.D. of either a fixed canon or text. The Essenes at QumraÆn exhibit no knowledge of this fixed text or canon. The same is true in the Hellenistic Jewish community in Alexandria, and in the early Christian communities. Until recently there has been a scholarly consensus that the acts of inclusion and exclusion that fixed the canon were completed only at the “Council of Jamnia (Yabneh)” meeting about the end of the first century of the Common Era.ⁿ However, recent sifting of the Rabbinic evidence makes clear that in the proceedings of the academy of Yabneh the Rabbis did not fix the canon, but at most discussed marginal books, notably Ecclesiastes (Qohelet)and the Song of Songs. The Rabbis asserted that both Ecclesiastes and the Song of Songs “defile the hands,” i.e., are holy books. They should thus be included in the canon. This decision thereby ratified the dicta of the house of Hillel in the case of 024Ecclesiastes, and probably in the case of the Song of Songs as well.¹⁶ Moreover, it must be insisted that the proceedings at Yabneh were not a “council,” certainly not in the late ecclesiastical sense. Whatever decisions were taken at Yabneh, they were based on earlier opinions, and they failed to halt continued disputes concerning marginal books: Song of Songs, Ecclesiastes and Esther of the “included” books, Ben Sira among the “excluded” or apocryphal. In any case, it is clear that Josephus in Rome did not take his clue from contemporary or later proceedings at Yabneh, nor did he manufacture a theory of canon from whole cloth. Thinly concealed behind Josephus’s Greek apologetics is a clear and coherent theological doctrine of canon that must stem, we believe, from the canonical doctrine of Hillel and his school.¹⁷

We cannot press the date of the fixation of the Pharisaic canon earlier than the time of Hillel, as an occasional scholar has attempted to do. Our evidence comes from the so-called Kaige Recension referred to in endnote 11. The Kaige Recension, at the end of the first century B.C., revised the Greek Bible to accord with the proto-Rabbinic text, not with the later fixed Rabbinic Recension. Similarly, the revision embodied in the Kaige Recension extended to the Book of Baruch and the longer edition of Daniel, works excluded from the Rabbinic Recension. This effort to “update” Baruch and the longer edition of Daniel would be most difficult to explain if at the time of the preparation of the Kaige Recension, the Book of Baruch and the additions to Daniel had already been excluded from the Pharisaic canon. Since the recensional labors in the Kaige Recension can be dated to about the turn of the Common Era, and its Pharisaic bias is clear, it follows that as late as the end of the first century B.C., an authoritative, canonical list had not yet emerged, at least in its final form, even in Pharisaic circles.

I am persuaded by the accumulation of evidence that the same circumstances that brought on the textual crisis that led to the fixation of the Hebrew text—varied texts and editions, party strife and sectarian division, the systematization of hermeneutic principles and halakic dialectic^o —were the occasion as well for a “canonical” crisis, requiring the fixation of a Pharisaic canon, and further, that Hillel was the central figure in sharpening the crisis and responding to it. The fixation of the text and the fixation of the canon were thus two aspects of a single if complex endeavor. Both were essential to erect “Hillelite” protection against rival doctrines of cult and calendar, alternate legal dicta and theological doctrines, and indeed against the speculative systems and mythological excesses of certain apocalyptic schools and proto-Gnostic sects. To promulgate a textual recension, moreover, one must set some sort of limit on the books whose text is to be fixed. In electing the text of one edition of a book over the text of an alternate edition—in the case of Jeremiah or Chronicles or Daniel—one makes decisions that are at once textual and canonical. 025Ultimately, the strategies that initiate the fixation of the biblical text lead to the de facto, if not de jure, fixation of a canon.

The principles guiding the exclusion of works from the Pharisaic canon reflected in Josephus’s notices no doubt also operated in eliminating works offensive to Hillel and the house of Hillel. The host of pseudepigraphical works written in the name of Enoch, Melchizedek, the sons of Jacob, Amram, and the like, which became popular in Hellenistic times, and which fill the library of QumraÆn, were excluded from the canon. The prophetic sequence began with Moses. There can be little doubt, moreover, that the Rabbis recognized the recent date of certain apocryphal and pseudepigraphic works since such cycles as Enoch and the Testaments of the Patriarchs were still in their creative, fluid phase of composition, unfixed as literary works, into the Roman Period. The principle of excluding works of “post-Prophetic” authorship pennitted also the suppression of the propagandistic Book of Maccabees, certain of the Hellenistic novellas, and Ben Sira, although the case of pseudepigraphs written in the name of the “Prophets,” especially the Jeremianic apocrypha, Baruch and the Letter of Jeremiah must have caused difficulty and dispute. Ezekiel, Song of Songs, and Ecclesiastes were controversial works, in all probability, because of their content, but were sufficiently old and recognized to prevent their being excluded from the canon.¹⁸ Most mysterious is the selection for inclusion of an edition of Daniel not earlier than the Maccabean Age, although it contains earlier material, and of Esther. It must be said, however, that in general, the Rabbis chose for inclusion in their canon, works or editions that in fact reached their final literary form (that is, when compositional activity ceased) by the end of the Persian Period (late fourth century B.C.).

If I am correct in perceiving the hand of Hillel in the promulgation of a Pharisaic text and canon, and in recognizing a reference to this achievement in the Rabbinic saying, “When Israel once again forgot [the Torah], Hillel the Babylonian came up and reestablished it,” I must nevertheless acknowledge that this canon and text did not immediately supplant other traditions or receive uniform acceptance even in Pharisaic circles. The ascendancy of the Hillelite text and canon came with the victory of the Pharisaic party and the Hillelite house in the interval between the two Jewish Revolts against Rome. After that, the text and the canon of the Hebrew Bible—despite Rabbinical queries about marginal books from time to time—remained fixed and guarded down to our own day.

12 This is Part I of a two-part article; the second part will appear in the next issue of Bible Review. Part 2 will discuss the recovery of a missing passage in the Book of Samuel, as well as new developments in our understanding of late biblical religion.—Ed. Nearly 40 years have passed since that fateful spring day in 1947 when a young Bedouin shepherd threw a stone into a cave in the cliffside on the northwestern shore of the Dead Sea and heard the sound of pottery shattering inside. When he and a companion later gathered nerve to […]