There are weaknesses in our books, and there are things we wish we had emphasized more or stated differently but, honestly, we are becoming tired of knee-jerk reactions to our work that reflect a superficial understanding of its content and purpose. For example, we have seen the claim repeated that the aim of our arguments is to prove a late date of all biblical literature. This is a truly bizarre claim. We are pretty clear that we are saying that all linguistic dating arguments, both for early or late dates, don’t work.
See Unhistorical Hebrew Linguistics: A Cautionary Tale
By Ronald Hendel
By Robert Rezetko
Faculty of Philosophy, Theology and Religious Studies,
Radboud Universiteit Nijmegen
Honorary Research Associate,
Department of Hebrew, Biblical and Jewish Studies,
The University of Sydney
Department of Hebrew, Biblical and Jewish Studies,
The University of Sydney
Department of Biblical Exegesis,
The University of Copenhagen
We appreciate the opportunity to respond to Ron Hendel’s op-ed1 and to clarify our real views and arguments.
Unfortunately Hendel takes out of context the short “potted summary” of our books2 and fails to engage our real views and arguments. We hope he will read the books on which the popular summary is based.3
Hendel’s op-ed and additional comments so far mention six main issues:
- Mamlakah and malkut
- Persian loanwords
- Septuagint, together with Joosten’s review of our books (in press; on the web)
- Textual history
- Historical linguistics
- Extra-biblical linguistic data
Obviously we are not going to repeat in full what we have already written in more detail in the books. And neither do we want to write full responses to misunderstandings and misstatements that we will address in much more detail in other publications in preparation and press. But we are willing to set forth again some of our main ideas and a few points of detail, so here are some boiled down responses.
Textual History. We agree with Hendel wholeheartedly when he says elsewhere: “In the case of the Hebrew Bible it is difficult to define what the ‘original’ means, since each book is the product of a complicated and often unrecoverable history of composition and redaction. The ‘original text’ that lies somewhere behind the archetype is usually not the product of a single author, but a collective production, sometimes constructed over centuries, perhaps comparable to the construction of a medieval cathedral or the composite walls of an old city” (Hendel, “The Oxford Hebrew Bible: Prologue to a New Critical Edition” , p. 332).
Of course we are aware of the complexity of the formation of the books of the Tanak, and that the historical linguistics of biblical Hebrew should not be separated from other disciplines such as literary criticism and textual criticism. Yes, absolutely, “textual criticism is entirely compatible with historical linguistics, and, indeed, ...the two pursuits are necessary adjuncts.” See chapter 13 in volume 1 of Linguistic Dating of Biblical Texts (LDBT), and other publications by us cited in the bibliography in volume 2 related to literary criticism and textual criticism. It seems almost unbelievable to us that Hendel actually claims “The authors’ argument evinces a tenuous grasp of the practice and implications of textual criticism.”
Historical Linguistics. We are hardly completely ignorant or dismissive of historical linguistic methodology. Of course ancient Hebrew changed! Of course historical linguistic methodology should be applied to the Tanak! But there are some big problems, routinely addressed in historical linguistic literature, that many Hebraists have often neglected, such as the issue of authentic and localized (in time and place) manuscripts and the issue of composite texts. We think Hendel will agree that we have nothing close to authentic (or “original”) manuscripts and that most and perhaps all biblical writings are composite. This does not mean that historical linguistic methodology cannot be applied to the Tanak, but it does mean that that should be done with more sophistication and integration of various disciplines, including also a realistic understanding of the complex nature of the Tanak’s composition/editorial/transmission history. Therefore, the argument we are formulating in a book in progress, Diachrony in Biblical Hebrew: Steps Toward an Integrated Approach, is that historical linguistic research on biblical Hebrew (setting aside the anomalous and unsound matter of linguistic dating, what our previous books were all about and sought to discredit) must more carefully and fully integrate historical, literary, textual and linguistic disciplines, the latter including also corpus linguistics, a variationist approach, and typological linguistics. The book is historical linguistic in scope and makes extensive use of historical linguistic research on other languages as comparisons to what is done and what should be done in the study of biblical Hebrew. (As Robert Holmstedt said to us, “I’m guessing that the likes of Brian Joseph and Lyle Campbell would simply roll their eyes at what passes for methodology in historical Hebrew studies.” Undoubtedly!)
Septuagint. Joosten’s arguments in these articles are less clear and certain than Hendel suggests, though they are important articles and offer a novel way perhaps to gain a glimpse of a few particular issues in late Second Temple Hebrew. We touch on several of his related points in LDBT (e.g. LXX translation of paragogic heh, vol. 1, p. 79, n. 71; LXX Jeremiah, vol. 2, p. 158). Rezetko has also recently sent for publication a very detailed review-article of a monograph that has a similar line of argumentation.
As for the book review by Jan Joosten that is cited by Hendel,4 we found it very disappointing. Actually, there are indications in the review that Joosten did not read both volumes of LDBT completely and carefully. Misinformation and disparagement aside, we have written responses to most of his criticisms, but here we will only discuss briefly his five examples of supposed weakness (“unevenness”) in philological analysis (pp. 6-7) that he cited to demonstrate that “Inaccuracies like these do not inspire confidence in Young and Rezetko’s [and Ehrensvärd’s?] ability to deal seriously with the linguistic data.” We feel we can safely assume he selected and discussed only these five specific examples because he was fairly clear and certain about them. So, let’s take a quick look at the five specific examples he discusses. (To fully understand the issues involved we refer readers to Joosten’s review and the discussions and literature cited in LDBT.)
1. “On pp. 72-74 of vol. 1, the syntax of 2 Chr 30:1, 5...” We agree that Joosten is probably right about 1 Sam. 10.14 and 2 Sam. 24.21 (// 1 Chron. 21.22), though 2 Sam. 24.21 is debatable when viewed alongside synoptic 1 Chron. 21.22. So, we are left with ten of twelve examples in Samuel, which are also cited by Miller (referenced by us; and she cites many other examples in EBH). Two probable mistakes do not invalidate a three page argument.
2. “On p. 104 of vol. 1, the syntax of 2 S 6:16...” Apart from Joosten’s text-critical oversight— most likely the LXX translators were translating wayehi and not wehayah as Joosten assumes (cf. 4QSama, MT/LXX Chronicles)—in 2 Sam. 6.16 this is probably not introductory wehayah (“and it happened...”) but rather a periphrastic construction (“and was...entering”). English translations (and also commentaries) have both options: introductory: JPS, NASB; periphrastic: TNK, NAB, NIV, NJB, NRSV. But the construction sets up a parenthetical scene and thus the introductory interpretation/translation is less compelling. Also, the syntax in 2 Sam. 6.16 is similar to 1 Sam. 25.20; both are better understood as having periphrasis (e.g. Endo, The Verbal System of Classical Hebrew in the Joseph Story , p. 184, n. 110). There are many literary connections between the Abigail story in 1 Samuel 25 and the Michal story in 2 Samuel 6, so it is somewhat interesting that we should find the periphrasis with wehayah + participle and also a following lqr’t in only these two chapters in Samuel. Rezetko addresses the text-critical and literary issues in his ark book. And MT accents? Joosten has got to be kidding! They have no more interpretative authority than the opinions of other commentators throughout the ages.
3. “On p. 114 of vol. 1, the LBH usage of qbl, Piel, ‘to receive,’...” We do not assert what Joosten deduces from his shallow reading. We are not trying to say that the items in that column prove that EBH had the same words/roots in similar uses etc. Quite the contrary, the point is that “items in the table in the EBH column are used alongside the items in the LBH column” in LBH. We simply cite the root in the “Occurrences Elsewhere” column for the sake of thoroughness, not to say that EBH had the same word/root/use. Our true intention is clearly stated on p. 113: “We can illustrate this by gathering together all LBH lexical features in our appendix regarding which the lexeme itself or a particular ‘late’ meaning, referent, stem or syntagm occurs 10 times or more in (mostly) undisputed postexilic books and not in core EBH books.” Joosten’s eye simply must have caught the reference while skimming through this material because he completely fails to grasp the context of the argument.
4. “On pp. 303-304 of vol. 1, dat, ‘law’...” See the remarks below on Persian loanwords.
5. “On p. 129 of vol. 2, Eccl 4:17...” Callaham, in his dissertation/monograph on the modality of the verbal infinitive absolute, takes the infinitive as an imperative, as we say, though Ehrensvärd himself was uncertain in earlier publications (2003, 2006). The possibility of an imperative is discussed elsewhere in the literature, e.g. Crenshaw, Murphy, Seow, and considered a possibility, but then set aside in favor of the substantive option. However, even Murphy says: “But it is also possible to interpret it as an imperative (‘enter’ ...) parallel to ‘watch,’ as most of the ancient versions and many commentators (e.g., Lohfink) have done” (Murphy, Ecclesiastes , p. 45, n. 17). Longman also reads it as an imperative. English translations have both options: substantive: TNK, NJB, NRSV; imperative: JPS, NAB, NASB, NIV. So it is quite amazing that Joosten should pick this example to demonstrate our lack of “ability” in philological analysis. We are hardly alone in our analysis!
So, in summary, on the basis of his presentation of only these five specific examples Joosten states: “Inaccuracies like these do not inspire confidence in Young and Rezetko’s [and Ehrensvärd’s?] ability to deal seriously with the linguistic data.” Really?!
Mamlakah and Malkut. Of course we know the textual data Hendel cites, but the point we are making in the “potted summary” is that if one works on the basis of the MT—as many recent Hebrew linguists have done, and in fact have asserted should be done, e.g. Hurvitz, A Linguistic Study of the Relationship Between the Priestly Source and the Book of Ezekiel (1982), p. 21—then it is clear that malkut was an available option for writers in the pre-exilic period. As Hendel himself says: “The [sic - they, meaning Young, Rezetko and Ehrensvärd] rightly note that if the word occurs in early texts, then the word cannot be an indicator of late language. Rather, it must be an option at any time. Therefore LBH features aren’t necessarily late.” Hendel seems to have overlooked Num. 24.7; it is unmentioned in his op-ed. How does Hendel account for malkut there? We are aware of different kinds of explanations for this “late” word in the “early” Balaam oracles (e.g. poetic usage, foreign speech), but, however it is to be explained, it would seem at least to “occur in an early text” (to paraphrase Hendel’s words) and thus to have at least been an option for some early writers. Actually, though, a thorough and accurate analysis of the distribution of occurrences of malkut and mamlakah and several other related nouns (and their uses) using a corpus-linguistic and variationist approach to all ancient writings is quite interesting and instructive, and this is something that has never been undertaken until recently, as far as we know, so stay tuned. (Interestingly, mamlakah-malkut variation has traditionally been considered one of the clearest and strongest pieces of evidence for diachrony in biblical Hebrew, thus it is often cited as a “classic” illustration [e.g. Hurvitz, Rooker], but even this pillar of diachrony is far less persuasive when the whole truth is told.) (Finally, by the way, and this takes us back to the issue of textual criticism, textual evidence shows that there were multiple directions of linguistic modification, e.g. LBH forms replacing EBH ones in some texts/traditions/editions [as Hendel points out] but also EBH forms replacing LBH ones in other texts/traditions/editions. For some examples see LDBT, volume 1, chapter 13. This is an issue we illustrate and discuss further in our book in progress.)
Persian Loanwords. Once again note what we were arguing and in what context in regard to Persian loanwords. We were first of all countering the claims made by MT-only scholarship that it is a very significant result that no Persian words are found in “early” sources. We pointed out that this is a circular argument, since the various Persian words in the MT are explained away on the assumption that early texts can’t have Persian words! We were very clear about what we were doing: “The main purpose of compiling this list is rather to counter the suggestion that such a list cannot be made” (LDBT, vol. 1, p. 303), and there is “no doubt that Persian elements do occur in the current text of EBH books” (LDBT, vol. 1, p. 309; emphasis added). This leads to a discussion of how dubious the use of individual linguistic elements like loanwords to date the original composition of a text is, in the light of the fluidity of the biblical text during its scribal transmission. In this context, among other points, we mention how a good case can be made that Isa. 24.16 is a later addition to the text. Please understand what we are arguing, and read the detailed discussion of Persian loanwords (as well as Egyptian, Akkadian, Greek and other loanwords) that lies behind the specific argument in our “potted summary,” in chapter 11 of volume 1 of LDBT (and cf. Young’s article on Pesher Habakkuk in JHS 2008). There we include a variety of other considerations about the difficulties of using Persian loanwords in linguistic dating, such as the fact that absence of Persian loanwords is characteristic of various definitely post-exilic works, and hence the argument from absence for an early date is not compelling, or the evidence that Hebrew came into contact with Iranian languages before the exile. But what we have said here is enough to say: to understand our broader position on Persian loanwords, please read our broader work.
Extra-Biblical Linguistic Data. In a subsequent response to Philip Davies, Hendel states: “there are sufficient extra-biblical linguistic data to corroborate at several points the standard model of the history of Hebrew.” We note that we have dealt with this argument extensively in LDBT, volume 1, chapter 6 (on Hebrew inscriptions; cf. Young’s 2003 article cited there), chapter 8 (Aramaic), chapter 9 (Mishnaic Hebrew), and chapter 10 (Qumran Hebrew and Ben Sira; cf. Young’s article on Pesher Habakkuk in JHS 2008). In fact we argue that the extra-biblical evidence argues against the standard model.
There are potentially many other interesting issues to discuss. We think a particularly instructive topic would be scholarship on P, and the different kinds of (chronological) conclusions that have been reached by biblical commentators/literary critics, historians of religion, and Hebraists. We summarize the main points in LDBT, volume 2, chapter 1. The place of P in the debate on the chronological relationship of different varieties of Hebrew needs a lot more work. (We know of some ongoing linguistic-oriented work, and we will add our contribution soon.)
1. Did ancient Hebrew have a history, did it develop? Yes, of course, all natural languages have histories! We have never said or intimated otherwise.
2. Should historical linguistic methodology be applied to ancient Hebrew? Yes, why not, and we have never said or intimated otherwise.
3. Has historical linguistic methodology been effectively applied to biblical Hebrew? No, not yet in our opinion, not in the light of what we know now about the Tanak’s composition/editorial/transmission history, not without a substantial amount of circular reasoning, and not in recent times largely because of increased specialization in Hebrew and Hebrew Bible studies which has often kept historians, historians of religion, literary critics, textual critics, and linguists from working together and addressing particular issues in unison. That should change, but it will require humility.
4. Can biblical Hebrew texts be dated on a linguistic basis? No, not in our opinion, as we argued in LDBT. We will be very clear about this: The point of Linguistic Dating of Biblical Texts was not to write a history of biblical Hebrew, nor to develop a historical linguistic approach to biblical Hebrew, but to confront head-on recent attempts to assign dates of origin to biblical writings on the basis of linguistic data. This is pretty clear in the full title/subtitle to volume 1: Linguistic Dating of Biblical Texts: An Introduction to Approaches and Problems. In other words, we were reviewing scholarship up to that point and indicating and expanding on problems that had emerged in recent discussion. Any interpretation or review of our work that does not keep this objective in mind is bound to misunderstand our argument and misrepresent our views. Statements to the effect that we believe or assert that Gesenius, Driver, Kropat, Kutscher, Polzin, Hurvitz, Joosten, Eskhult, etc. all were mistaken, are senseless and do not represent our real views or published words.
There are weaknesses in our books, and there are things we wish we had emphasized more or stated differently but, honestly, we are becoming tired of knee-jerk reactions to our work that reflect a superficial understanding of its content and purpose. For example, we have seen the claim repeated that the aim of our arguments is to prove a late date of all biblical literature. This is a truly bizarre claim. We are pretty clear that we are saying that all linguistic dating arguments, both for early or late dates, don’t work. Obvious practical applications of our work include Young’s arguments against the late dating of Job (VT 2009) and Qoheleth on linguistic grounds (indeed Young argues for an earlier date for Qoheleth than almost any other scholar on the planet: interesting coming from scholars dedicated to a late-dating agenda!). And Rezetko has been clear in his publications that the book of Samuel had a long and complex history of composition/editing/transmission. So, read our work! And read our forthcoming work!
Thanks again, Ron, for the op-ed that gave us yet another chance to clarify our real views and arguments!
thank you for a reasoned, well argued, factually based and informative rejoinder.
#1 - Jim - 09/28/2011 - 20:40
YR&E begin and end their response to my column by writing: “we are becoming tired of knee-jerk reactions to our work that reflect a superficial understanding of its content and purpose.” They’re entitled to view my column as “superficial” and a “knee-jerk reaction.” What they fail to do is to contest any of my specific criticisms of their arguments, analyses, and assumptions. I take it that these criticisms stand.
I am happy to respond to their query about the word malkut in Numbers 24:7 and its implications. I didn’t address this word in my column because this is not a CBH text, and hence it’s not germane to an analysis of CBH vs. LBH. As YR&E know, the corpus of old poetry (here, Balaam’s Oracles) has a different linguistic profile than CBH. The word occurs in Old Aramaic (e.g. the Sefire inscription), and therefore it’s not surprising to find it in this text. (And it may suit Balaam as an Aramean seer.) What I did show is that YR&E’s argument that malkut occurs in CBH texts is thoroughly flawed. They now qualify their argument by stating that “if one works on the basis of the MT—as many recent Hebrew linguists have done…—then it is clear that malkut was an available option for writers in the pre-exilic period.” But this is the crux of the issue. One cannot just work “on the basis of the MT” if one is doing historical linguistics of biblical Hebrew. Yet YR&E persist, writing that “this pillar of diachrony is far less persuasive when the whole truth is told.” Restating a flawed argument doesn’t make it any better.
Regarding extra-biblical linguistic data, YR&E maintain that “the extra-biblical evidence argues against the standard model.” Well, this only holds if one accepts the dubious assumptions that they’re working with. For example, in the Qumran chapter, they rightly observe that there are “a number of LBH forms in Pesher Habakkuk” (vol 1, p. 262). But then they say that it doesn’t have more such features than some CBH books. After comparing the numbers, they conclude: “One of the Qumran text samples, Pesher Habakkuk, falls squarely in the middle of the EBH books” (274). This conclusion is marred by two methodological problems: (1) their “MT-only approach” includes problematic or illusory readings, as I have shown, and (2) they exclude from consideration LBH syntax and the verbal system. Had they considered the verbal system and syntax, the raw numbers of LBH features in Pesher Habakkuk would cascade. In sum, YR&E are misrepresenting the linguistic data on multiple levels, which skews their statistics to near meaninglessness.
#2 - Ron Hendel - 09/28/2011 - 23:05
A word on Avi Hurvitz and MT. YR&E state that they adopt the “MT-only approach” on Hurvitz’s authority. Now, I assume that Hurvitz adopted this approach out of an abundance of caution, because it shields him from the accusation of cooking his data. But he knows (and says) that this is, in principle, incorrect. Others have followed him in this stance, as YR&E document. This decision mars some of Hurvitz’s results, such as his discussion of ‘ad in Job 1:18, which is clearly a scribal error (or simply a defective spelling) of ‘od, which occurs in the same formula three times in vv. 16-18. But this methodological problem usually doesn’t affect Hurvitz’s larger arguments, which deal with well-attested linguistic oppositions of CBH vs. LBH. With Hurvitz this methodological problem is a minor and occasional flaw. YR&E take this principle to it logical extreme, basing large arguments on scribal errors and strange hapax legomena. In doing so, they demonstrate that the “MT-only approach” to Hebrew linguistics is wholly untenable.
With this method, they find LBH features in every book, such as the Persian word dat (“law”) in the old poem of Deuteronomy 33. They now write: “there is ‘no doubt that Persian elements do occur in the current text of EBH books.’” This is the problem. Their method precludes them from conceding that this word in MT is the result of scribal error. The preferred reading for Deut 33:2 is arguably אשרו אלם (“the gods march”), which is reflected in the LXX (this was proposed by Cross and Freedman). This reading has nothing to do with a Persian loanword, which is a scribal phantom in MT (produced by a couple of graphic errors: resh/dalet and aleph/taw, the latter arguably in the old Hebrew script). In other words, their “MT-only approach” ends up producing “false positives,” which YR&E know are textual illusions. In effect, they demonstrate the reductio ad absurdum of the “MT-only approach” to Hebrew linguistics.
YR&E claim to have established that “all linguistic dating arguments, both for early or late dates, don’t work.” I submit that they are only partly right, and that they have drawn the wrong conclusion from their own work. They have established, by taking it to its logical limit, that linguistic dating arguments based on the “MT-only approach” don’t work. Although this is a different conclusion than the one they think they have demonstrated, it is an important result. I hope that it will cause Hebrew linguists to consign the “MT-only approach” to the dustbin of history and embrace the full panoply of critical scholarship. As for YR&E, they have demonstrated that they are learned young scholars who were insufficiently critical about their methodology. If they learn this (hard) lesson, I’m sure that we can expect positive and substantial contributions from them in the future.
#3 - Ron Hendel - 09/28/2011 - 23:06
Robert, Ian, and Martin,
My private email comment from early 2008 required some contextualization (and was too flippant for a public forum). I wish you would have checked with me on that.
Moreover, since you know that I take significant issue with your methods and conclusions in LDBT and your follow-up papers, there is a high irony in quoting me. Indeed the remainder of that old email detailed what I identified as the greatest shortcoming of your work: you used the same assumptions and methodologies of those you criticize (thereby perpetuating the problems) instead of pushing the field towards an approach more in line with historical linguistics in general. Thus, in the end, my Campbell and Joseph comment applies equally to LDBT.
(Note to readers: to contextualize my comments, see the interaction with Young, Rezetko, and Ehrensvärd's work at http://ancienthebrewgrammar.wordpress.com/2011/03/02/biblical-hebrew-diachrony-continued/
I hope that your next work, the diachrony volume you mention, is not simply a defence of your conclusions in LDBT. It will serve the study of Hebrew much more positively if you detach yourselves from the concurrent dialects conclusion and the *texts* focus you (largely, but not entirely) had in LDBT and focus on the historical development of *linguistic features*.
And I encourage you to maintain a better focus than in LDBT. Your description above about integrating "historical, literary, textual and linguistic disciplines", etc., raises all sorts of red flags. There is no way that you (or anyone else at this point) is prepared for such a synthesis. The only way to do this project correctly is to take the long road and consider all the data and possible diachronic paths with no regard to theories on textual development. Only when you have a number of plausible paths, for dozens of features, should you turn to the synthesis and see whether any constellation(s) complement some of the major theories of textual development, or not.
Actually, this is at least a 10-year project. Let it bake longer, much longer.
#4 - Robert Holmstedt - 09/29/2011 - 02:36
Just a thank you to the fearsome trio.
Logics have never beem a problem to biblical studies. The idea of my 2008 book was that most of 200 years of scholarship in this field has dependen on circular argumentation, and for that reason, it is based on false methodology and can easily be discarched. Not a way to get many friends.
But a false argument is a false argument.
It is a problem to any historical lingvistic analysis if it say that P is from the pre-exilic period (Hurvitz), and then goes on to study the language of P as pre-exilic. It is pre-exilic because -- at the end -- the language is, and the language is because it belongs to P.
#5 - Niels Peter Lemche - 09/29/2011 - 11:31
Dear Robert H.,
Thank you for your comment (#4).
In our opinion your Joseph and Campbell remark did/does not apply equally to LDBT since the books are not, properly speaking, a “historical linguistic” work. We have always been very clear that LDBT was intended to introduce and then criticize the widespread (and idiosyncratic) phenomenon in biblical studies of linguistic dating of biblical texts (principles, methods and conclusions) (vol. 1, pp. 3-5). “Linguistic dating” (of texts) barely plays any role in “historical linguistics,” as you and we both know, and which seems to be news to many Hebraists in the field.
You are correct that LDBT mainly was about *texts*, and not the historical development of *linguistic features* as is usual modus operandi in historical linguistics (generally on the basis of authentic, localized and non-composite manuscripts, to be sure). Again, that focus relates to the objectives of our books. And we explicitly stated that we would be focusing on texts and not features (vol. 1, p. 4, 61 n. 32). Of course, in LDBT we do identify, track and discuss hundreds of features, many of which, as we argued, do not fit well in the EBH-LBH paradigm or in any other commonly argued scheme of linguistic or literary development. In any case, we agree with you that the linguistic focus should be on many linguistic features and possible diachronic paths of change, especially in the framework of typological linguistics.
In your last point you comment that “There is no way that you [Rezetko and Young] (or anyone else at this point) is prepared for such a synthesis.” You are absolutely 100% correct! However, the book in preparation, Diachrony in Biblical Hebrew: Steps Toward an Integrated Approach, is not a full-blown synthesis that seeks to re-write the history of biblical Hebrew; rather, the book deals with starting afresh and taking some solid first steps in the framework of a holistic perspective. Therefore, we essentially disagree with your final remark: “The only way to do this project correctly is to take the long road and consider all the data and possible diachronic paths with no regard to theories on textual development. Only when you have a number of plausible paths, for dozens of features, should you turn to the synthesis and see whether any constellation(s) complement some of the major theories of textual development, or not.” Such a project, in our opinion, would move in the wrong direction, by still privileging linguistic analysis over other facets of scholarship (explicitly: “textual development”). That is precisely what most Hebraists, especially those concerned with linguistic diachrony, have been doing. We refer to our discussions of “objectivity” in vol. 1, pp. 16-18, 60-64. In contrast, what we are arguing is that the only effective way for Hebraists to write a solid history of biblical Hebrew is to “embrace the full panoply of critical scholarship,” to use Ron Hendel’s excellent statement, and that means considering historical, literary, textual and linguistic issues simultaneously, “integrated-ly.” We are worried that your language-only or language-mainly or language-first approach might become reductionist in that it could distort the complexity of the matter and would then, as we will argue, amount to inadequate historical linguistic methodology.
Robert, Ian and Martin
#6 - Robert Rezetko - 09/29/2011 - 14:12
comment, part 1:
Thank you very much for some helpful remarks.
A few of the issues are still unresolved, in our opinion, and they will probably remain unresolved until more substantial research is published. We include here mamlakah/malkut. People treat this variation as if it were a clear-cut example of diachronic lexical change. We cannot blame them since this has been used as a “classic” example of lexical change for a very long time. Until now our discussions of these terms have been in the guise of Hurvitz, Rooker, etc. with the purpose of showing on the basis of their MT-only approach that the distribution of malkut is problematic. We feel our treatments of malkut accomplished what we wished in the contexts of those discussions. Nevertheless, we think the article in preparation mentioned above will change some minds about the unassailability of this “classic” illustration, given the complete panorama of data (textual variation included).
Another unsettled issue, in our opinion, is verbal syntax. In actual fact we did not exclude the verb system from consideration in general (see comment #24 on your op-ed) or in relation to Qumran Hebrew and Pesher Habakkuk in particular (see e.g. LDBT, vol. 1, pp. 277-278; to which we can now add some other publications, e.g. Geiger). Again, we simply believe that the situation is more complicated than often portrayed (e.g. ABH>EBH>LBH>Ben Sira/QH>MH).
As regards Avi Hurvitz and the MT, he may or may not *know* that his MT-only approach is, in principle, incorrect, but we would be very interested to know where he *says* this. Can you tell us? We hope you or somebody will answer that question but, in any case, since his MT-only principle sidesteps the entire composition/editorial/transmission process, “regardless of textual alterations, literary developments and editorial activities” (see comment #21 on your op-ed), we are skeptical that it does not affect his larger arguments, and actually we would affirm the opposite (e.g. in regard to his pre-exilic [only] dating of P; cf. LDBT, vol. 2, pp. 11-17).
comment continued in part 2
#7 - Robert Rezetko - 09/29/2011 - 22:12
comment, continued, part 2:
Since we are discussing methodology and you have criticized ours in regard to textual criticism, perhaps you could give us your take on the problem at hand.
You have suggested that we commit a methodological error because we take the books as they stand when criticizing recent attempts at linguistic dating even if those attempts (by others) have an MT-only modus operandi. Using the same MT-only approach (for the sake of argument, so to speak; cf. comment #25 on your op-ed) we observed that all samples of biblical books, whether commonly considered early or late, have conventional LBH features. You seem to suggest that if we had properly used text-critical methodology we would have found that none of these forms exist in “early” books.
We hope you don’t mind us asking you a few questions, to clarify what you are actually saying – which is still not clear to us. How do you know that none of the LBH features in the “early” books are original (if this is actually what you are saying)? Is it because you know that those features are late? How do you know that those features are late, if they appear in “early” books? Is it because you know that “early” books CANNOT have LBH features? How, if all books have LBH features? Is this your argument: “Since I know that ‘early’ books cannot have LBH features, all the actual LBH features in them MUST be late scribal changes?” That is, only books that you know are late can have LBH features in them? We are guessing that this is not your argument, but please clarify, if you don’t mind.
An obstacle to the idea that LBH features cannot be found in early books is the surprising number of supposedly post-exilic LBH features in pre-exilic inscriptions (LDBT, vol. 1, ch. 6, and Young 2003). Those data seem to indicate that there is something wrong with the idea that such forms are “late.” Those data also make it difficult to make a blanket claim that early books cannot have LBH features. We agree that the linguistic profiles of the current books are very uncertain evidence of the language of such a thing as original authors, but it seems that total removal of all inconveniently situated LBH forms is not an option. However, which, if any, of the linguistic features of the current books are “original” to an “author”?
That is the text-critical issue as we see it. As a text-critical scholar whose work we appreciate, we would be glad for you to work on this and other problems with us.
Robert, Ian and Martin
#8 - Robert Rezetko - 09/29/2011 - 22:13
I appreciate your question, but I don't think the methodological principles I am advocating are complicated. I didn’t know ahead of time that your citations of malkut and Persian loanwords in CBH texts were flawed. There’s no way to know this a priori. I examined each of the instances you cited with the tools of critical scholarship and drew what I regard as warranted (and, in this case, non-controversial) conclusions. One must weigh the alternative arguments in each case, have a lot of background knowledge and reference books, and try to make reasonable judgments. And be willing to say “I just don’t know.” That’s philology.
Regarding what you describe as LBH features in pre-exilic inscriptions, there are several logical possibilities. (1) It’s not an LBH feature, in which case one should prune it from the list; (2) It’s a feature in vernacular pre-exilic Hebrew that did not occur in the literary register of CBH, but became dominant in LBH; (3) It’s a minor feature in CBH that became dominant in LBH. Each of these is a live option, and there may be more options that I haven’t listed.
So, for example, you point out that the numerical sequence in the Siloam inscription – מאתי[ם ו]אלף אמה – has the smaller category first (hundreds before thousands), which is typical of P and LBH, but not typical of CBH (vol. 1, 155). (By the way, I think it’s best to regard P and Ezekiel as transitional, not as LBH proper). This is a genuinely interesting observation. Which option is it? Then you write that “the construction is also attested in EBH sources, e.g. 1 Kgs 5.12.” This muddies the waters, because this reading is a textual problem: MT reads 1005 (חמשה ואלף), but LXX reads 5,000 (probably חמשה אלף). Since we’re talking about Solomon’s wisdom, my bet is on the big round number in the LXX (which may differ only by a waw). My criticism here is that your “MT-only approach” prevents you from acknowledging that this is a textual problem, and therefore occludes your path to a warranted analysis. Unless there are other examples in CBH, I would infer that this is most plausibly option #1, a feature of vernacular speech that later became dominant in LBH. This inference could be wrong, but it seems the most plausible. The same inference applies for the other feature that you flag in this inscription, the construct of “100” (מ[א]ת). Regarding the third feature that you label as LBH – וילכו (“[the waters] went”) – I assume that this is an editorial error on your part, since it’s obviously not an LBH feature.
Each instance is interesting in its own way. There’s no simple rule, except to use the tools of critical scholarship in a circumspect manner. Errors will be made, but part of the task of scholarship is to engage in critical conversations with others, so that errors can be identified, and better construals of the evidence constructed. That’s what we’re paid to do (and we’re lucky for it).
#9 - Ron Hendel - 09/30/2011 - 03:29
Avi Hurvitz is the first scholar who explicitly reflected on the methodology of diachronic analysis of Biblical Hebrew. In his book Beyn lashon le-lashon (1972), he shows full awareness of the “imperfect” nature of the MT: textual alterations, redactional processes, intertextual references and other factors may have modified the linguistic profile of any given text (pp. 67-69). The way he deals with this problem is by designing the criterion of accumulation: the linguistic profile of a text can only be decided on the basis of a plurality of features (p. 69). His “MT only” methodology is not a principled posture, but a pragmatic decision: the linguist has to start somewhere, and the MT is not such a bad place. If the MT is corrupt, redacted, or composite, the criterion of accumulation will assure that the global picture remains more or less accurate.
Hurvitz would presumably be happy to learn that malkut in 1 Sam 20:31 is textually doubtful and that mamlakah is the better reading. If this is indeed true, it enhances the distribution of the term over the “CBH” and “LBH” corpora. But to Hurvitz it won’t make a big difference. The big picture is clear even with three or four exceptions: CBH uses mamlaka, while LBH often uses the more Aramaic-like malkut instead. Exceptions can be expected in a text that has been transmitted over hundreds of years.
#10 - Jan Joosten - 09/30/2011 - 08:19
One of our main hopes for LDBT was that we could move the discussion, previously about “linguistic dating,” on to a new stage.
In light of that, we appreciate the potentially fruitful discussion of our ideas exemplified by your latest response where you, for example, seriously consider the significance of LBH forms in Hebrew inscriptions.
Obviously, we still disagree on quite a few important issues. For example, we have documented how common so-called “late” features are in EBH texts and in Hebrew inscriptions and would resist any attempt to minimize the significance of this evidence. And since the large majority of “late” features in EBH texts are clearly not related to any attested textual variation, and since the textual variation attests multiple directions of linguistic modification (as mentioned above in the article), we consider it methodologically problematic to simply remove all “late” features from “early” texts by claiming that they are textual variants (of course some are but the majority and certainly not all are not).
But we have appreciated the opportunity to clarify the aims and methods of what we have done in LDBT, especially in view of some of the (to us: surprisingly) heated responses we have received to our work, and the misunderstandings we have mentioned.
We have never considered that we said the last word in LDBT, and we invite reasoned responses to our work which seriously engage with the data.
Thanks to all who have followed this discussion with such interest.
Ian, Robert and Martin
PS: The mention of HLK in LDBT is not an editorial error. The point was related to Frank Polak’s observation that the proportion of B) increases in LBH texts in proportion to HLK. We were pointing out that the pre-exilic inscriptions show the same ratio as LBH Esther in this feature, Esther having the highest ratio of B) to HLK in the Hebrew Bible. See again LDBT, vol. 1, p. 154.
#11 - Robert Rezetko - 09/30/2011 - 13:06
Robert, Ian, and Martin write:
"For example, we have documented how common so-called “late” features are in EBH texts and in Hebrew inscriptions and would resist any attempt to minimize the significance of this evidence."
This is the kind of statement that reflects the serious methodological flaws with LDBT and your subsequent efforts. First, categorizing texts as "EBH" and "LBH" and then pointing out how an item from the LBH books occurs a couple times in an EBH text is absurdly circular. Logically, you must either stop using the categories or try to explain how and why the item can exist in both. Your criticisms about previous studies start with promise but, due to statements like the above, end up as muddled rhetorical nonsense.
Second, you have been told again and again (by me, by Dresher, by Shalom Paul at the 2009 NAPH meetings) that the presence of one, two, or three examples of an item in an early texts does not necessarily disprove the diachronic story that is being told about the item, since linguistic change isn't a sudden process. It is quite possible that an item can begin as a borrowed word, used in rare contexts, and then later displace the earlier item to become dominant in later texts. Hence, the "accumulation" principle of Hurvitz's and the S-curve within the variationist literature (which you have said you're reading, but does not seem to have affected your arguments).
Repeating the same flawed arguments does not somehow make them correct.
#12 - Robert Holmstedt - 09/30/2011 - 16:34
Dear Robert H.,
The distribution and accumulation of typical LBH or “late” features in EBH or “early” texts is not a moot point. It stands at the heart of Hurvitz’ methodology. That is why we emphasized the accumulation of such features in pre-exilic Hebrew inscriptions and in all other biblical and non-biblical texts we studied.
As for individual features, we have studied a lot of them within the framework of the traditional approach, and we have found that the overall distribution often does not increase along the conventional EBH > LBH path or in agreement with books and sources that are traditionally conceived of (by linguists, literary critics, etc.) as pre-exilic or post-exilic. Often the problem is one of incomplete coverage of both the so-called early and late contrasting features. So more work needs to be done, as we, you and others know and are doing.
We agree with you that the “historical linguistic” working principles and methodology can change for the better in studies of biblical Hebrew, and so we have begun to incorporate in our work the S-curve within the variationist approach, largely due to insistence by you, Dresher, and others. (Indeed, Robert R.’s 2010 SBL paper and handout had already begun to integrate the methodology and vocabulary. That it has affected our thinking and argumentation will be evident in the book in preparation.)
Robert, Ian and Martin
#13 - Robert Rezetko - 09/30/2011 - 18:37
Sorry guys, but your claim that the verb "went" (converted imperfect of hlk) in the Siloam inscription is a feature of Late Biblical Hebrew is nothing short of bizarre. You have confirmed my initial diagnosis that you're playing fast and loose with the evidence. I'm signing off now from this exchange.
#14 - Ron Hendel - 10/01/2011 - 01:05
By no means do we say what you assert. You say that we say: “the verb “went” (converted imperfect of hlk) in the Siloam inscription is a feature of Late Biblical Hebrew.” We do NOT say this!
In fact, we say the OPPOSITE! The lexeme hlk and the waw consecutive form of hlk (wylkw) in Siloam 4 are clearly NOT features of LBH (vol. 1, pp. 154-155).
However, at the bottom of p. 154, including also n. 15, we point out that the proportion of b) to hlk in the inscriptions is 9 to 1, that is, 9 b) and 1 hlk. That is the same proportion, for example, as in Esther, 9-1. According to the work of Polak, LBH is marked by a rise in frequency of usage of b) compared to hlk.
The point then in the discussion on pp. 154-155 is NOT that the Siloam inscription has a feature of LBH, but that it is the EXCEPTION, because the inscriptions clearly prefer the lexeme b) as in LBH. That overall preference for the lexeme b) in the inscriptions (NOT the particular lexeme/form in Siloam!) is the link with LBH that is stressed on p. 156 (#14).
Robert, Ian and Martin
#15 - Robert Rezetko - 10/01/2011 - 08:40
Like you we all respect Avi Hurvitz’s work, and have constantly striven to stress those areas of continuity between Avi’s work and ours.
Our current state of research on the textual problem would indicate that Avi was right to stress the very large accumulations of “LBH” features in the core LBH texts of Esther, Daniel, Ezra, Nehemiah and Chronicles, as being possibly significant even in the context of extreme textual fluidity that characterized the transmission of ancient texts. Hence we still consider those books a distinct group. However, no other texts have such a degree of accumulation, and the other texts that Avi has discussed like the Prose Tale of Job have only a handful of “LBH” features, which could easily all be added or subtracted in scribal transmission (we have ancient texts which add or subtract similar numbers of “LBH” features).
Thanks for stressing the important role that Avi Hurvitz’s work has played in shaping this field.
Ian, Robert and Martin
#16 - Robert Rezetko - 10/01/2011 - 08:41
Two worlds that shall never unite:
The discussion between Ron Hendel and Ian, Robert and Martin is probably leading nowhere as the presuppositions are too different.
Let me illustrate this with a quote from Ron Hendels Remembering Abraham p. IX:
"The remembered past is the material with which biblical Israel constructed its identity as a people, a religion, and a culture."
Apart from the old fashioned way of defining ethnicity here, it says that what we have in the OT is the memory of biblical Israel. The logic problem is obvious: The thing remembered is biblical Israel. It is nowhere except in the Old Testament. So the argument is circular: As a matter of fact, Hendel is saying that biblical Israel is construing itself.
However, when this argument is lifted into the present discussion, it means that biblical Israel is a reality to Ron Hendel, although we only have it in the Bible. Following the track in the OT of this Israel, it becomes easy to date texts as belonging to this biblical Israel, otherwise the whole exercise becomes meaningless.
Bringing in one word from the Sfire inscriptions here is meaningless except when said that the Sfire appearance is a terminus a quo, but definitely not a terminus ad quem. We should never confuse these. Sfire is from the second half of the 8th century, making them in Ron Hendel's optic (I believe) contemporary with Isaiah. We can make speculations about the importance of Sfire here--there should at least be a demand for more similarities than one single word, something more systemic, but I don't think that it dates anything.
The supposed archaic language of Numbers 22-24 has of course been noted, as should Luther's reference here to the Macedonians. This is a special mantra which the late Fred Cryer often commented on, saying (including Debora)that speaking about language differences, we have local ones, dialectal ones, sociological ones, and still without linguistic anchors -- a concept probably not understood -- we cannot date any text and thus make it into an anchor. Again the argumentation will become circular.
And just to repeat: a circular argument is a false argument. A false argument can be dismissed right out of hand: We don't need to pay attention it anymore (and should not).
So, bring up lingvistic material based on inscriptions from the Iron Age and compare this to what we may have of inscriptions from, say the Persian or the Hellenistic period. Create a framework for the development of datable Hebrew. Then it might be timne to get to the biblical Hebrew and see what is going on.
Niels Peter Lemche
#17 - Niels Peter Lemche - 10/02/2011 - 07:38
Dear Ian, Robert and Martin,
I wasn't trying to stress the important role of Hurvitz's work (in the present context that would be somewhat self-defeating), but to react to the query in one of your posts: "As regards Avi Hurvitz and the MT, he may or may not *know* that his MT-only approach is, in principle, incorrect, but we would be very interested to know where he *says* this." He knew, and said, and devised the criterion of accumulation to "pick up the slack".
#18 - Jan Joosten - 10/02/2011 - 11:03
To start anew the BH historical linguistics sounds like a very ambitious and challenging goal, but one has to be a trained linguist for that sake. LDBT does not posit any proper linguistic problem and does not work with any sound linguistic methodology. It simply reformulates the data collected by Hurvitz and other scholars in order to make a pure theological claim - biblical texts can not be dated (linguistically). Moreover linguistics does not work with separate words and features, it has to deal with the whole system. Even Avi Hurvitz who might be not the most up-dated linguist nowadays suggested very systematic analysis of BH lexicon and phraseology. Joosten, Par-El, Cook and others are doing the same with the syntax. Just let everyone do their own job.
#19 - Tania Notarius - 10/02/2011 - 13:37
What we are suggesting is that everyone should “do their own job” but that there should be significantly more collaboration between trained linguists, trained textual critics, trained literary critics, trained historians of religion, trained historians, and so on. The formation of the Bible and, therefore, the history of the language attested in the Bible are simply too complex for any one of these to sort out all the chronological details or for linguists on their own to write a history of biblical Hebrew. The work of the people you mention (Joosten, Pat-El, Cook) and many others (e.g. yourself, Naudé, Holmstedt, Bar Asher) is a crucial component but it will be more valuable if thoughtfully integrated with other kinds of data.
Martin, Ian and Robert
#20 - Robert Rezetko - 10/02/2011 - 22:52
One last footnote. In their discussion of the Siloam inscription, YR&E “suggest three possible links with LBH (14, 17, 18).” (vol. 1, p. 156). Feature number 14 is וילכו. I assumed (charitably) that this was an editorial error on their part. They reply that it is not an error, yet they agree that וילכו is “clearly NOT” an LBH feature. They say that the LBH feature in question is the frequency of the verb “come” in other inscriptions. Now, while I enjoy the logic of Alice in Wonderland, it has no place in scholarship. Only the Red Queen can make words mean whatever she wants - and their opposite.
#21 - Ron Hendel - 10/03/2011 - 00:20
We see your point that the sentence on p. 156 is tersely and unpropitiously worded, giving the first impression that "(14) וילכו" itself is the LBH feature in question, but the discussion on p. 154 and n. 15, and indeed even how the discussion of point 14 ends on p. 155 ("...in some ways [the Siloam Tunnel inscription] is NOT typical of the inscriptions as a whole [which prefer the LBH ratio]" [emphasis added]), construe how the wording on p. 156 should be read in context.
Robert, Ian and Martin
#22 - Robert Rezetko - 10/03/2011 - 06:05
The Red Queen is a character in Through the Looking Glass who runs hard to stay in the same place - half admirable, half perverse. Several Oxford scholars may have served as models. The claim to sovereignty over words is made by Humpty Dumpty, who is a mixture of pomposity,kindliness and (misplaced?)faith.
My job used to be in philosophy - perhaps, logic having been mentioned, I can make a few points.
The propositions -
1.Early composition implies early language
2.Early language implies early composition - do not imply each other (even in the event that both are true) and so do not form a circular argument. (If one prefers to say 'predominantly early' the logic is the same: 'p implies q' doesn't imply 'q implies p'.)
To infer 1 from 2 a further proposition, such as
3. Early writers do not innovate, at least not enough to affect the argument
is needed - and to infer 2 from 1 you need
4. Late writers do not archaise, ie write in early language, at least not enough, or not with enough plausibility, to affect the argument.
Common sense might suggest that 3 is fairly plausible, 4 comparatively doubtful.
So is the substantial point at issue the question of how this doubt should be resolved? Were late writers able and willing to write extensively - and significantly beyond our power to detect their hand at work - in early language and manner?
#23 - Martin - 10/03/2011 - 19:40
It was indeed Humpty Dumpty and not the Red Queen. My thanks to Martin for the correction. The full passage is as follows:
'When I use a word,' Humpty Dumpty said, in rather a scornful tone, 'it means just what I choose it to mean — neither more nor less.'
'The question is,' said Alice, 'whether you can make words mean so many different things.'
'The question is,' said Humpty Dumpty, 'which is to be master — that's all.'
Alice was too much puzzled to say anything; so after a minute Humpty Dumpty began again. 'They've a temper, some of them — particularly verbs: they're the proudest — adjectives you can do anything with, but not verbs — however, I can manage the whole lot of them! Impenetrability! That's what I say!'
#24 - Ron Hendel - 10/06/2011 - 16:39
A late comment on an issue for which I was quoted: the relationship of BW)'come' and HLK 'go' in the inscriptions. Robert, Ian and Martin argue that in my view the decline of HLK vs BW) is a matter of LBH. Actually I am speaking of t h r e e corpora in BH rather than of two: a classical, a pre/Early -Babylonian (VIIth/early VI century) and a Persian/Late Babylonian corpus. In my view the decline of HLK is already notable in the VIIth century corpus, roughly the same period as the epigraphic texts quoted by R,I and M.
As to Malkuth, it may have been an option in CBH, but appears three times only, and thus clearly was not a preferred option.
I don't think it is right to limit analysis to type, token frequency should be taken into account.
A third matter. Hurwitz' argumentation is not circular, as Lemche seems to think (maybe not after reading his work?). In many features 'P' language precedes Ezekiel typologically (unfortunately some of those features do not appear in the LXX of ezekiel and could be regarded as late additions). With regard to the date of P this phenomenon should not be disregarded. This argument is not circular.
Best regards, if you are still following this thread,
Best regards, if you still are following
#25 - Frank Polak - 06/11/2012 - 09:22