Dorival on the History of Canon Research

I've been writing about Gilles Dorival's new book on the LXX (here), particularly about his first chapter on the development of the Jewish canon (here). Dorival spends a few pages (pp. 7–11) of his first chapter surveying "Theories of the Formation of the Biblical Canon." In this post I note merely some corrections and confusions (on my part, or Dorival's). 

Several times in this book (pp. 7, 35–36, 171), Dorival attributes the first formulation of the Alexandrian canon theory to Grabe in the preface to his 1715 translation the Letter of Aristeas. But, as I've noted before, Grabe did not write that preface, and the preface has nothing whatever to do with the Alexandrian canon theory. Instead, the first formulation of the theory should be attributed to Francis Lee in 1719 (see here). 

In this section Dorival also describes the three-stage theory of canon formation (p. 8). In favor of the view that the Torah was canonized first, Dorival comments on Ezra's reading of the Law in Nehemiah 8 during the course of a morning. "Since half a day is time enough for the reading of the Law, but not the Law and the Prophets, the implication is that the Prophets were not yet part of the canon." This argument makes no sense. Now, I acknowledge that Dorival goes on to argue against this three-stage theory, so he would presumably respond to my previous sentence by saying, "yeah, I know." But I'm wondering whether anyone can ever have really brought forward this line of argument about Nehemiah 8. I've never heard it in those terms, before, and for good reason: Nehemiah 8 says nothing about Ezra reading the Prophets, so why would anyone think it would be a good argument to say that he didn't have time to read the Prophets? And another thing: what about Dorival's assertion that half a day is plenty of time to read the Pentateuch? I guess, but you'd have to read quickly. On the next page Dorival acknowledges that the half-day reading "suggested that the Biblical corpus was limited to the Torah or even to one book of it" (p. 9). 

The overturning of the Alexandrian canon theory occupies the next couple paragraphs, in the midst of which he makes an interesting statement about the discovery of Ben Sira in the Cairo Genizah: "For the first time, it was proved that a book hitherto considered as specific to the Alexandrian Bible had a prior existence in Hebrew" (p. 10). Did scholars in the nineteenth century really think that Sirach might have been written in Greek? Even though the translator's preface precedes the book in Greek? And even though Jerome had said (in his preface to the books of Solomon according to the Hebrew) that he had seen a copy of Sirach in Hebrew? And even though the book is quoted in rabbinic literature, a fact surely known to those nineteenth century scholars? A statement similar to Dorival's is made in an article by Natalio Fernández Marcos (in his essay in this book), when he claims that the Alexandrian Canon hypothesis was based in part on "the idea that most of the deuterocanonical/apocryphal books had been composed in Greek and on Egyptian soil" (p. 76), an idea refuted, he says, by the discovery of Hebrew Ben Sira in the Cairo Genizah and the discovery of Hebrew Ben Sira and Hebrew/Aramaic Tobit at Qumran. 

Look, maybe Fernández Marcos and Dorival are right about this—that scholars used to think these writings were composed in Greek—but I have my doubts. First, Jerome (problematically) claimed he had translated Tobit from a semitic text, so I don't see why western scholars would ever imagine that Tobit had been composed in Greek. I've already mentioned the evidence for Ben Sira. I don't feel like digging through nineteenth-century writings on the canon right now to see whether they thought the deuterocanonicals were written in Greek, but these two books (Tobit and Ben Sira) are the worst examples, because there was definite evidence for a semitic origin long before the discoveries in modern times. Furthermore, the heyday of the Alexandrian Canon hypothesis was the first half of the twentieth century (see here), and started to be widely accepted right around the time of the discovery of the Cairo Genizah. How could the Alexandrian Canon hypothesis be based on the idea that the deuterocanonicals originated in Greek if the strongest supporters of the hypothesis lived after the discoveries in the Cairo Genizah? 

Dorival's conclusion to this section: 

Because of these discoveries [= Cairo Genizah and Dead Sea Scrolls], one may conclude that, in the Judaism prior to Jabneh, a collection of holy books larger than the Jewish canon of twenty-two/twenty-four books existed. This collection appears to vary from group to group, with a stable of books common to them all. There is no direct connection between Alexandria and the deuterocanonical books. The Christian Old Testament is larger than the Rabbinic Bible because it comes from the larger collection of books that was understood as 'inspired' by one or several Jewish groups at the beginning of the Christian Era. (p. 11)

This view has similarities to the "majority canon" position of Timothy Lim, though Dorival does not cite Lim here. I myself think that this conclusion should be stated less confidently. (See my review of Lim.) It is, in fact, not clear that "a collection of holy books larger than the Jewish canon of twenty-two/twenty-four books existed," although I suppose that Dorival is correct that "one may [or may not] conclude" so. 

Dorival on Canon Terminology

As I mentioned last time, I've got some thing's to say about Dorival's first chapter, the one on the development of the Jewish canon. As it turns out, I'm going to split my thoughts on this chapter into at least two posts. 

Dorival's first major section in the chapter is called "Words and Concepts" (pp. 3–7). He runs through the well-known history of the word κανών and related terminology. Let me mention first something that I appreciated. Dorival acknowledges that a canon can exist even in the absence of the word "canon," and he thinks such was the case for some ancient Jews. 

First, even if the word 'canon' is lacking, the reality of the canon did exist in these ancient Jewish milieus: that is, a list of books understood as being in some sense normative. Greek-speaking Jews probably used the expression 'testamentary books' (ἐνδιάθηκοι) for this list. (p. 5)

For the term ἐνδιάθηκος, see Origen's Selecta in Psalmos 1 as quoted by Eusebius (Hist. eccl. 6.25), where the list of of "testamentary books" is attributed to "the Hebrews," from whom Origen may also have derived the terminology of "testamentary books."  

Moreover, Dorival argues that the Rabbis used the word seder for what Christians called a κανών, and he cites Jerome's Prologus Galeatus (where the term is ordo) in favor (pp. 5–6). 

In Jerome's text, the word ordo first refers to the succession of the books among the three categories of biblical books, but then also to each category of books. The word ordo has the meaning of category of books in Gelasius' Decretum ... and in Rufinus, Commentarius in symbolum Apostolorum 36. (pp. 5–6)

He connects ordo in these Latin sources to seder at b. B. Bathra 14b. "The suggestion is that the Sages called seder what the Church fathers referred to as canon" (p. 6). I think this is a good possibility. I feel like I may have made this suggestion in print, but I can't remember where. I'm glad to see it here in Dorival. (Or, maybe I just read it in the previous French version of this essay.)  

I do have a couple of critiques on small points about termionology. Dorival discusses the Hellenistic-era Alexandrian canons of classical authors, seemingly implying that the word κανών was used for these lists (p. 4). It was not. Later he says:

English historians assert that the first modern occurrence of the word 'canon' meaning 'the canon of the Scriptures' is David Rhunken [sic] in 1768. In fact, this word with this meaning is found in French writings of the late seventeenth century: in 1685, in the work of Richard Simon ('Canon juif') ... (p. 6, providing further examples)

These statements from Dorival are confused. What can he mean by telling us that the word "canon" in the sense of "canon of Scripture" is already so used in seventeenth-century French literature? He seems to mean that Simon's use of the word "canon" in this sense is an early example of this meaning, but hasn't Dorival already told us that this word is used in this sense in fourth-century Greek literature? In that case, Richard Simon was not innovating, even if the word "canon" was rarely used in this sense in the intervening years (about which I am not certain). Actually, now that I look back at Dorival's first few pages, I'm not sure whether he acknowledges that Athanasius used κανών and related terms to designate the canon of Scripture. He simply notes on p. 3 that Eusebius and Athanasius did use these terms, but he doesn't say what sense the words bore in those contexts (and Eusebius himself did not use κανών for "canon of Scripture"). As for the assertion about English historians and David Ruhnken, I believe Dorival has again misunderstood. In a previous post, I noted that Ruhnken is credited as the first person to use the word "canon"—not in the sense of "canon of Scripture," a usage that goes back to Athanasius, but—for the lists of classical authors drawn up by the Hellenistic-era Alexandrian scholars. Those Alexandrian scholars did not use the term "canon" to talk about "the canon of Greek orators" or whatever; that was the innovation of Ruhnken.   

Another thing: Dorival says, "The existence of the acronym Tanak (b. Sanhedrin 101a, b. Qiddushin 49a, b. Mo'ed Qatan 21a) seems to imply they did not have a word for canon" (p. 5). The French version (p. 12) makes it even more clear that Dorival means that the Talmud uses the acronym Tanak. But the Talmud does not use the acronym. In the three passages listed by Dorival, the Talmud uses the spelled-out names Torah, Neviim, Ketuvim (in Hebrew, or—in the case of Quddushin—Aramaic), not the acronym. I checked in the Soncino edition of the Vilna Shas, but you can also check it at Sefaria: it's §3 of Sanhedrin 101a; §12 of Qiddushin 49a; and §7 of Mo'ed Qatan 21a. Maybe I'm misunderstanding something?  

Dorival's New LXX Book

Gilles Dorival is well-known among Septuagint specialists, as he has contributed many French-language studies to the field, and has especially dedicated his energies to the study of the catenae, and he has also dabbled in the development of the Jewish and Christian biblical canons. Now he has published his Grinfield Lectures in a monograph, The Septuagint from Alexandria to Constantinople: Canon, New Testament, Church Fathers, Catenae (Oxford, 2021). 

The four subjects listed in the subtitle correspond to the four sections of the book, with each section constituting two chapters. After the eight main chapters, there is also a lengthy concluding summary (pp. 171–87).

Probably the most welcome part of this monograph is the final section on the catenae, a subject hitherto not well-served in English. As I mentioned, Dorival has contributed a great deal to this area of study, particularly on the catenae of the Psalms (see, e.g., here), and it is the Psalms from which he takes many of his examples in the current monograph. 

Nevertheless, it is the other sections of the book that dovetail most closely with my own interests, and it is nice to have Dorival's thoughts in English on these topics. In a subsequent post I will offer some appreciation and critique regarding certain things that Dorival says in the first chapter on the development of the Jewish canon, but for now I will simply summarize briefly his first six chapters. 

Chapter 1, "The Formation of the Jewish Canon" (pp. 3–33). As Dorival tells us in the first note, this is a translation of his essay in this book. I'll have more to say about this essay in a future post. 

Chapter 2, "The Septuagint and the Issue of the Canon" (pp. 34–47). Dorival rejects the Alexandrian Canon hypothesis. He argues that some Jews before the turn of the era had a category of books that were not read publicly but rather privately, and that this category influenced the reception of the deuterocanonical books in Christianity. 

Chapter 3, "Is the Septuagint the Old Testament of the New Testament?" (pp. 51–68). The answer is "yes, with caveats," the biggest caveats being that not every quotation aligns with OG. 

Chapter 4, "Was the Text of the Septuagint Christianized?" (pp. 69–91). This was a good chapter. The assumption among some scholars for some time has been that LXX manuscripts were occasionally Christianized, in the sense that OT verses quoted in a variant form in the NT were adapted in the LXX to align with the NT form of the verse. This view has been challenged, and Dorival joins that challenge here. He shows how little of the LXX can be said to have been Christianized. But he doesn't limit himself to OT verses quoted in the NT. He deals first with the insertion of actual Christian content into the LXX, which does occasionally happen, such as the interpolation of Rom 3:12–13 into the text of Psalm 13 in the majority of manuscripts, and the Christian texts (such as from Luke) in the Odes. As for the the OT quotations in the NT, Dorival examined the ones in the Psalms (25 total) and found that hardly any Christianization had occurred in many manuscripts of the LXX, with only the debated quotation of LXX Psa 39:7b being found in the majority of LXX manuscripts in its "Christian" form (if this form is not, indeed, the OG). 

Chapter 5, "Is the Septuagint the Old Testament of the Church Fathers?" (pp. 95–116). Here Dorival answers the question "yes and no." While the LXX served the majority of Greek and Latin Fathers as their OT, this was not usually true for Syriac-speaking Christians, and even Greek- and Latin-speaking Christians recognized that the LXX was a translation of the Hebrew and sometimes needed to be understood—or even corrected—according to the Hebrew or a closer translation of it. In his concluding chapter's summary of Chapter 5, he wraps up by saying: "In sum, whilst the domination of the Septuagint in patristic times is a fact, it requires qualification" (p. 178). I would like to think that my dissertation had some influence on Dorival's thinking here; he does cite it a time or two. 

Chapter 6, "The Vocabulary of the Septuagint and the Church Fathers" (pp. 117–31). Dorival summarizes Swete's presentation and updates it. 

And as I mentioned, the last two chapters introduce the catenae. 

Chapter 7,  "An Overview of the Catenae" (pp. 135–54)

Chapter 8, "The Catenae and the Septuagint" (pp. 155–70)

Moberly on von Rad, and Augustine

In preparation for an upcoming course, I've been reading R. W. L. Moberly, The Theology of the Book of Genesis (Cambridge, 2009), which is, of course, wonderful. Of the two chapters that deal with the call of Abraham (Gen 12:1–3), the first concerns the correct construal of the blessing formula in Genesis 12:3. God promises Abraham, "in you all the families of the earth shall..." what? Be blessed? Bless themselves? Is Abraham being charged with becoming a source of blessing for others, or a model of others' blessing formulae? Is this some sort of a missionary text, as if Abraham is called by God for a purpose, to bring blessing to the world? Or is this a promise of divine protection and blessing, that Abraham will become so prosperous that others will use his name as they bless people: "May you become as prosperous as Abraham!"? 

It is quite common for Christians to read the text in the first way, as a quasi-missionary text, and Moberly cites some heavy-hitters favoring this reading: Westermann, Childs, Christopher Wright, Bauckham. Moberly himself argues for the other reading, that Abraham's name will be used in the blessing formulae of others, and he cites Gunkel as a proponent of a rather negative version of this second reading strategy.  

The scholar with whom Moberly interacts most in this chapter is Gerhard von Rad. Moberly quotes a long passage from von Rad's Genesis commentary, a quotation that takes up more than a page of Moberly's text (pp. 142–44), and then he quotes von Rad again for about half a page. Von Rad was a proponent of the missional reading, and von Rad connected the call of Abraham very strongly to the New Testament. Moberly argues against von Rad's position. But rejection is not Moberly's last word on von Rad's interpretation. 

Von Rad's original formulation of the significance of the Yahwist and Genesis 12:1–3 was in the context of 1930s Nazi Germany, and his specific situation was as a member of the Confessing Church working at the University of Jena, where National Socialist policies were strongly promoted. [Moberly cites this essay.] In such a context, where the authorities degraded the Old Testament and denied any positive enduring significance to it, von Rad's work was a profound and imaginatively serious contribution; his argument for strong continuity between the Old and New Testaments is an argument that is intrinsic to Christian faith and was particularly timely as a Christian Old Testament scholar's response to Nazi ideology. By contrast, Gunkel's reading of God's call and promises as an example of Israel's rather inflated sense of self-importance would in no way have made any (would-be) Nazi or anti-Semite think twice. Good theological interpretation of the Old Testament is not necessarily that which might aspire to be recognized as correct in any time or any place; rather, part of its rightness may be specific and contextual, in its ability to articulate biblical priorities in relation to particular situations of need. To say this is not to prioritize relevance over accuracy, but rather to recognize, with the sociology of knowledge, that human understanding and insight depend on many factors other than pure reason and do not achieve finality in any one situation. (pp. 158–59)

There is a  lot in this paragraph that calls for reflection, and I'm thinking particularly of the way Moberly here formulates the task of theological interpretation. For now I want merely to put Moberly's interpretation of von Rad in conversation with Augustine. 

So anyone who thinks that he has understood the divine scriptures or any part of them, but cannot by his understanding build up this double love of God and neighbour, has not yet succeeded in understanding them. Anyone who derives from them an idea which is useful for supporting this love but fails to say what the writer demonstrably meant in the passage has not made a fatal error, and is certainly not a liar. (De Doctrina Christiana 1.86, trans. Green)

Putting Moberly's reading of von Rad's interpretation next to Augustine's hermeneutical advice suggests that von Rad's incorrect interpretation of Gen 12:3 was more correct than a correct interpretation might have been. Of course, there were probably ways of articulating the "correct" interpretation more "lovingly" than did Gunkel (e.g., Moberly's own articulation of it), but one wonders how imaginable such an articulation would have been in von Rad's context.