Articles and book excerpts used in and referred to on Issues, Etc.
Summary and Conclusions
from The Oracles of God
At the beginning of this study we surveyed
older attempts to explain the formation of the OT canon. The Triple Canon
Theory was the dominant explanation for the formation of the OT from the end of
the nineteenth century to the middle of the twentieth century. However, since
its demise several competing theories that divide roughly into two types-early
date theories and late date theories-have vied to replace it. None of these has
gained widespread recognition. Now that we have examined the evidence over six
centuries, the time has come to evaluate the reasons why no current theory has
gained widespread acceptance.
Methodological Problems with Current Theories on the OT Canon
Imposition of Later Evidence on Earlier Evidence
One fault of the method used to arrive at nearly all current theories of the OT canon is that they all read the earliest evidence for the canon in light of later evidence. The later evidence is used to draw conclusions about earlier periods without allowing for significant development or changes along the way.
This is especially true of the way that early date theories assume the second- and third-century AD evidence for the tripartite Hebrew canon ought to be imposed on earlier evidence. Thus, the Prologue to Sirach, Philo and Luke 24:44 are read as if they offer evidence for a three-part canon. For the Prologue this means misreading its evidence. Its mention of the books that Ben Sira studied does not require that the three categories be equated with the three sections of the Hebrew OT attested by later evidence. The phrase "the Law, the Prophets and the Psalms" as used in Philo and Luke is interpreted to be equivalent to the later-attested three-part canon despite the fact that in neither Philo or the NT is Psalms ever used as a synonym for the Writings and that shortly afterward Philo defines what he means by the Psalms. In addition, the evidence from 4QMMT demonstrates that this phrase probably should be understood as a variation of a twofold arrangement of the canon as Law and Prophets with Psalms meaning the book of Psalms alone.
Moreover, the first clear evidence of a three-part canon from Josephus is often dismissed as idiosyncratic since it does not match the later three-part canon. Indeed, a close look at the evidence in chronological order reveals the arrangement found in Josephus to be a step in the development towards both the three-part Jewish canon and the four-part Christian canon. This would have been more obvious had attention been paid to the chronological order of the evidence.
However, late-date theories of the canon suffer from the same problem in a different way. While they are less likely to impose the tripartite canon on earlier evidence, they demand something that the earlier evidence does not provide: a definitive list of books that comprise the canon. Since later evidence does provide this, the date of the closing of the canon is placed among this evidence in the late first century or later. Therefore, earlier evidence from Ben Sira through Josephus is discounted because it is less specific in its details. This is despite the fact that Josephus clearly and forcefully states that Jews long before his day possessed a definitive collection of Scriptures. It is contrary to the clear assumption of a canon in a number of places in the NT and the weight of the evidence as early as Ben Sira.
In essence, those who hold to late-date theories of the OT canon demand that the canon cannot have been accepted and acknowledged until its contents were listed. This is imposing a later view of the canon (i.e., that of a well-defined list) on the earlier evidence. However, some early evidence points to the canon as defined by an official Temple archive and not as a list (e.g., 2 Maccabees and Josephus). Thus, the late-date theories are also guilty of imposing conclusions from later evidence on the earlier evidence.
Moreover, these late-date theories purposely discount all evidence that present less than the maximum threshold for defining a canon. This is somewhat akin to a prosecuting attorney refusing to try a murder case without videotape of the murder clearly showing the perpetrator, the victim and the weapon. However, circumstantial evidence is relevant in proving the guilt of an accused lawbreaker beyond a reasonable doubt. In the same way, the evidence from times before the canon was explicitly defined as a list of books is relevant. When this evidence is properly examined, we can draw conclusions from it that leave no reasonable doubt about the existence of a canon and little doubt about its contents.
Using Post-Second Temple Definitions of Canon to Define the Canon During the Second Temple Period
This leads to the second problem with nearly all current theories: They attempt to use a later definition of canon as a list of accepted authoritative books as a part of their investigation of the canon. However, the concept of the canon as a list may be only a later development due to historical circumstances. Given the evidence, I would propose that the canon came to be a list as a result of one extremely important event: the fall of Jerusalem and its Temple to the Romans in AD 70. Before this time there was little need for lists to define the canon. The canon was the collection of holy, inspired, authoritative books in the Temple. The canon could be assumed to be known and acknowledged by most Jews because of this normative archive. (This explains the references in the NT which assume the existence of a commonly agreed upon Scripture.) Only with the destruction of the Temple did there arise a need to define the canon as a list that could gain common acceptance. Moreover, this need was addressed in two different ways by the two religious communities that valued it: Judaism and Christianity.
The problem arises when we insist that there was no canon until it was defined the way we presently define it. We could draw an analogy to the definition of the unit of length called a meter. Originally a meter was to be one ten-millionth of the distance from the equator to the North Pole along the meridian that runs through Paris. French scientists estimated this distance and divided it into ten-millionths. Later, when it was discovered that the earth was not perfectly spherical, the meter was redefined as the distance between two lines inscribed on a platinum-iridium bar kept in Paris. Later, when more precision and reproducibility was needed, the meter was redefined as a certain number of wavelengths of red light from a kiypton-86 source. Finally, in 1983 when scientists needed even more precision, the meter was redefined as the distance light travels through a vacuum in a certain period of time. We certainly would be in error if we held that no meter existed in the nineteenth century or most of the twentieth century because it was not defined the way we have defined it since 1983. In the same way, it would be an error to insist that no canon existed simply because it was not defined as a list-a definition which did not exist until the late first century at the earliest. Yet this is what most theories of the canon do. This is especially true of the late date theories, but it is also implicit in the attempts by those who hold to early-date theories when they attempt to impose later canonical organizations and lists on earlier evidence. However, the evidence from before AD 70 indicates that the canon, though it existed, was not conceived of as a list at this time.
Looking for Evidence for the Formation of the Canon Instead of Evidence for Its History Subsequent to Its Formation
A final methodological fault that is highlighted when the evidence is considered in chronological order is that scholars have often been looking for the wrong thing. The search has been for evidence of the formation of the canon or at least of the closing of the canon. However, none of the evidence appears to point in that direction. Instead, the evidence appears to present the history of the canon subsequent to its formation. As Schechter noted a century ago, Ben Sira appears to be a post-canonical book that makes extensive use of the canon. With the exceptions of Ruth and Esther (and possibly Daniel, if its composition is placed after Ben Sira), Ben Sira knows and uses all the books of the OT canon. As we have seen, the finds from Qumran reinforce this. From at least the last third of the first century BC this canon can be referred to as "the Law and the Prophets" or "the Law, Prophets and Psalms." Therefore, it is more likely that subsequent developments are not evidence of the formation of the canon, but of its history subsequent to its formation.
A New Theory of the History of the Canon
What is needed, then, is a theory of the OT canon that explains the evidence and accounts for the developments that are seen when the evidence is examined in chronological order. The theory I propose rests on three conclusions drawn from the evidence:
I. The OT canon was formed before the second century BC. The only possible exceptions are the inclusion of Esther and possibly Daniel (if a date of approximately 164 BC is accepted for its final form in Hebrew. Note that evangelical The canon is collected in the Temple archives and become accepted by Jews sometime in the late Persian period. The canon is in theory open, if another prophet should arise. However, in practice it is closed. It is probably in this period that the canon is divided into two parts: Law and Prophets. I Subsequent authors such as Ben Sira and Philo study the canonical books and base their writings on them. Some write pseudonymous books whose putative authors are mentioned in canonical books, especially the Pentateuch. Though valued by some Jewish groups, none of these are generally accepted by Jews as canonical and none is ever referred to as among the Law and the Prophets. scholars would date Daniel much earlier and would, therefore likely argue that Daniel should not be considered a possible exception). However, it would appear that the evidence for excluding Esther is mostly accidental except for later statements by some Christian Church Fathers.
2. Before the fall of Jerusalem to the Romans in AD 70 the canon was understood to be a set of holy books accepted as normative by all Jews. These books were collected in an official archive in the Temple.
3. After the fall of Jerusalem the Temple archives ceased to exist. During the following centuries the canon made a transition to a list of holy books accepted as normative. This transition took place in different ways among Jews than among Christians.
Figures 5 and 6 outline this theory in more detail.
The History of the OT Canon from 200 BC to AD 500
History of the Organization of the Canon
The Canon Was Formed Before the Second
While some have sought to place the formation of the canon as late as the third century AD and others as early as the mid-second century BC, nearly every source from Ben Sira onwards appears to presume the existence of a canon. Moreover, there appears to be broad agreement on the general contents of this canon. Some doubt can be raised about one or two books, but the existence of the general designation "the Law and the Prophets," a common phrase used by writers from diverse theological viewpoints, would appear to argue that a widespread consensus existed as to the contents of the Prophets as well as the Law. Moreover, the view that inspired prophecy in some sense ceased in the Persian period enjoys a long tradition from the second century BC to second or third century AD. Table 10 summaries the evidence:
Witnesses to the Cessation of Inspired Prophecy
|Witness||Type of Evidence|
|Ben Sira||Implicit (Ben Sira's Praise of Ancestors)|
|1 Maccabees||Implicit (prophecy had ceased)|
|New Testament||Implicit (perhaps Matt 23:34 - 36. Luke 11:49 - 31, the written prophetic revelation seen as distinct from apostolic revelation and continuing prophecy, e.g., 2 Pet 3:1 - 2, 15 - 16)|
|Josephus||Explicit ( Against Apion 37 - 43)|
|2 Esdras||Implicit (canon as 24 book collection from Ezra)|
|Rabbinic Writings||Explicit (t. Sota 13:2)|
Many of the items in Table 10 are implicit
testimony to the writing of the last books of the canon in the Persian period.
In and of themselves they would not provide much evidence. However, the weight
of the several implicit references combined with explicit statements,
especially of Josephus in the first century, tips the scales in favor of this
This evidence for the formation of the canon during the Persian period agrees with the conclusion of David Noel Freedman mentioned in chapter 1, although Freedman does not cite any of this evidence. Yet, Freedman's instincts appear to agree with the evidence. The only exception is the book of Daniel which, like a large number of scholars, Freedman dates to about 164 BC.
If Darnel was written about this date, it was quickly accepted into the canon apparently without controversy as a genuine book from the early Persian period. A copy of Daniel only about thirty years removed from 164 BC was found at Qumran. A large number of Jewish writings, including those from Qumran and Josephus as well as the NT accept Daniel as a genuine prophetic book written by Daniel the prophet in Babylon. Collins notes, "The Book of Daniel was widely accepted as a reliable and authoritative document by the end of the second century BCE.
If Daniel dates from as late as 164 BC and gamed swift acceptance as a holy book, a number of difficult, but ultimately unanswerable questions about its near-immediate acceptance by all Jewish groups need to be addressed. Unfortunately, exploring these questions is beyond the scope of this study. These include:
o Why did no Jewish sect-Pharisees, Sadducees, Essenes or others-dispute its inclusion in the canon? (Daniel would have received general acceptance about the time these groups were forming. Yet, as far as we know, no one disputed its prophetic credentials and it was given a place among the Prophets, unlike many pseudepigraphal books.)
o If Daniel 11 accurately portrays the career of Antiochus Epiphanes up to 11:39 but fails to accurately predict the Antiochus' death in the ensuing verses, why was it immediately received as prophecy and not quickly rejected as false prophecy soon after 164 BC when it became apparent that Antiochus did not die the way the book predicts?
o How was Daniel placed among the Prophets when other books of similar vintage claiming prophetic origins were not? How did a well-respected book like Ben Sira escape being placed among the Prophets? (I know of no Jewish source that treats Ben Sira as a prophetic book.)
o How are the apparent uses of Daniel in Ben Sira to be explained? o How is it that material attributed to Daniel (much of it in Hebrew, to judge from Qumran) was not included in the Hebrew book? Were these written after 164 BC? If Daniel as we have it in Hebrew is a composite work, when did the addition of newer material end, and why did it end?
However, evangelical scholars will have no
problem with holding that Daniel was part of the canon from an early date,
since they advocate a date of composition in the Persian period. A Persian
provenance is the clear claim of the text of Daniel. Compelling evidence
supports that assertion, especially for readers who regard the text's claim as
superior to post-enlightenment construals and theories of origin.
Nothing in this study resolves the question of the book of Daniel, though the evidence from Ben Sira suggests that Daniel may have been written earlier than 164 BC. However, only two alternatives exist for its inclusion in the canon:
1. It is earlier than supposed by many scholars, perhaps as early as the Persian period if evangelical scholars are correct.
2. Its final form dates from the second century and in this form it was swiftly accepted as a book from a genuine prophet of the late Babylonian and early Persian period without any known objections from any Jewish group or sect.
Thus, the canon was closed for all practical
purposes in the Persian period. If Daniel was a later composition accepted into
the canon, it was because it was understood to be a genuine work from that
time. However, the absolute closing of the canon took place with the fall of
Jerusalem and the accompanying loss of the Temple's archives. While the canon
had been closed in practice before this, no books could possibly be added to
the archives of holy books after this time.
It is interesting to note with Barton, that with the exception of Sirach, no book of the Hebrew canon or the wider Greek canon has as its putative author anyone living later than Ezra.243 Barton adds, "There are no recent upstarts among the authors of books that were taken seriously as prophecy in the years around the turn of the era. And so far as I know there is no case before about the third century AD of a Jewish work pseudonymously attributed to anyone later than Ezra."244 Therefore, although much of the evidence for the effective closing of the canon is implicit (though some is explicit), there can be no doubt about the tradition that the canon was first assembled and recognized in the Persian period. This tradition is as old as our evidence, is long-standing, and is persistent.
Before AD 70: The Canon as an Official Archive in the Temple
Both 2 Maccabees and Josephus speak of books stored in the Temple. However, it is not until after the fall of the Temple that we have the first evidence for the number of books in the canon or for its subdivisions (Josephus, 4 Ezra) or of a list of books in the canon (the Talmud, Mehito). This fact has been used by many who would date the final acceptance of the canon to the late first century or later to argue that no canon existed prior to the times of these lists, even though something very much like the present Hebrew OT is called "the Law and the Prophets" in earlier times.
What has not been noted heretofore is the significance of the Temple archives in shaping how explicitly the canon was defined. As long as the Temple archives existed, phrases such as "the Law and the Prophets" or "the Law, the Prophets and the Psalms" were adequate. it is likely that no one felt a pressing need for further definition. Only after the fall of the Temple, when an official archive could no longer be maintained, was the need for such a definition pressed upon Jews and Christians.
Many, especially among those who view the acceptance of the canon as a late development, argue that the Prophets was not a well-defined collection, but a fluid category that may have included books for some that others rejected (e.g., Jubilees, 1 Enoch). This position is maintained because we do not possess a list of the books belonging to the Prophets. However, we should note that in most of our sources we do not have a list of the books of the Law, yet no one has serious doubts about its contents. The Law is not defined by Ben Sira, 1 Maccabees, 2 Maccabees, 4 Maccabees, the NT, or any Qumran document. Nevertheless, almost all scholars are comfortable with understanding references to the Law in these documents as the five books of Moses, Genesis through Deuteronomy.
After AD 70: The Canon as a List
The situation after the Roman capture of Jerusalem required a shift in thinking about the canon. No longer a collection in the Temple archives, the canon was more susceptible to sectarian tinkering-both additions and subtractions could have been foisted upon Jews or Christians. Therefore, the canon needed to be more tightly defined. By the end of the first century it was beginning to be defined according to the number of books it contained-either twenty-two or twenty-four books, depending on how some books were combined. In the second century the books were not only numbered but also listed.
From this perspective, it makes little sense to argue about which way of listing and organizing the books is older-the Christian twenty-two books in four divisions or the Jewish twenty-four books in three divisions. While the roots of the Christian twenty-two book enumeration appears to be slightly older (Josephus and Palestinian Jews), so do the roots of the three-division scheme followed by Jews. However, the final products of the differing Christian and Jewish organization of canon appear to be the results of parallel developments. In fact, from the common groupings of books in each, it would appear that they are both children of a common heritage of canonical organization that may be much older. Note the groups that they have in common as presented in Table 11
Groupings of Books in the Jewish and Christian Canons
|Group||Place in Jewish
to the Talmud
|Genesis - Deuteronomy||Torah||Pentateuch|
|Joshua - Kings||Early Prophets||Historical Books|
Ezekiel, the Twelve
|Latter Prophets||Prophetic Books
(placed after the
|Job, Psalms, Proverbs
Ecclesiastes, Song of Songs
Books (placed before
the Prophetic Books)
|Note: The Jewish Bible does not include Ruth in the Joshua-Kings group, the Christian OT does. The Christian OT includes Lamentations with Jeremiah. In the Talmud the order of the first two Poetic books is reversed (i.e., Psalms then Job). As noted above in chapter 4, Job was one of the last books to find its place in Christian lists.|
Most of the books of the Hebrew canon are
included in Table 11. Only Ruth, Chronicles, Ezra-Nehemiah, Esther and
Lamentations are grouped differently. Of these, it would appear that the
Christian order has the slightly older claim that Ruth is to be grouped with
Judges and Jeremiah with Lamentations. The other books are placed differently
in the two canonical arrangements because of different views of how the canon
ought to be organized. The Christian canon considers Chronicles, Ezra-Nehemiah,
and Esther historical books whereas these are placed in the Writings in the
Jewish canon because they were not part of the Sabbath liturgical readings.
Thus, even the order of the books in present-day Christian and Jewish canons points to an incipient organization of the canon that appears to be older than either one and perhaps antedates the fall of the Temple. Josephus' testimony could perhaps be seen as evidence that this ordering of the canon was already in progress in his day.
Finally, the wider Christian canon is a development that is due both to the shift in the canon from archival collection to list and to popular usage in the church. Even at this, there is no evidence that predates Origin in the late second century for including any material not found in the Hebrew canon and no evidence until the fourth century in the West for including books not somehow considered part of the canonical books of Jeremiah, Daniel, or Esther.
From this survey we are left with one question. Which OT canon is the OT canon? This, of course, was not a question that this study was designed to explore at the outset, but it is, nevertheless, one which ought to be briefly addressed. For Jews there is no question. No conclusive evidence exists that Jewish traditions ever accepted any books as canonical outside the twenty-four books of the Hebrew Bible.
For Christians, however, the question is more complicated. For those Christians who view the canon as an inheritance from the Jews of Jesus' day, for whom the view of the canon could be expressed by Paul's words that "they were entrusted with the oracles of God" (Rom 3:2), the answer is the shorter Christian canon that corresponds to the Hebrew Bible of Judaism. This, of course, is the attitude of Jerome and the Eastern Fathers and, later, of the Reformers, for whom Scripture was the source of theological truth. This is consistent with their principle for authority in the church. But for Christians who value the church's continuing traditions as a source of truth alongside Scripture, the answer is not so simple. Here Augustine's attitude will prevail as it did in the West, so that Roman Catholics and others will hold to a larger canon based on the church's prerogative to define its own canon even if this cannot be demonstrated to be the canon of the Jews in Jesus' day. Indeed, for Christians who value the continued witness of the church or its traditions during the past twenty centuries as an equally important source of truth, this is a consistent position also.
Thus, for Christians the question of the extent of the OT canon will never be a purely historical question, but is bound up with the question of the source of teaching authority in the church. If teaching authority depends solely upon the prophets and Jesus and his apostles, only one canon is possible. If teaching authority extends beyond Scripture to tradition, then the canon is defined by the tradition one accepts.
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.
Management Techniques Incorporated
has provided this article archive expressly for Issues, Etc. The articles in
this archive have been formatted converted for internet use, by Management
Contact MTI webmaster