Articles and book excerpts used in and referred to on Issues, Etc.

Modern discussions of the Old Testament canon date to Luther's rejection of the apocrypha in 1519. In a debate with Johannes Eck at Leipzig, Luther admitted that Eck had correctly quoted 2 Maccabees 12:46 and did not dispute that it encouraged prayers for the dead. Luther did, however, deny the canonical status of 2 Maccabees. Yet, Luther was not the first. Questions of the extent of the canon have been a subject among Christians since the second century. Originally the question was largely a debate over whether certain books belonged in the Bible (with a few exceptions such as Esther, this involved the apocryphal/deuterocanonical books). However, for Christians from various traditions today the question of the extent of the OT canon has been settled for their particular tradition.¹ Roman Catholics accept a wider canon that includes books not found in the Jewish canon. Protestants reject these additional books, although they differ on their usefiulness.²

For Protestants this rejection of the apocrypha sterns from a historical question of the extent of the canon used by Jesus and his apostles. The Reformers answered this question when they accepted the Hebrew OT. They asserted that this was the canon in use by Jews in Jesus' day. Protestants, therefore, have rejected the apocrypha. For the Roman Catholic Church the question of Christian tradition is of primary importance (which itself implies a historical question). The Council of Trent affirmed the canonical status of the wider OT canon. Therefore, Catholics prefer to call the books labeled apocrypha by Protestants deuterocanonical books.

This study addresses the historical question of the OT canon. Simply put, this question is: Was the canon formed and accepted among Jews by the time of Jesus and the apostles, or was it still to be formed? This question is of primary importance for deciding between the competing claims on the canon. If the canon was not formed in Jesus' day, then the church would have nowhere to look except to its own traditions to determine the extent of the canon. However, if a generally accepted canon did exist in Jesus' day, then later tradition may not be valid, especially for those Christians who reject tradition as determinative of doctrine.

To answer this question, I will examine the historical evidence for the OT canon from the second century BC through the third century AD, with some references to the fourth century as appropriate. This chapter will deal with preliminary questions: definition of canon, modern theories on the formation of the canon, and the method of investigation into the question of the canon. Subsequent chapters will examine the evidence for the canon from the second century BC through the fourth century AD. The final chapter will summarize the evidence and the conclusions about the OT canon.


A Definition of Canon

The English word canon was derived from the Greek word K(WC~M) which means reed or measuring slick. An additional meaning - rule - was derived from this basic meaning. The New Testament (Gal. 6:16) and the early church fathers used the word with this additional meaning. Later, the word came to be used to describe a collection of books viewed as sacred and authoritative for Christian faith and life.³ Thus, the word canon when applied to scripture primarily denotes a list of books viewed as authoritative. It secondarily implies (connotes) the concept of their divine origin (inspiration) which leads to their authoritative nature. Its synonym scripture denotes sacred written revelation of divine origin. Therefore, while the concept of canon cannot be totally separated from the question of inspiration, the primary focus of canon is on the collection of books. James Barr notes, "The word 'canon' meant simply 'list', i.e. the list of books that counted as holy scripture." He adds, "This is, and has always been, the normal meaning of the word in English when applied to scripture."⁴ Eugene Ulrich states, "A strict definition of canon will also include these latter concepts: conscious decision [to include or exclude certain books], unique status [divine origin], necessarily binding [authoritative]."⁵

While this seems straightforward enough, a number of scholars have taken issue with this definition. Brevard Childs notes that some writers distinguish sharply between canon and scripture.⁶ They argue that scripture denotes a body of authoritative writings whereas canon adds the denotation of restriction, implying that some books have been excluded from the collection. For these writers canon implies a closed collection to which no more books may be added. Therefore, while some scholars may speak of a period when the canon was open (i.e., the theoretical possibility existed that more books could be added), these scholars would argue that the term canon implies a closed list of books which cannot be supplemented by further authoritative writings.

The problem with this approach to canon is that it implies an all-or-nothing view of the concept of canonicity. That is, it denies the possibility some books were already disqualified as scripture while others were already considered scripture before the canon became the complete, closed collection we have today. It would seem to imply that no books were generally accepted authoritative until the entire collection was generally accepted. Considering the several centuries over which the OT books were written, this seems to be highly unlikely.

Sid Leiman offers another definition of canon. Based on his study of rabbinic literature he concludes that two categories of canonical books were recognized by Jewish authorities: inspired canonical literature and uninspired canonical literature.⁷ Inspired canonical literature is co-terminous with the OT. Uninspired canonical literature includes all other books seen as authoritative for Jews (Mishnah in its oral form and Megillath Taanith in written form).

Leiman's definition would seem to combine a scriptural canon with a collection of other books held to be authoritative, but not of divine origin (i.e., a canon of religious literature, but not a canon of scripture). That religious communities oftentimes accept other collections of books as authoritative but not on the level of scripture does not mean that they have one canon divided into two categories. Instead, it implies that they recognize two collections: a collection of Scripture and a collection of other books that, though useful, are not recognized as both authoritative and inspired. Thus, Lutheranism could be said to recognize a scriptural canon of the Old and New Testaments, but also to recognize as author-itative the works that are collected in the Book of Concord (a canon in the sense that it is a collection of authoritative books, but not a scriptural canon). However, these are really two distinct collections. Lutherans do not view the Book of Concord as part of a larger canon containing both Scriptures and their confessional documents.

In the same way the rabbis cannot be said to have recognized one canon with two categories of literature in it. Rather they recognized two distinct, but related, collections. A canon of scripture (the OT) and a collection of other documents that came to be regarded as authoritative (the Talmud).

Lee McDonald, following the definition of canon proposed by Gerald Sheppard, makes a similar distinction in the definition of canon.⁸ This definition of canon defines canon in two senses. Canon I designates any authoritative voice, whether written or oral. Canon 2 designates writings in a temporary (open canon) or permanent standardization (closed canon). This, according to McDonald, explains why some writings that were once considered inspired and authoritative within the Christian community (canon 1) were later excluded from the authoritative collection of literature for the church (canon 2).

The problem with McDonald's approach to canon is that it purposely confuses two different meanings of canon in the definition of canon I for ideological reasons. The meaning of canon as rule or norm is combined with the meaning of a generally recognized authoritative collection of inspired books in order to argue that the canon was not closed until a relatively late date. McDonald argues that since some in the church accepted certain books as authoritative or argued about the sacredness of certain books, the church had no OT canon (canon 2) until the fourth or fifth centuries, and the Jews had no scriptural canon during the first century.⁹

The flaw in McDonald's approach is that he equates any question raised within the church about whether or not a book belonged in the canon with an indication that the canon was not yet closed (no canon 2 yet existed). However, that the rabbis debated the merits of books Iike Ezekiel and Ecclesiastes or that some Christians considered Sirach or Wisdom inspired does not necessarily indicate that the canon had not been closed. It only indicates that some within a religious group may from time to time express doubts about the extent of a canon that may well have been considered closed for some time.¹⁰

In this regard, Luther's comments on certain books of the New Testament, notably James, are significant. No one would argue that the New Testament canon had not long been a closed collection in Luther's day. Luther's objections to James did not change that, and, in fact, it has had little influence even on Lutherans, the vast majority of whom have always accepted the canonicity of James.

Therefore, McDonald's redefinition of canon is motivated by his conclusions about the canon, especially his conclusion that the canon was closed at a late date. However, the mixing of two definitions of canon is less than helpful in the debate about the close of the canon. Certainly canon can mean rule. It can also mean authoritative collection of books. But this does not imply that one can use one sense of this term against another sense of this term in order to aid in proving an argument about the date of the closing of the canon. That would be like using the word diamond in the sense of a field on which baseball is played to argue that the gem known as diamond was not recognized until the nineteenth century.

Most recently Philip R. Davies has suggested a looser definition of canon.¹¹ Though he never offers a rigorous definition, he seems to operate with a definition of canon that would include any collection of works deemed to be classics. Thus, Davies can state, "A work becomes canonized (a) by being preserved by copying until its status as a classic is ensured; and (b) by being classified as belonging to a collection of some kind."¹² The problem with this minimal definition is that it does not distinguish between collections of books that are revered as good literature (i.e., a definitive collection of the works of Shakespeare) and books that are considered divinely inspired and authoritative (i.e., a scriptural canon) for a religious community. The discussion of the canon of Scripture turns precisely on the notion of religious authority, something that Davies' view does not seem to recognize or appreciate.

How then should we define canon when it is used to describe Scripture? In this study, the term canon will mean a collection of authoritative and divinely inspired books accepted as such by an overwhelming majority in a religious community.

This definition implies several corollaries:

The acceptance of books as authoritative means that they exercise an authority not only for the present generation, but also (in some sense) for all generations.¹³

The acceptance of some books as authoritative and inspired immediately implies that other works are excluded because they areviewed by the community as either not authoritative or not inspired or both).

The canon may be open. That is, the community can recognize that additional books may yet be written and will be added to the canon as authoritative and inspired without denying that it has already chosen to include some books and exclude others. However, by accepting some books as canonical it is asserting that none of these books can be later excluded. (To do so would deny that such books are authoritative.)

The canon may be closed. That is, the community can recognize that it will no longer allow any other books, already written or yet to be written, yet to be added to the canon.

The peculiar views of individuals or groups within the community about the addition or deletion of a book or books does not necessarily imply that no canon exists or that it is still in formation. It may imply thatsome individual or group is challenging an accepted canon's contents.


Modern Theories on the Formation of the Canon

The Triple Canon Theory

From the end of the nineteenth century to the middle of the twentieth century a generally held theory about the formation of the OT canon prevailed among scholars. This theory held that the OT canon formed in three stages corresponding to the three divisions of the canon found in Jewish Bibles. The Torah was the original section of the canon, and it was recognized by 400 BC. The Prophets were accepted by 200 BC. The Writings rounded out the canon and were officially recognized no later than AD 90 by a rabbinic council held at Jamnia. This council was responsible not only for adopting the Writings but also for closing the canon.

This consensus among scholars was challenged by some but remained the working hypothesis of most scholars for over fifty years. H. E. Ryle in his book The Canon of the Old Testament put forth the classic summary of this theory.¹⁴

This theory was supported by an earlier theory, Alexandrian Canon Hypothesis. Since the closing of the canon occurred after the beginning of the Christian church, Christians struggled on their own with the extent of the canon. Some eventually adopted the Hebrew scriptures of the rabbis. Christians in the West eventually adopted a wider canon. In order to explain this wider canon, the Alexandrian Canon Hypothesis was proposed.

The Alexandrian Canon Hypothesis held that Diaspora Judaism, especially in Alexandria, honored a larger set of books as sacred. These books were preserved in Greek, the language of the Diaspora, and were eventually adopted by the early church, whose language was also Greek.

It was noted that the codices of the Septuagint do not group the books of the OT into the Hebrew Bible's threefold division. They do group the books of Moses together, but the books of the Prophets and the Writings are not grouped the same way as in the Hebrew Bible. This was seen as an indication that the Jews in Alexandria were developing their own view of the canon before 200 BC when the Prophets were adopted into the Hebrew canon. This development continued in that a number of books not included in the Hebrew canon were adopted as Scripture by Alexandrian Jews. When the rabbis were closing the Hebrew canon at Jamnia, Christians had already begun to adopt the Greek Scriptures of Diaspora Judaism as their own. Later, Christians in the East fell under the influence of the rabbinical canon of Jamnia, and many of them eventually rejected the wider canon. Christians farther away continued to use the wider canon of Alexandria. Eventually, this wider canon became the one adopted by the Roman Catholic Church at the Council of Trent, while the Hebrew canon was endorsed by Protestants on the notion that it predated Jesus and was the canon he used.

The consensus for the Triple Canon Theory was built around four major assumptions:

1. The threefold division of the Hebrew Scriptures into Torah, Prophets, and Writings is an indication of the history of the canon, and not a later arrangement imposed upon the canon.

2. Many of the books contained in the Writings were completed too late to be considered canonical in Jesus' day. In particular Psalms, Ecclesiastes, Job, and Daniel were held to have been written in the Maccabean period, allowing little time for their recognition as Scripture.

3. The meeting of rabbis at Jamnia about AD 90 functioned like a church council in adopting a list of books that ever after defined the canon.

4. The OT books as preserved in the great Septuagint codices exemplify the Alexandrian Canon: Codex Alexandrinus, Codex Vatincanus, and Codex Sinaiticus, as well as others.

Although the Triple Canon Theory seemed to explain the formation of the canon and the source for different canons among Christians, it increasingly came under attack in the latter half of the twentieth century. While it still survives in modified form and has its defenders (e.g. McDonald¹⁵), each of its four major assumptions has been shown to be erroneous at least in part, if not entirely. It is the evidence that discredites these assumptions to which we now turn.

The Collapse of the Triple Canon Theory

The collapse of the Triple Canon Theory that was formulated in the late nineteenth century came about during the last half of the twentieth century. The details of the evidence discussed below that discredit the old consensus will be explored in subsequent chapters. Here I will off only a summary of the evidence that undermines each of these assumptions.

The first of the four assumptions listed above depends on the traditional arrangement of the books of the Hebrew OT into three sections. While this arrangement is an ancient one, the first firm evidence for it is references in the Talmud. The oldest of these passages that can be confidently dated come from the early second century and are attributed to Elisha ben Abuyah (110-135), Gamaliel (80-110) and Ben Azzai (110 - 135).¹⁶ Considering that the statements attributed to these men were not recorded in their own day, were collected and edited around AD 200, and may not have been reduced to writing until as late as the fourth century, this evidence can only be a provisional indication that the canon was divided into the three traditional sections even as early as the second century. Scholars who hold that the threefold division of the canon predates the early second century usually buttress their contention by referring to statements in the Greek prologue to Ben Sira (c. 132 BC), Philo's The Contemplative Life (early first century), Luke 24:44 and Josephus' Against Apion (late first century). However, none of these contains the threefold division of the OT corresponding to the latter Jewish divisions.

The Prologue to Ben Sira explicitly mentions the Law and the Prophets, but the supposed third division is variously referred to as "those who followed them," "the others books of our fathers," or "the rest of the books." One cannot be certain that these are even references to canonical books.

Both Philo and Luke mention only one book-Psalms-apart from the Law and the Prophets.

Josephus, on the other hand, divides the canon into three parts: Five books of Moses, the Prophets (thirteen books), and hymns to God and precepts for human life (four books). Josephus' arrangement of twenty-two books in three divisions is different from the traditional threefold arrangement of 24 books. Only the first division containing the five books of Moses is identical in both Josephus and the traditional threefold division of the UT canon.

Each of these pieces of evidence will be examined in more detail in later chapters. However, given the fact that no unambiguous evidence exists for the traditional threefold division of the canon before the second century AD, and that the most common way of dividing the OT canon in the NT is a twofold one (the Law and the Prophets), some scholars have challenged the notion that the threefold division of the canon has any relevance to the history of the canon's formation.¹⁷

The second major assumption behind the Triple Canon-that certain books contained in the third section of the canon were of Maccabean origin and, therefore, too young to be included in a canon by Jesus' day-has been abandoned because of evidence from both the Cairo Geniza manuscripts of Ben Sira and the manuscripts from Qumran and Masada. The Cairo Geniza manuscripts of Ben Sira in Hebrew discovered in the late nineteenth century, were first published by Solomon Schechter. Schechter' s examination of the Hebrew text revealed many passages in which Ben Sira employed phrases, expressions, and even entire verses from nearly every UT book. From this evidence Schechter concluded, "in the case of all the canonical books, with the doubtful exception of the Book of Daniel, these books must as a whole have been familiar to B.S., and must therefore be much anterior to him in date."¹⁸

Since Ben Sira wrote his book sometime between 200 and 180 BC, this evidence ruled out the Maccabean dating of Psalms, Ecclesiastes and other books. However, some scholars challenged Schechter's conclusion, maintaining that the medieval Geniza Ben Sira manuscripts were actually a retroversion into Hebrew from either Syriac or Greek versions. With the discovery of the scrolls at Qumran and Masada this argument proved false on three counts:

1. The first century AD Ben Sira manuscripts discovered at Masada show remarkable agreement with the Geniza manuscripts.

2. Non-biblical manuscripts from Qumran clearly demonstrate that the Hebrew of the Hasmonean era is different from that of the Psalms, Ecciesiastes, and other books.

3. The existence of Qumran manuscripts of Psalms, Ecclesiastes, Job, and other books, (some as early as the mid-second century BC) make it highly unlikely that these books could have been Maccabean compositions.

No one currently holds to a Maccabean date for any book of the OT, with the exception of Daniel, which many scholars still date to about 164 BC.

The third assumption, that the rabbis closed the canon about AD 90 at Jamnia, was discredited by Jack Lewis in l964.¹⁹ The Council of Jamnia assertion was built on a passage in the Mishnah indicating that the rabbis at Jamnia debated the status of two books, Ecciesiastes and Song of Songs-not about the Writings in general.²⁰ Nothing in this passage indicates that the rabbis were deciding the contents of the canon or seeking to close either the third division of the OT or the canon as a whole. In fact, the debate over some books of the canon continued for some time, as other passages in the Mishnah indicate. Leiman offers this conclusion about the supposed Council of Jamnia:²¹

In summary, all that can safely be said is that at a session of the academy at Jamnia that convened sometime between 75-117 C.E., it was decided that Ecclesiastes and the Songs of Songs defile the hands. (According to R. Akiba, a decision was necessary and made only with regard to Ecclesiastes.) These decisions were still being questioned 100 years later. The widespread view that the Council of Jamnia closed the biblical canon, or that it canonized any books at all, is not supported by the evidence and need no longer be seriously maintained.

Consequently, few, if any, scholars today maintain that a council at Jamnia closed the canon around AD 90.

The final assumption, that of the Alexandrian Canon Hypothesis, was thoroughly discredited by A. C. Sundberg in 1964, the same year that Lewis discredited the Council of Janrnia.²² Sundberg assumed that the Jewish canon was closed at Jamnia. Despite this flaw, his proof that no separate canon existed among Jews of the Diaspora remains valid.

The heart of Sundberg's analysis was a comparison of the lists of canonical books found in the church fathers to the contents of the great Septuagint codices. He demonstrated that little agreement could be found either in their contents or order, indicating that the church did not simply adopt a canon from Diaspora Judaism. Since the codices were only an indication of later Christian usage, they fail to prove the existence of a distinct canon that was used by Jews in pre-Christian Alexandria or anywhere else. As a result of Sundberg's work the Alexandrian Canon Theory has been completely abandoned.

Newer Theories about the Closing of the Canon

With the destruction of the foundation upon which the Triple Canon Theory rested, newer theories about the formation of the canon began to be proposed. These fall into two categories: theories that place the closing of the canon in the late first century or even later and theories that place the closing of the canon sometime before Jesus.

Late Date Theories for the Closing of the Canon

Many scholars continue to date the closing of the canon at the end of the first century AD, even though scholars acknowledge that it was not fixed by a council of rabbis at Jamnia. Instead, for many scholars the witness of Josephus to a canon of 22 books, the witness of the book of 4 Ezra (written late first century or early second century) to a canon of 24 books, and the appearance of lists of canonical books after that time would seem to make this the latest possible date for the closing of the canon. The primary argument against dating the canon earlier than the late first century is that no lists of books belonging to the canon can be dated earlier than the second century AD and no enumeration of the books can be dated prior to Josephus. Scholars who take this view include Eugene Ulrich²³ and Joseph Blenkinsopp.²⁴ Moreover, this view is commonly found in Bible dictionaries and encyclopedias.²⁵

John J. Collins offers a more detailed view of this theory.²⁶ Collins argues that the canon was closed for all Jews around the end of the first century, but that the canon chosen was a canon that took shape among one sect of Jews (presumably certain Pharisees) before the fall of Jerusalem in AD 70. This theory is a possible way of bridging the gap between the late date theory and the early date theories yet to be examined. Its weakness is that it can point to no date for the closing of this canon. Moreover, it assumes that there were differences among various Jewish sects in the first century BC and first century AD despite a strange silence in the sources concerning any friction between this one sect of late Second Temple Judaism and other sects that may have wanted to include or exclude other books.

The strength of this late date theory is that it places the closing of the canon at the time when enumerations and lists of canonical books were appearing. However, the weakness of this theory is that it fails to explain earlier evidence for the canon that appears to arrange the canon into a twofold (Law and Prophets) or threefold arrangement. Authors before the late first century seem to assume a fixed content.²⁷

In a more radical approach, Lee M. McDonald argues that the canon was not closed until much later.²⁸ He argues that in Jesus' day the Jews recognized an amorphous collection of books as sacred. This collection apparently had a core of books accepted by nearly everyone, but also contained other books on the fringes that were accepted by some but viewed suspiciously by others. For Jews, the closing of the canon began with the rabbis responsible for shaping the Mishnah around AD 200 and within 100 years resulted in the Hebrew OT we have today. For Christians the process took longer. Christians accepted some of the books eliminated in the process followed by the rabbis. Eventually church councils accepted the larger canon in the fifth century.

In this way McDonald can explain the two divergent canons. He also believes that his theory better accounts for the evidence. Chief among these are the continuing discussions about the status of some books of the OT by the rabbis as recorded in the Talmud and the reference to noncanonical books in the NT.

On the other hand, McDonald has to discount some evidence as unreliable in order to push the date of the canon's closing as back as far as he does. Most notable is the evidence from Josephus. According to McDonald the canon put forth by Josephus is partly exaggeration for apologetic purposes and partly Josephus' idiosyncratic view. Against Josephus, McDonald asserts that Jews generally accepted some books later deleted from the canon as of divine origin and possessing authority in Josephus' day.²⁹

Finally, we should note that other scholars have argued that the pluriform nature of Judaism in the first century (BC or AD) argues against any early date for a canon among Jews.³⁰ However, if pluriformity is an indication of lack of canonical formation, then modern Protestantism has no canon!

Early Date Theories for the Closing of the Canon

On the other side of the debate on the closing of the canon are scholars who believe the canon was closed before Jesus' day. Among these are David Noel Freedman. Freedman argues that, with the exception of Daniel, the canon was formed and closed in the time of Ezra.³¹ Fundamental to Freedman's approach is his view that the composition of each of the OT books is to be dated close to the last recorded episodes in it. Therefore, the last books of the canon were Chronicles and Ezra-Nehemiah (not counting Daniel). Freedman's theory would appear to agree with tradition that dates back at least to Josephus that the canon was closed during the Persian period.³² Indeed, modern scholars, in contrast to scholars in the nineteenth century, have tended to date all of the books of the OT except Daniel to the Persian period. Thus, Freedman's theory that the canon formed near this time is attractive, since there is no intrinsic need to date it later.

Freedman's argument for the closing of the canon is based on symmetry that he discerns in the books of the Bible themselves and in their arrangement in the canon. To achieve this symmetry Freedman divides the Prophets into two sections. The Former Prophets (Joshua, Judges, Samuel and Kings) are grouped with the Pentateuch to form what Freedman labels the Primary History. The Latter Prophets (Isaiah, Jeremiah, Ezekiel and the Twelve) are grouped with the Writings.³³ These two groupings are approximately the same size. The Primary History contains 149,641 words, whereas the remaining books (minus Daniel) contain 149,940 words.³⁴ This and other patterns in the OT convince Freedman that the canon was a conscious arrangement of books at the time of the composition of the last of them. That is, this symmetry could only be achieved once all of the canonical books were written and assembled in a collection for the first time. Therefore, twenty-three of the twenty-four books of the OT were canonized in the days of Ezra. Daniel, which Freedman dates to the Maccabean era, was added shortly after it was composed. Therefore, according to Freedman the canon was essentially formed by about 400 BC and took its final shape around 160 BC.

However, his thesis could be criticized for two weaknesses. First, his theory depends entirely on internal biblical evidence. He does not cite any evidence external to the biblical text for dividing the OT canon as he does. His evidence depends on his detection of certain patterns within the OT itself, but he offers little proof that his observations of these patterns are indeed what the editors/compilers of the canon had in mind.³⁵

Second, part of his evidence is the traditional arrangement of the Hebrew OT. I have already discussed the lack of hard evidence for this arrangement of the books before the early second century AD and will examine this in detail in the following chapters. Therefore, Freedman's theory remains debatable in its details.

Nevertheless, based on the historical arguments he adduces else-where as well as in his book on the Canon, Freedman certainly has made a strong case for the OT canon as formed early.³⁶ His theory that the entire canon was formed early not only is supported by historical evidence, but also by the currently scholarly consensus on the date of the composition or final editing of all of the OT books (with the exception of Daniel).

Several other scholars date the closing of the canon to Maccabean times. Sid Leiman marshals an impressive array of evidence from the Talmud and Midrash that point in this direction.³⁷ Roger Beckwith assembles Jewish and Christian evidence from the second century BC through the fifth century AD to argue, like Leiman, that the canon was closed by Judas Maccabeus around 160 BC.³⁸ E. Earle Ellis also believes the canon was closed before Jesus' day.³⁹

In all three cases these scholars rely on external witnesses to the canon, beginning with the second century BC. This evidence will be examined in the following chapters. However, all three of these scholars still rely on the assumption that the canon was accepted in the traditional threefold arrangement from early times. Therefore, as we examine the evidence on the canon we will need to carefully investigate the possible references to the divisions of the canon.

These early date theories have the opposite problem of the late date theory: They explain why the canon can be referenced by its main divisions as early as the second century. However, the supporters of this theory have yet to explain why the enumeration and listing of the books of the OT canon do not begin to appear until 250 years or more after the canon has been closed.


Toward a New Theory of the Canon

Since it is obvious that no theory takes into account or explains all the evidence, a new theory is needed. Such a theory should be able to offer a reasonable explanation for evidence before the late first century when references seem to indicate a canon and an explanation for references from the late first century forward that not only begin to enumerate and list the books of the canon, but also at times incorporate books not found in the present Hebrew OT. In order to do this, we will have to examine the data in chronological order. Chapter two of this study will examine the references to the canon before Jesus' time. Chapter three will investigate the evidence from the first century after the time of Jesus. Chapter four will look at the evidence after the first century. Along the way I will build a theory to explain the evidence. Chapter five will summarize the evidence and present a united theory on the OT canon.

1. The only possible exceptions are the ambiguous approach of the Eastern Orthodox churches, the wider canon of the Ethiopic church, and the Eastern Syriac canon. See Elias Oikonomos, "The Significance of the Deuterocanonical Writing in the Orthodox Church," in The Apocrypha in Ecumenical Perspective, ed. Siegfried Meurer. UBS Monograph Series 6 (Reading, UK: United Bible Societies, 1992), 16-32; Hans Peter Ruger. "The Extent of the Old Testament Canon," in The Apocrypha in Ecumenical Perspective. 151-60.
2. Traditionally Lutherans and Anglicans have retained the apocrypha and placed them between the Testaments as useful reading (see Luther's German Bible or the original arrangement of books in the King James Version). Both have at times included these books as part of the lectionary readings. However, due to the influence of Reformed and, later, evangelical Christians, especially in the United States and Canada, many Episcopalians and nearly all Lutherans in North America use Bibles without the apocrypha.
3. Hermann Wolfgang Beyer, Kavc~M' TDNT, 3.596-602.
4. James Barr, Holy Scripture: Canon, Authority, Criticism. (Philadelphia: Westminister 1983), 49.
5. Eugene Ulrich, "The Canonical Process and Textual Criticism," in Sha 'arei Talmon: Studies in the Bible. Qumran, and the Ancient Near East Presented to Shemaryahu Talmon. ed. Michael Fishbane and Emanuel Toy (Winona Lake, IN: Eisenbrauns 1992). 270.
6. Brevard S. Childs, Introduction to the Old Testament as Scripture. (Philadelphia: Fortress. 1979), 50.
7. Sid Z. Leiman. The Canonization of Hebrew Scripture: The Talmudic and Midrashic Evidence. Transactions of the Connecticut Academy of Arts and Sciences 47. 2nd ed. (New Haven: Connecticut Academy of Arts and Sciences, 1991), 14 - 15.
8. Lee M. MeDonaki The Formation of the Christian Biblical Canon. (rev, and expanded ed. Peabody. MA: Hendrickson, 1995), 20 - 21: also "The Integrity of the Biblical Canon in Light of Its Historical Development," BBR 6 (1996), 101 - 3; Gerald T. Sheppard, "Canon" in The Encyclopedia of Religion. ed. Mircea Eliade. (New York: Macmillian, 1987), 3.62-69.
9. Though he does admit that the Hebrew canon was recognized as authoritative by most Jews from the first century on. See McDonald, Formation, 21.
10. Noteworthy here is the observation by Roger Beckwith that "the failure to distinguish between the canon which a community recognized and used and the eccentric view of individuals about the canon" is one of the fallacies that has vitiated much of the writing about the canon. See Roger T. Beckwith, The Old Testament Canon of the New Testament Church. (Grand Rapids, MI: Eerdrnans, 1986), 8.
11. Philip R. Davies. Scribes and Schools: The Canonization of the Hebrew Scriptures. (Louisville: Westminster/John Knox. 1998).
12. Davies, Scribes and Schools. 9.
13. For example. Christians still accept Leviticus as authoritative, though not in the sense that it was authoritative for ancient Israel. For ancient Israel it was not only an authoritative guide to God's law and an authoritative presentation of the Gospel in the forgiveness offered in the sacrificial system, but it also authoritatively dictated specific worship practices. Though Christians do not view it as authoritative in dictating worship practices and other aspects of God's law, other authoritative functions (its moral underpinnings and its Gospel presentation) are still functional for Christians.
14. Herbert Edward Ryle, The Canon of the Old Testament: An Essay on the Gradual Growth and Formation of the Hebrew Canon of Scripture. 2nd ed. London: Macmillam, 1982.
15. McDonald, Formation, 30 "Ryle's only unreasonable proposal is his dating of the threefold development of the Hebrew Bible" and 49 "H. E. Ryle's theory that the three tiered OT canon gained recognition by the time when the so-called council of Jamnia met to discuss such matters has been challenged by a number of scholars. A three-stage development of the Jewish biblical canon, however, is not as unlikely as some have supposed, even though it has little direct evidence." (emphasis mine)
16. Leiman, Canonization, 66 - 67. Note: Luke 24:44 is often assumed to refer to the tradtional three sections of the Hebrew OT. However, it most likely does not. See the analysis of Luke 24:44 beginning on page 90.
17. Barr. Holy Scripture. 54 - 56; R. Laird Harris. "Was the Law and the Prophets Two-thirds of the Old Testament Canon?" BETS 9 (1966), 163 - 71. An early' challenge to the threefold division of the canon as indicative of the development of the OT was Willis J. Beecher, "The Alleged Triple Canon of the Old Testament," JBL 15 (1896), 118 - 28.
18. S. Schechter. and C. Taylor, The Wisdom of Ben Sira: Portions of the Book of Ecclesiasticus from Hebrew Manuscripts in the Cairo Genizah Collection Presented to the University of Cambridge by the Editors. (Cambridge: Cambridge, 1899), 35.
19. Jack P. Lewis, "What Do We Mean by Jabneh?"JBR 32 (1964), 125 - 32.
20. The debate was about whether these books "defile the hands." Many scholars understand the rabbinical concept of books defiling hands as a mark of sacredness and canonicity.
21. Leiman. Canonization. 124.
22. Albert C. Sundberg, Jr., The Old Testament of the Early Church. Harvard Theological Studies 20. Cambridge: Harvard.. 1964.
23. Eugene Ulrich, "The Canonical Process, Textual Criticism, and Latter Stages in the Composition of the Bible" in "Sha 'arei Talmon ": Studies in the Bible, Qumran and the Ancient Near East Presented to Shemaryahu Talmon. (Winona Lake, IN: Eisenbrauns, 1992), 267 - 91.
24. Joseph Blenkinsopp, Prophecy and Canon: A Contribution to the Study of Jewish Origins. (Notre Dame, IN: Notre Dame, 1977), 126.
25. E.g., "Canon" inABD, 1.840:
26. John J. Collins, "Before the Canon: Scriptures in Second Temple Judaism," in James Luther Mays, David L. Petersen and Kent Harold Richards, eds. Old Testament Interpretation: Past, Present, and Future. Essays in Honor of Gene M Tucker. (Nashville: Abingdon, 1995). 225 - 41.
27. One could claim that the New Testament's reference to "the Law and the Prophets" does not necessarily refer to a closed and well-defined collection of books, but that it is a general way of referring to a sort of proto-canon that had yet to be strictly defined. However, the Law was already a strictly defined collection of the five books of Moses. It would be a strange collocation if 'the Law and the Prophets" meant "five prophetic books from Moses and an undetermined number of books from an undetermined number of prophets."
28. McDonald, Formation and "Integrity."
29. McDonald, Formation. 53 - 58; "Integrity," 108 - 11.
30. David M. Carr, "Canonization in the Context of Community: An Outline of the Formation of the Tanakh and the Christian Bible," A Gift of God in Due Season: Essays on Scripture and Community in Honor of James A. Sanders. (JSOTSup 225. Sheffield: Sheffield, 1996), 22 - 64. Carr is following the approach laid out by James Sanders.
31. David Noel Freedman, The Unity of the Hebrew Bible. Distinguished senior faculty lecture series. (Ann Arbor: University of Michigan, 1991): also "The Symmetry of the Hebrew Bible," ST46 (1992) 83 - 108.
32. However, Freedman himself does not mention this tradition.
33. Freedman, Unity. 5.
34. Freedman, Unity, 79.
35. For a more detailed critique of Freedman's thesis see my review of The Unity of the Hebrew Bible in The Michigan Academician, 54 (1993), 108 - 9.
36. David Noel Freedman, "Canon of the Bible," in G. Wigoder, W. M. Paul, B. T. Viviano, OP. and E. Stern, eds., illustrated Dictionary and Concordance of the Bible (New York: Macmillan, 1986), 21 1 - 16: "The Formation of the Canon of the Old Testament: The Selection and Identification of the Torah as the Supreme Authority of the Post-Exilic Community." in E. B. Firmage, B. G. Weiss and J. W. Welch, eds. Religion and Law: Biblical-Judaic and islamic Perspective (Winona Lake, IN: Eisenbrauns. 1990). 315 - 33.
37. Leiman, Canonization; also "Inspiration and Canonicitv: Reflections on the Formation of the Biblical Canon." in Jewish and Christian Self-Definition. 2nd ed. E. P. Sanders, A. I. Baumgarten and Alan Mendelson, eds. (Philadelphia: Fortress, 1981), 2.56 - 63.
38. Roger Beckwith, The Old Testament Canon of the New Testament Church. (Grand Rapids, MI: Eerdmans, 1986); also "Canon of the Hebrew Bible and the Old Testament." in The Oxford Companion to the Bible. Bruce M. Metzger and M. D. Coogan, eds. New York: Oxford, 1993, 102 - 4; "Formation of the Hebrew Bible" in Mikra: Text. Translation, Reading and Interpretation of the Hebrew Bible in Ancient Judaism and Early Christianity. CRINT. M. J. Mulder. ed. Minneapolis: Fortress, 1990, 39 - 86: "A Modern Theory of the Old Testament Canon." VT41 (1991), 385 - 95
39. E. Earle Ellis, The Old Testament in Early Christianity: Canon and Interpretation in the Light of Modem Research. Grand Rapids: Baker, 1991; also "The Old Testament Canon in the Early Church," in Mikra: Text, Translation. Reading and Interpretation of the Hebrew Bible in Ancient Judaism and Early Christianity CRINT. M. J. Mulder and H. Sysling. eds. Minneapolis: Fortress. 1990, 653 - 90.

Taken from The Oracles of God, copyright 1999. Used by permission of Concordia Publishing House, St. Louis, MO 63118-3968. You can order The Oracles of God for a total of $20 by calling the Issues, Etc. resource line at 1-800-737-0172.