Beckwith on the Order of the Hagiographa

Roger Beckwith's book on the Old Testament canon has held a prominent place in the discussion since it was published in 1985 (The Old Testament Canon of the New Testament Church, Eerdmans). Nevertheless, one of its odd theses--that Judas Maccabeus is responsible for the canon list found now in b. Baba Bathra 14b--can be difficult to evaluate (other than simply declaring it 'odd') because the arguments supporting it are scattered about on many pages. (Beckwith was rather wordy in this book, resulting in a 500-page tome.)

This post is my attempt to make Beckwith's argument more straightforward, and thus to see whether it has merit. I wont be able to offer much evaluation here; that would make the post even longer than it is. But I don't think I hide what I think about the idea.

Most of the main argumentation appears in Beckwith's ch.4, "The Structure of the Canon," pp. 110-66, with notes on pp. 166-80. It goes more-or-less like this:

1. The tripartite canon (Law, 5 books; Prophets, 8 books; Writings, 11 books = 24 books) was established by the second century BCE (pp. 111-18). Evidence:

(a) Sirach prologue (pp. 110-11), which mentions three times "Law, Prophets, Other Books," or some similar formula.

(b) Luke 24:44 (pp. 111-15), which mentions "Law, Prophets, Psalms." Beckwith spends some time arguing that "Psalms" = Writings, that is, that "Psalms" in Luke 24:44 refers not just to the Psalter but to the 11 books of the Writings as a unit. His argument for this is necessarily based on the idea that when Luke 24:44 refers to the "Prophets", it must mean the 8 books of the Prophets according to the current divisions of the Hebrew Bible. He does not even question this (on which, see Barton), and the same goes for his treatment of Sirach's prologue. His argument that "Psalms" can mean "Writings" goes as follows: (1) it would be odd for Jesus to omit the books of the Writings from his canon in Luke 24:44 since he quotes from Daniel and the like so frequently; (2) the prologue to Sirach does not have a set title for the third section of the canon; (3) Philo uses hymnoi as the title for some books beside the Law and Prophets, and this is probably also the third section of the canon; (4) the tenth-century Arabic writer al-Masudi spoke of the "Law, the Prophets, and the Psalms, which are the 24 books"; (5) some passages in rabbinic literature (t. Kelim 5.8; y. Meg. 2.3; Sepher Torah 2.3-4; Sop. 2.4) mention the "Fifths," which might mean the Psalter (divided into five books), which might stand for the entire collection of Writings.

(c) Matthew 23:35 // Luke 11:51 (p. 115), which implies, on Beckwith's interpretation, that Chronicles constituted the last book of the canon, which again might imply that the canon was tripartite.

(d) Philo, De vita contemplativa 25 (pp. 115-18), which mentions the Law, Prophets, and Hymnoi (and other books), and hymnoi means again the entire third section of the canon.

"It is thus a well-attested fact that, by the first century AD, the division of the canon into three groups of books was widespread in the Jewish world, and that it was familiar to Jesus" (p. 118).


2. The distribution of books between the Prophets and Writings was the same as in the modern Jewish canon and in the Talmudic baraita found in b. Baba Bathra 14b (pp. 118-27).

The sources discussed so far do not list the books in the different divisions of the canon, so for that we must rely on Josephus, Jerome, and the Talmudic baraita. Josephus (Against Apion 1.37-43) has a canon of 22 books, divided into 5 books of Law, 13 books of Prophecy, and 4 books of "hymns to God and precepts for the conduct of human life." On the other hand, Jerome (Prologus Galeatus) and the Talmud both distribute the books more-or-less the same (Jerome attests two ways of distributing books, both very similar to each other, one exactly equivalent to that in the Talmud). Beckwith favors Jerome & the Talmud over Josephus in terms of representing the earliest and best distribution because:

(a) Jesus (Matt. 23:35 // Lk. 11:51) implies that Chronicles concludes the canon, as it does in the Talmud, to which arrangement Jerome nearly attests, while Josephus is way off.

(b) The Sirach Prologue implies that late historical works (e.g., Daniel, Esther, Ezra-Nehemiah, Chronicles) are included in the third section of the canon, which is impossible in Josephus' arrangement. You'll have to read Beckwith's discussion for how he draws this inference from Sirach's prologue (p. 123).

(c) Josephus' arrangement can be explained as a modification of the Talmudic arrangement, but not vice versa.

3. Judas Maccabeus divided the canon into its three sections and distributed the books according to the manner of the baraita in Baba Bathra 14b (pp. 152-53).

Beckwith claims that the division of the non-Torah books into two distinct sections--Prophets and Writings--took place exactly in the mid-second century BCE, because while the Sirach prologue attests to its existence, the lack of an exact title for the third section--the prologue simply calls it 'the other books'--means that this third section cannot have long existed (p. 142, 152).

"The exact date of the division into Prophets and other Books was probably about 164 BC" (p. 152)

 For this surprisingly precise dating of the arrangement of the canon, Beckwith cites 2 Maccabees 2:13-15.

(13) The same things are reported in the records and in the memoirs of Nehemiah, and also that he founded a library and collected the books about the kings and prophets, and the writings of David, and letters of kings about votive offerings. (14) In the same way Judas [the Maccabee] also collected all the books that had been lost on account of the war that had come upon us, and they are in our possession. (15) So if you have need of them, send people to get them for you.

 How can Beckwith possibly see in this report the arrangement of the tripartite canon of scripture?

Judas knew that the prophetic gift had ceased a long time before (1 Macc. 9.27; cp. also 4.46; 14.41), so what is more likely than that, in gathering together the scattered Scriptures, he and his companions the Hasidim classified the now complete collection in the way which from that time became traditional, dividing the miscellaneous non-Mosaic writings into the Prophets and the other Books? (p. 152)

And that is his argument. [Maybe you think I've left out part of it? No, I have not.] Beckwith further speculates that since Judas could not have collected all the scriptural books into a single volume, thus resulting in a fixed arrangement, he "must have done it primarily by compiling a list" (p. 153).

One last stage in the argument: though we now know that Judas Maccabeus divided the canon into three sections and assigned the books to the three sections precisely in accordance with the baraita of Baba Bathra 14b, we have not yet established that the exact sequence of books reported in the baraita derives from the action of Judas. But Beckwith can also prove this. 

4. The internal order of the Prophets and Writings as reported in b. Baba Bathra 14b goes back to Judas Maccabeus (pp. 156-62).

(a)  Beckwith asserts that "Judas's division of the miscellaneous non-Mosaic books"--which has in a few pages gone from a bizarre proposal to a fact of history, apparently--"was a deliberate and rational act," and so the order thus created should show signs of this, as the baraita does, Beckwith thinks (p. 156).

(b) The concluding position of Chronicles, to which Jesus attests (Beckwith thinks; see esp. pp. 211-22), corresponds to the order transmitted by the baraita (p. 156).

Beckwith attempts to illuminate the rationale for the division and sequence of books. He says that each section--Law, Prophets, and Writings--contains two types of books: history and another type. The Torah contains history and law; the Prophets, contain history and prophecy; and the Writings contain history and wisdom. The history books for each section narrate progressively later history. Beckwith then discusses what he perceives to be the problems to this account of things (pp. 158-62), which I do not feel like surveying right now.

But let me just note one of his points. The reason Ruth is in the Writings and not after Judges in the Prophets (as in other arrangements of the books, reflected now in the Christian arrangement of the OT) is because it serves as an introduction to the Psalter, since it ends with a genealogy of David. Also, the reason Chronicles is in the Writings rather than with the other history books in the Prophets (as in the Christian arrangement of the OT) is because it stands "at the end of the canon as a recapitulation of the whole biblical story, from the Creation [Adam] to the return" (p. 158). He goes on to say:

It follows that the presence of Ruth and Chronicles in the Hagiographa cannot be isolated from the order in which they stand in the Hagiographa, and that the same hand which divided the Hagiographa from the Prophets must also have determined the internal order of the two sections. It further follows that the hand responsible must have been later in date than the Chronicler, who apparently intended his work to precede Ezra-Nehemiah, not to follow it, while of course being earlier than the grandson of Ben Sira, who refers to the existence of the three sections of the canon in his prologue to Ecclesiasticus. (pp. 158-59)

Strangely, some scholars would disagree with Beckwith by saying that he dates all of this too late, and instead it was the Chronicler himself, in the Persian era, who put the finishing touches on the canon (Julius Steinberg, his teacher Hendrik Koorevaar). Others (e.g. Georg Steins) would also attribute this action to the Chronicler, but would date him to the Maccabean date in line with Beckwith's timetable. More on them some other time.

Those are the fundamental points of Beckwith's argument that the baraita of b. Baba Bathra actually goes back to the second century BCE, nay, even to 164 BCE and the action by Judas Maccabeus. His next chapter--ch. 5, "The Order of the Canonical Books"--is mostly taken up with arguing against other ancient orders, such as reported in patristic lists and alternative Jewish orders. I might have more to say about this chapter some other time.

Qoheleth and Sukkot

The Book of Qoheleth (a.k.a. Ecclesiastes) forms one of the Five Megilloth in the third section (Ketuvim = Writings = Hagiographa) of the modern Jewish biblical canon. Since Medieval times (twelfth century, or so), it has been customary in many Jewish communities to assign each of the Five Megilloth to a particular holiday, so that it is read in the synagogue on that day.

Song of Songs during Passover
Ruth during Shavuot (Pentecost)
Lamentations at the Ninth of Av
Esther at Purim
Qoheleth during Sukkot (Booths)

There is a fairly clear relationship between each book and its holiday. Esther, of course, describes the origins of Purim. Lamentations works well on a day commemorating the two destructions of the Jerusalem Temple (by the Babylonians and by the Romans). The story of Ruth takes place during the Barley harvest (Ruth 1:22), which would correspond to the time of Shavuot. Song of Songs traditionally receives an allegorical interpretation according to which it recounts God's love for Israel, especially as displayed in the exodus from Egypt, thus befitting Passover.

However, there is not an obvious relationship between Qoheleth and Sukkot, and I don't believe there's any traditional interpretation for the relationship. It's been a few years since I tried to gather opinions on this issue, and now I can't find my notes on it. So, I thought I'd start collecting some opinions here.

First, Wikipedia's article on Ecclesiastes says this (under "Traditional Judaism"):

It is read on Sukkot as a reminder to not get too caught up in the festivities of the holiday, as well as to carry over the happiness of sukkot to the rest of the year by telling the listeners that without God, life is meaningless.

Marvin Sweeney's new book Tanak: A Theological and Critical Introduction to the Jewish Bible (Fortress, 2012) contains these thoughts:

Qoheleth is read on Sukkot to explore the transitory nature of human life while affirming commitment to G-d. (p. 24)

Because the book of Qoheleth reflects on the transitory nature of life, it is associated with the festival named after the transitory life that Israel led at the concluding harvest during the wilderness wanderings. (p. 438)

Neither Sweeney nor the Wikipedia article cites any ancient or medieval Jewish authority for these opinions. That corresponds to what I recall from the last time I looked at this issue.

I'm not going to look further right now, but at some point I might add to this post with further opinions. 

One more thing: there's also the issue of when the Five Megilloth became connected to the festivals, and when they were grouped together as a subsection within the Ketuvim in our manuscripts. As I recall, there are different answers to this as well. I think some scholars believe that the liturgical connection to the holidays came first, and they were subsequently grouped together in the manuscripts to reflect their use at festivals. Other scholars argue or assume the opposite, that they were grouped together in the manuscripts first, and this became a reason to associate all five of the Megilloth with particular holidays, especially once 2 or 3 of the Megilloth became associated with their own holiday. This would offer a particularly nice explanation for how Qoheleth became associated with Sukkot (not a suggestion original with me).

Does Matthew 1 Narrate a Virginal Conception?

Yesterday, I received the alert that in the new issue of JSNT appears an article by Andrew T. Lincoln, "Contested Paternity and Contested Readings: Jesus' Conception in Matthew 1.18-25," 34.3 (2012): 211-31.

The article evaluates two ways of reading Matthew 1:18-25 with regard to the issue of the virginal conception. The traditional view sees in Matt. 1 a virginal conception because Matthew calls Mary a παρθένος, emphasizes that the product of her womb is begotten of the Holy Spirit, and informs his readers that Joseph did not have sexual relations with Mary until after the birth of the child.

The revised view (represented by Jane Schaberg, David Catchpole, and Robert J. Miller) does not see a virginal conception here, and charges the traditional reading with being unduly influenced by data outside Matthew's text, such as Christian tradition and the Lucan infancy narrative, which definitely (I think) contains a virginal conception (Luke 1:34; of course, yesterday I would have said that Matthew definitely contains a virginal conception). According to the revised view, the word παρθένος does not always mean virgin (more on this below), the Holy Spirit's involvement in the conception does not eliminate the possibility that a human father was also involved (see, e.g., Gen. 4:1-2; Judg. 13), and the reason for stressing that Joseph did not have sex with Mary until after Jesus was born was not to ensure a virginal conception, but to highlight once again (cf. Matt. 1:19) that Joseph was a righteous man, who therefore avoided the impurity of sexual relations with his pregnant wife (cf. Josephus, Against Apion, 2.199, 202).

At the end of the article, Lincoln offers this conclusion:

The provisional conclusion of this article, then, is that the revised reading needs to be taken seriously as a minority report that raises significant questions about the traditional reading, questions that should cause the latter's adherents to re-think its justification, but that, on balance, it is not compelling enough to make them abandon it. Matthew, though obliquely, probably remains a witness to a virginal conception. But, with apologies for the puns, the article will have succeeded in its aim if it has raised suspicions about the legitimacy of both the revised and the traditional readings of Jesus' conception in Matthew and planted the seed for further reflection and discussion. (p. 229)

Lincoln thus seems to lean toward the traditional view, which apparently represents something of a revision of his own position (see the conclusion here). 

But what about the word παρθένος? Does it really not mean virgin? Let me first note that Lincoln's article re-affirms my own interpretation that Matthew's main reason for citing Isa. 7:14 was not the word παρθένος--that is, not as a prediction of the virginal conception--but because of the child's name, Immanuel, for this child of Joseph and Mary would truly be "God with us".

Lincoln can cite some evidence that παρθένος does not mean 'virgin' but simply 'young woman' (p. 215 n. 10). This includes some Classical Greek references, but also a couple of instances in the LXX (Gen. 34:3; Joel 1:8). The idea is not that παρθένος never means 'virgin', but that this is a specialized meaning of the word that usually has not implications regarding sexual experience. It turns out, then, that the Greek translators of Isaiah got their translation correct, and, on the revised reading of Matthew, the Evangelist cited the verse without meaning to stress the virginity of the παρθένος.

[Note, however, that Lincoln gives two contradictory explanations for παρθένος in LXX Isa. 7:14. On p. 215, he says: "The LXX translators do not, then, change the force of the Hebrew. [...] [I]t makes perfectly good sense to read παρθένος [in Matt. 1:23], in line with the Hebrew and the LXX, as having its more general meaning of 'young woman' [...]."

But on p. 221, Lincoln writes: "There is, then, still no decisive reason against thinking that Matthew understood Isa. 7:14 in the way the LXX does: as referring to one who was at the time a virgin but who would conceive naturally."

I think this latter view--LXX Isa. 7:14 means that a woman who is now a virgin would conceive naturally (thus being a virgin no longer) and bear a son--was the one supported in R. Brown, Birth of the Messiah. But it is different from saying that παρθένος is actually the correct translation of almah, as Lincoln seems to say on p. 215.] 

Lincoln purposefully omits Luke from discussion, because he wants to focus on how we read Matthew apart from Luke. But Luke might give some support to the revised reading of Matthew, at least in regard to the meaning of παρθένος. About Mary's virginity, Luke is much more explicit than Matthew. In the crucial verse that establishes beyond doubt that Mary is a virgin (Luke 1:34), Luke does not rely on the word παρθένος (which he had already used about Mary; 1:27), perhaps because he knew it to be ambiguous. Instead, he has Mary say, ἄνδρα οὐ γινώσκω. I think this might support the idea that παρθένος cannot carry the meaning 'virgin' without the help of context.

Nevertheless, at the end of the day I still agree with the traditional interpretation of Matt. 1. Certainly, Lincoln's article has helped me to be a more critical reader by identifying some presuppositions that I bring to the text. Yet, if I were presented with Matthew's Gospel with no prior knowledge of Christianity, I think I would understand it to be asserting that Jesus was conceived without any male human agency, before Mary had sexual relations with any man, thus a virginal conception.

'Footprints' Is Not in the Bible

Just a thought, based on a comment from a chapel presentation last week by Nathan Daily.

Too often contemporary Christian spirituality relies on extra-biblical writings such as devotional books or 'poems' (is it a poem?) like 'Footprints'. One of the major problems with this is that Footprints is so theologically bland. Certainly it is comforting to think that God carries you through the hard times of your life. But is it even correct to say this? Does scripture support such an assertion. I think, rather, that the Psalms might accuse God of exactly the opposite (cf. Psa. 44; 88; etc.).

The attraction to something like 'Footprints' is that it is easy to understand and offers no challenge whatsoever. But it should give us pause when we consider that this description cannot in any way apply to the Bible. God has given us scripture which is profoundly difficult, both to understand and in the challenge that it presents. It does not seem that God is interested in giving us easy, bland theology. He is not interested in representing his relationship with us simply as one of his carrying us through the hard times. What would Job have made of such a theology?

So, once again, we must challenge the church actually to read scripture. If we want to understand God, and ourselves, 'Footprints' is not going to get it done. Rather, Ezekiel, or Isaiah, or the Chronicler, or Paul, all of these theologians have something to say about God, but one can benefit from them only through struggle. But, really, precisely for that reason they are of exceedingly great value.

Article on Jerome's Canon

I just received the proofs for my article entitled "The Old Testament 'Apocrypha' in Jerome's Canonical Theory," to be published in The Journal of Early Christian Studies 20.2 (2012): 213-33, forthcoming in June, I believe.

I don't think I've mentioned this article on this blog before. It is a greatly expanded form of a few pages from my dissertation that did not make it into the published form of my book. I presented the paper at the 2010 SBL, and then submitted it to JECS. Here's the abstract.

In his Preface to Samuel and Kings (the Prologus Galeatus), Jerome sets forth a theory of the Old Testament canon that allows for no room between the canonical books and the apocrypha. However, Jerome elsewhere maintained a more neutral or even positive view of some of the non-canonical books, even accepting their use within the ecclesiastical liturgy. Jerome’s seemingly inconsistent attitude toward some books he classifies as “apocrypha” has led scholars to posit a development in Jerome’s canonical theory, such that his earlier position was accepting of books that he later excluded, and to suppose that Jerome’s use of the word “apocrypha” in the Prologus Galeatus relied on a neutral definition of the term. This paper examines the evidence for these claims and finds them wanting. While Jerome consistently regarded the books labeled “apocrypha” in the Prologus Galeatus as outside the canon, he chose to propagate an especially harsh judgment against these books especially in this preface. The confusion arising from Jerome’s comments may be explained as a consequence of a multi-faceted plan to realign the church’s Old Testament with the Hebrew Bible, a plan that Jerome articulates only partially on any given occasion.

 I'll note the article again when it is finally published. 

The Messiah in the Old Testament

Several posts on this blog recently have assumed that some passages in the OT look forward to a Messiah, that is a future Davidic king. This idea of a Messiah in the OT has a long history in Christianity, and one may say it has a long history of abuse. Christians often like to see Jesus everywhere in the OT. I don't mind seeing Jesus everywhere, if in doing so I'm following the example of Paul (as articulated by someone like Richard Hays), but I do want to understand what meaning these same passages might have had for the ancient Israelites, who would not have wanted to look for Jesus in every verse.

I have just finished reading an essay by J.J.M. Roberts on this topic, and I thought it would be helpful to summarize here some of the things he says. The essay is titled, "The Old Testament's Contribution to Messianic Expectations," and it is reprinted in Roberts' collection, The Bible and the Ancient Near East (Eisenbrauns, 2002), 376-89.

First, the word messiah (מָשִׁיחַ) cannot serve as a guide to the messianic passages in the OT, because its 39 appearances in the HB never look forward to a future Davidic king. It mostly refers to a contemporary Israelite or Judaean king, sometimes to priests or prophets, and once to the Persian king Cyrus (Isa. 45:1). Note that Roberts agrees with the majority of scholars (but not the majority of Christian tradition) in interpreting the two appearances of 'messiah' in Dan. 9:25-26 in terms of Jewish figures before the first century.

Second, some passages that have usually been interpreted throughout Christian history as involving a messianic hope actually probably did not originally carry this meaning but were rather written as a reflection on the contemporary monarch. Into this class, Roberts places Psalm 2, Psalm 110, and Isa. 8:23b-9:6 (Hezekiah).

Third, Roberts does identify some texts that "do in fact envision a future ruler not yet on the scene" (p. 381). These texts begin to appear in 8th-cent. documents. Roberts imagines that "the political disasters of the late eighth century, including the destruction of the northern kingdom and the deportation of a significant portion of the population of the southern kingdom, produced widespread longing for the unity, strength, and justice of the idealized united monarchy of the past" (p. 381). And so we have the legitimately messianic texts (without the word 'messiah') in Isa. 11; 32:1-8; Hos. 3:5; Amos 9:11-12; Mic. 5:1-5; Zech 9:1-10. This hope is also found later in Jeremiah (23:5-8; 30:9; 33:14-26) and Ezekiel (17:22-24; 34:23-24; 37:15-28).

While Zech. 3-4 has contemporary figures in view (Zerubabbel and Joshua), the linking of royal and priestly figures here (reflecting the earlier Jer. 33:14-26) probably became an inspiration for later conceptions of a priestly Messiah either alongside the Davidic Messiah (Qumran) or identical to him (Christianity).

One more thing worth mentioning from this essay. There is nothing in the OT that indicates that the coming Davidic king would himself rule forever. Rather, what seems to be in view in these passages of hope for a Messiah is the re-establishment of the Davidic dynasty, which would function like any dynasty, with sons replacing fathers on the throne. Hardly any other meaning can be gotten out of Jer. 33:22, where God promises to "multiply the seed of David," that is, give him a long line of descent, which would be valuable only if death still comes even to the Davidic king. This assumes, of course, that the Israelites did not consider the coming king to be the Second Person of the Trinity. That seems like a reasonable assumption!

The Curse of Death

A new(-ish) book by Rodrigo Morales, The Spirit and the Restoration of Israel: New Exodus and New Creation Motifs in Galatians (WUNT 2/282; Mohr Siebeck, 2010), has recently been reviewed in RBL by Sigurd Grindheim. It looks like an interesting book, and I especially took notice when Grindheim summarized Morales' argument such that it parallels an interpretation of a passage unmentioned by Grindheim in his review (and Morales in his book?) that I offered on this blog recently.

I was talking about Acts 13:39--which says that Jesus has rescued you from something the Torah could not rescue you from--and I suggested that even more than sin, the verse referred to death as that from which Jews needed a rescue. Now I see that Morales says something similar, though I'm not sure that he brings Acts 13:39 to bear on the issue. I quote from the second page of the review by Grindheim:

The most detailed discussion in the book is devoted to the notorious crux in Gal 3:10–14. Morales argues that Paul did not speak in hypotheticals when he quoted the curse from Deut 27:26. Rather, this curse was already a fact that affected all Israel due to their disobedient heart. Relying on an article by Joel Willitts, Morales shows that Ezekiel and Nehemiah also understood this curse to be in effect. In contrast to James Scott and N.T. Wright, Morales observes that Paul understood the curse not as exile but as death. Israel’s predicament, as it is reflected in Galatians, is not that they continue to be in a state of exile but that they cannot be free from death.

UPDATE: Morales' book is searchable on Google Books and I can confirm that Morales does not cite Acts 13:39 in his book.