One of the most significant problems in the debate on canon formation is how we define “canon.” The term was not used to refer to a collection of authoritative biblical books until well after the time of Christ (circa 4th century C.E). The first rabbinic list that includes all the books of the Hebrew Bible is found in tractate Baba Bathra, folio 14b in the Babylonian Talmud (c. 2nd-3rd century C.E.) Albeit, the sequencing of the books is slightly different from the modern Masoretic arrangement. For example, Baba Bathra puts Jeremiah and Ezekiel before Isaiah, and Ruth at the beginning of the Writings instead of Psalms. An early Christian canon list from the 2nd century C.E. (Melito’s) is included in Eusebius’s Ecclesiastical History. Melito’s canon included all the books of the Hebrew Bible except Esther and he possibly added Wisdom of Solomon. Some scholars argue there was no canon until these lists were confirmed and closed. Of course, this definition requires the list to explicitly spell out all the books of the Hebrew Bible we accept today. This view tends to dismiss earlier references to canon lists (e.g. Josephus) that might have included but don’t explicitly state every book. Some scholars maintain that authoritative texts used for faith and praxis should be considered canonical even if the list is not closed.
Another significant debate involves the dating of the canon. When did the canon close? A prominent 19th century three-stage theory was championed by Herbert Ryle (among others). In this theory, the canon developed at three different time periods: the Torah closed by the 5th century based on the 432 B.C.E. date for the Samaritan schism, the Prophets were finalized around 200 B.C.E., and the Writings closed by 90 C.E. based on the alleged (now refuted) Jewish council at Yavneh.Most scholars no longer accept this three-stage theory. However, they have not discarded it entirely. The majority believe the Torah and Prophets were determined earlier on, with the Writings confirmed later. Some believe the Torah and Prophets were closed in the Persian period (5th century B.C.E.), others cite the Maccabean period (2nd century B.C.E.) Similarly, some scholars believe the entire canon including the Writings was closed by the 2nd century B.C.E. and others argue that the canon remained fluid even until the 2nd or 3rd centuries C.E.
Yet, despite disagreement on how the term “canon” should be understood and when the canon closed, virtually everyone agrees authoritative scriptures existed in antiquity for Jews and Christians. The term “canon” was not used by these faith communities, however, other terminology clearly referred to collections of sacred texts (e.g. “the writings,” “Moses and the prophets,” and “the holy books”). Many of these texts were later included in the official canon(s). G.T. Sheppard proposed the concept of Canon 1 and Canon 2. Canon 1 refers to collections of authoritative texts prior to the final closing of the canon. Canon 2 is the finalized, closed canon. However, some scholars only acknowledge Canon 2, preferring to categorize Canon 1 as simply “authoritative scriptures.” In order to unpack the possibilities for how to conceive the formation of the Old Testament canon, several scholars are summarized here.
Diverse Views on Canon Formation
Roger T. Beckwith
- A complete tripartite canon (Torah, Prophets, Writings) existed no later than 130 B.C.E. based on the prologue of Sirach (the translator refers to the Torah, Prophets, and “other ancestral books”). However, the canon likely closed even earlier in 164 B.C.E. At this time, Judas Maccabaeus gathered together scriptural texts (unspecified) after Antiochus Epiphanes attempted to destroy them (2 Macc 2:14). Judas probably finalized the list of Prophets and Writings (the Torah already long since closed).
- Some continued scribal activity may have occurred beyond 164 B.C.E. but the books remained canonical even as they were elaborated upon and completed.
- Jesus’ reference to the law, prophets, and psalms indicates he knew a tripartite canon (Luke 24:44). Philo also refers to the law, prophets, psalms and “other books” (On the Contemplative Life 25). The title “psalms” is meant to include all the Writings.
- The 1st century C.E. debate on the canonicity of Ecclesiastes and Song of Songs, as well as earlier rejections of Esther is insignificant. Disagreements were represented only by minority voices.
- Terminology, the temple archives, and general consensus on certain books are evidence for an early canon.
- There are many terms referring to authoritative texts, including “what is written” (Jn 20:9; Gal 3:22; I Pet 2:6), and “the holy writings” (Philo, Josephus, 2 Tim 3:15). These titles are from the 1st century C.E. However “the law and the prophets” is a century older (4 Macc 18:10; 2 Macc 15:9), and a tripartite canon is alluded to in the prologue to Sirach written by the author’s grandson around 130 B.C.E. Albeit, the third division of the canon is not officially titled (e.g. “other ancestral books”). The Letter of Aristeas indicates a closed Torah no later than 3rd century B.C.E., and even older biblical texts refer to the law of Moses (at least parts of it). This terminology clearly indicates Jews recognized a canon of sacred writings.
- These writings were stored in sacred space proving their significance. Texts were put in the ark of the covenant or Temple (e.g. Exod 25:16, 21; 40:20; Deut 10:1-5; 2 Kgs 22:8; 2 Chron 34:15, 30; 2 Macc 2:13-15). Texts such as genealogies and letters of the kings about votive offerings were kept in this library/archive as well. But non-canonical books would not normally be brought into the Temple. The Temple library included not only the Pentateuch, but other biblical books. The evidence for this is in Josephus’ Antiquities—a reference indicates the book of Joshua was placed in the archive. A later rabbinic reference of “the Fifths” means the Psalms and therefore all the Writings were kept in the Temple (T. Kelim Bava Metsia 5:8). The canon was compiled only by Temple scribes.
- Many Jewish writers refer to biblical books. Philo quoted from all five books of the Pentateuch using the conventional formulas for citing scriptural texts. Scriptural citations in the works of Josephus, the New Testament, and deuterocanonical/apocryphal texts (e.g. Tobit, Baruch, Sirach) demonstrates Jewish writers named all the books of the Hebrew Bible except Ruth, Song of Songs, and Esther. This suggests general agreement about what was considered Scripture. Not all the books of Josephus’ canon are specifically named. However, Josephus cites biblical books elsewhere in his work and only omits Ecclesiastes or Song of Songs.
- The beginnings of the canon reach back as far as Moses when God gave the Israelites the Ten Commandments (c. 14th century B.C.E.) However, the Israelite exile (c. 6th century B.C.E.) was the impetus for bringing together all the authoritative texts that were located in the temple, court, or prophetic circles. During this time the Pentateuch, historical books, and prophetic texts underwent redaction.
- Final editing of the Torah and Prophets probably occurred in the post-exilic period. The Writings were completed sometime later.
- The canon was fixed by late first century C.E. as evidenced by revision of the Septuagint toward a proto-Masoretic reading.
- The editorial process itself indicates the Israelites were conscious of their canon. By editing they were intentionally compiling and preserving sacred texts. For example, the reason superscriptions were added to the prophetic books was to classify them as revelation from God. The canon did not come together haphazardly. Instead the canonical shape of the books—the way books were sequenced—indicates deliberation. Not only were individual books redacted, but the shape of the Hebrew Bible itself is an intentional project as evidenced by intertextual devices. In fact, the concept of canon is precisely what led to prolific copying, resulting in diversity because of the realities of ancient text production. In other words, canon consciousness gave birth to pluriformity.
- The Ten Commandments that God gave Moses and the people on Mt. Sinai were authoritative and thus canonical from that time. The safekeeping of the Decalogue in the ark of the covenant confirms this. The Israelites also deposited sacred writing in the sanctuary (Josh 24:25-26; I Sam 10:25; 2 Kgs 22:8). Other biblical texts affirm the Israelites accepted the Torah as canonical. For example, Psalm 19 and 119 are preoccupied with the law of God. The prophets constantly refer back to Torah when they rebuke the people for not following it (e.g. I Kgs 18:40; Ex 22:19; Hos 4:2; Amos 1:3-2:5). They also saw their own words—revelation from God—as authoritative and to be preserved for future generations (Isa 8:16-21; 30:8; Jer 36).
- Indications of a bipartite canon (Torah and Prophets) date from early second century B.C.E. The deuterocanonical/apocryphal books Tobit and Baruch refer to the “law of Moses” and the “prophets of Israel” or “words of the prophets.” Other texts that refer to the law of Moses and the prophets include 2 Maccabees 15:9, 4 Maccabees 18:10, the Damascus Document (CD VII, 15-17), the Rule of the Community (IQS 1, 2-3), and the New Testament (e.g. Matt 5:17; 7:12; 22:40; Luke 16:16; Acts 13:15; 24:14; Rom 3:21).
- There is also evidence for a tripartite canon in Sirach (2nd century B.C.E.). Sirach refers or alludes to the majority of the books in the Hebrew Bible, and seems to discuss them in a sequence reflecting canonical consciousness. The prologue by the author’s grandson also refers to the law, prophets, and “other ancestral books.” The third canonical division is not given a name, but undefined terminology does not necessarily mean flexibility in canon. Another document supporting a tripartite collection is 4QMMT, dated to first or second century B.C.E., albeit, this is a reconstructed text. It reads: “the Book of Moses and the Books of the Prophets and David.” Similarly, 2 Maccabees 2 refers to the law, prophets and David. In the New Testament Jesus refers to the law of Moses, prophets, and psalms (Luke 24:44). Finally, by the 1st century C.E. Josephus refers to a 22 books canon that includes texts from the Writings (Against Apion 1.37-43).
- There was no Hebrew Bible canon before the end of the 1st century C.E., only a canon-in-the-making. Prior to the 1st century C.E. there was a canonical list for the Torah, and some books of the Prophets and Writings were confirmed. But since the list was still subject to fluidity at this time, there was no fixed canon yet. Agreement on all the texts to be included in the Hebrew canon did not occur until after the 1st century C.E.
- Use of the term “canon” is inappropriate until the closure of the list of books. A canon cannot be a canon if it is “open.”
- There was no Hebrew word for “canon” until long after the 1st century C.E. The term does not show up in Christian writings until the 4th century C.E.
- A collection of authoritative books used for religious practice and doctrine is insufficient to define canon.
- The canon is a result of the “canonical process.” This canonical process involved scribal editing, revising, and expanding as the Jewish people developed and passed on sacred texts generation to generation. In other words the canonical process is synonymous with the redaction process. Scriptural texts were formed, not at one particular moment, but over a long process wherein texts were reformulated to provide new meaning to new audiences. Thus, the canon was fixed only as the fluid texts gradually reached stabilization.
- The redaction process, and therefore canonical process, would have continued indefinitely, but ceased as a result of external factors such as Roman persecution and the rise of Christianity. Christian use of sacred texts raised new questions about what was authoritative and what was not. The creation of the canon was the result of reflexive judgment on the use of certain texts and confirmation of their authoritative status (or not).
Timothy H. Lim
- The Letter of Aristeas provides evidence of an established Torah by the 3rd century B.C.E. An even larger collection of scriptural texts was affirmed by the 2nd century in Sirach (and the prologue). Sirach references all the books of the Hebrew Bible except Ruth, Song of Songs, Esther, and Daniel. However, the canon was not closed as Ben Sira considered his grandfather’s own work to be part of that collection (and ultimately it was not included in the canon).
- The rabbinic canon, the basis for the Masoretic Text, was officially closed between 150-250 C.E. This is evident in the Baba Bathra list. However, the sequencing of the books of the Prophets and Writings was not stabilized until the 10th century with the Aleppo Codex.
- There was no three-stage canon formation. The Samaritan schism is not evidence of a closed Torah in the 5th century. The dating of the schism is inconclusive and thus unreliable. Similarly, Judas Maccabaeus’s unspecified collection of scriptural texts is insufficient evidence for an established library, and therefore a tripartite canon, in the 2nd century B.C.E.
- The Jewish temple library did not define the canon. Several different collections of scriptural texts were held by various groups before a common canon developed. The diverse collections include those of the Essenes, Josephus’ 22 books canon, the authoritative texts of the Therapeutae, as well as the Apostle Paul’s undefined biblical canon. Paul’s use of the Septuagint is exaggerated and therefore not indicative of adherence to a Septuagintal canon. However, Paul may have held to a Pharisaic collection.
- The canon did not develop from diversity into unity. Rather, the differences in collections held by various groups stemmed from what they were willing to include. All the groups had overlaps in the texts they affirmed, but some groups adopted a wider group of authoritative texts.
- There was a majority canon alongside sectarian collections. Ultimately, the scriptural collection of the Pharisees, the majority canon, won the day by the end of the 1st century C.E. and became the canon of Rabbinic Judaism (Masoretic Text). The Pharisees were the primary participants at the Yavneh gathering.
- However, the canon was not officially closed at the end of the 1st century C.E. There continued to be debates on the status of Ecclesiastes, Song of Songs, Ruth, Esther, and Sirach.
- Canon formation stemmed from a variety of factors, including: 1) many books had long been considered traditional scriptures as evidenced by study and liturgical use; 2) the influence of non-Jews (Josephus contrasts the Jews’ 22 reliable books vs. the Greeks’ inconsistent collections, the desire to distinguish Jewish sacred text from Hellenistic ones, and a reaction to Christian use of scriptural texts); 3) the biblical books are a product of transmission history—revision, re-writing, and editing; 4) Persian and Hellenistic imperial authorization (i.e. ruling kingdoms sometimes encouraged the use of local sacred texts/laws as a way of maintaining order).
Karel van der Toorn
- The canon first developed around 450 C.E. under Ezra who was commissioned by the Persian authorities to provide Judah with a national constitution after the exiles were allowed to return home. This imperial practice of local governance was not uncommon in the ancient Near East. The Prophets and the Writings were confirmed together two centuries later around 250 B.C.E.
- There is a two-stage canon formation not a three-stage. The first stage is grounded in the 5th century Torah of Ezra (with a nod to a 5th century Samaritan schism), and the Prophets and Writings were brought together simultaneously in the 3rd century. Justification for combining the Prophets and the Writings is Josephus’ reference to some of the Writings as “prophets.” Similarly, the refrain “Law and the Prophets” is a common designation for Scripture in the Gospels and also shows up in 4 Maccabees 18:10-19. However, there was not a finalized canon yet but all the ingredients were there.
- Significant editorial activity in the 2nd century B.C.E. also contributed to the formation of the canon. During the Hellenistic period there seems to have been a greater demand for national literature read by an educated public. During this period the scribes worked on preparing biblical books for publication. This national publication of particular books created a kind of canon without necessarily closing it.
- The Twelve Minor Prophets were published together on a single scroll by 250 B.C.E. (suggested by its messianic themes). Sirach (c. 200-180 B.C.E.) specifically mentions the Twelve (49:10). Copies of the Twelve at Qumran are dated to the first half of the second century.
- According to Josephus and other ancient Jewish commentators prophecy ceased after the time of Ezra. With no more prophecy, the making of scriptural texts would naturally reach its end. Thus, the scribes did not have to draft a definitive list to close the canon so much as announce that there would be no more books to add since revelation had ceased. The number twelve also suggests the Twelve Minor Prophets scroll was intended to convey completion of prophecy.
- Between 250 B.C.E. and 50 C.E. when Josephus provided his canon list, there was not a closed canon. Discussion was still underway as to which books prior to the cessation of prophecy were to be included. Some, for example, did not think Esther or Ecclesiastes were canonical. The canon was finally closed in the 1st century based on the consensus of Pharisees.
- Baba Bathra’s canon list preserves an earlier tradition seen in the 1st century works of Josephus’ Against Apion and 2 Esdras 14:44-46. Both 2 Esdras and Baba Bathra have 24 books (70 of the books of 2 Edras are secret; thus 24 public). Josephus counts 22, but the difference is not in number of books or content, but in the way books are divided. For example, some canon lists will cite certain books either together or separately (e.g. Ezra and Nehemiah or Ezra-Nehemiah).
- There are deficiencies in three contemporary proposals: canon as organic process, canon as library catalogue, and canon as scribal curriculum.
- In current scholarship there is a tendency to describe the canonization process as gradual, natural process without definitive and conscious decision-making on the part of the community of faith. Texts became part of the canon spontaneously with use and through constant revision as the texts were applied to new generations. This perspective is problematic because it amounts to a “self-authenticating word.”
- The “library theory” is also insufficient. The temple library catalogue did not comprise the canonical list. Evidence some scholars cite for this view includes reference to Nehemiah’s library (2 Macc 2:13-15), as well as other indications the Israelites deposited texts in temple archives (e.g. 2 Kgs 22; 1 Sam 10:25). There was certainly a temple library. However, it could only be accessed by authorized personnel. The texts were not available for individual, public use. Ancient libraries functioned more as storage rooms than our modern understanding of libraries. These libraries were not public, yet a canon by definition is public.
- We do not know specifically what the Jewish temple collection included. Second Maccabees indicates there were texts stored there that were not included in the canon. Thus, the library was more comprehensive than the canon list. Inclusion in the library did not automatically lead to canonization, but it’s probably true that inclusion in the library was a prerequisite for becoming canonical.
- Another insufficient canon theory is that of scribal curriculum. This view claims textbooks used in the schools became canonical. Unlike the library list that could be comprehensive, a curriculum was usually selective for practical reasons. The best and most honored texts were chosen for study. One problem with this theory, however, is that most scribal curriculums in the ancient Near East tended to be technical. The Hebrew canon is mostly literary. At the same time, the canon is likely influenced by scribal schools in a way that the temple library also paved the way for the canon. Scribal priests were the ones working with the sacred texts that eventually ended up on the canon list.
What are we to make of this evidence and various views of canon formation? I will address this question as well as draw overarching conclusions on this series in my next and final post.
 The Hebrew Bible is traditionally divided into three sections: the Torah, Prophets, and the Writings. The Torah is the first five books of the Bible. The Prophets refers to books attributed to prophets (e.g. Jeremiah, Amos, Isaiah, etc; the historical books like 1 and 2 Kings are called the “Former Prophets” to distinguish them from the “Latter Prophets” who prophesied closer to the time of the exile). The Writings refer to wisdom and other texts such as Proverbs, Job, Song of Songs, as well as the liturgical Psalms.
 G. T. Sheppard, “Canon,” The Encyclopedia of Religion (ed. M. Eliade; 10 vols.; New York: MacMillan, 1987) 3.62-69.
 Roger Beckwith, “Formation of the Hebrew Bible” in Mikra: Text, Translation, Reading and Interpretation of the Hebrew Bible in Ancient Judaism and Early Christianity (ed. by M. J. Mulder. Assen/Maastricht: Van Gorcum and Minneapolis: Fortress, 1990) 39-87.
 Stephen Dempster. “Torah, Torah, Torah: The Emergence of the Tripartite Canon” in Exploring the Origins of the Bible: Canon Formation in Historical, Literary, and Theological Perspective (ed. by Evans, Craig A. and Emanuel Tov; Grand Rapids: Baker Academic, 2008) 87-128.
 Eugene Ulrich, The Dead Sea Scrolls and the Origins of the Bible (Grand Rapids: W.B.