I care about how we do things in biblical studies, not because I give a flip about current scholars (I can simply not read work I don’t think worth my time), but because I am sensitive to what young minds gravitate towards. And if there is one thing that is insidiously attractive to young minds it is easy thinking. And I think any approach that deliberately eschews clear methods for handling data and clear theories for interpreting data is very dangerous. And this is precisely what I sense in the movement to reclaim philology as a useful term in contemporary biblical scholarship.

Below are excerpts from an essay that has its origin in the joint linguistics and philology session at SBL 2018. If anyone was wondering why I wrote my phrenology (I mean, philology) post, this perhaps explains why.

Facts are the world’s data. Theories are structures of ideas that explain and interpret facts.
Stephen Jay Gould

1. Introduction

It may be inevitable that every discipline enters into periods of self-reflection and reorientation to the object of study, the methods used for studying that object, and relationships to overlapping disciplines. Such reflection has the potential to be a very a healthy process, though the degree of health depends greatly on the nature of the catalysts and the tenor of the ensuing discourse. The status of philology as an activity and then a definable academic discipline has changed a great deal in the last two centuries, as much if not more than any other scholarly endeavor:

… probably no other discipline in the vast spectrum of academic fields has undergone as sweeping a transformation as philology has during its history. Since the days of Karl Lachmann, it seems that nearly every aspect of it has changed radically, be it the subject, scope, or methodology. (Bajohr et al 2014: 1; similarly, see Pollock 2015: 2-3)

In the latest round of soul-searching, which has unfolded over the last decade or so in general, the concern has been less about philology and linguistics carving out their own turfs within academia, but more about whether philology has any turf at all. This question has become more intriguing in the last few years for those in Biblical Hebrew studies, for two reasons. First, the most visible and accessible scholarly venue for sharing current linguistic work on Biblical Hebrew was eliminated as an independent unit at the discipline’s largest academic society (see here for background); for reasons that were neither entirely transparent nor coherent, the unit can only organize sessions in collaboration with other units, such as the units representing philology or Hebrew poetry. At the same time, two independent collaborations began: one with the focus of “renewing” philology, the other with the focus of bridging the divide between philologists and linguists.

Circumscribing fields of inquiry need not be an exercise in limitation, but one of empowerment, enabling scholars to identify among a myriad of theories and methodologies which may be most appropriately applied to a given research question. Given the current state of Biblical Hebrew studies, it is thus eminently useful for current and future scholars to explore how “philology,” a prestigious term associated with centuries of rigorous and important scholarship, may or may not have continued utility in future academic discourse.

2. A Brief History of the Study of Biblical Hebrew

….

3. What Future Has Philology?

….

4. The Role of Theory in Language Study

….

5. Conclusion
Language research without theory is not research, but scientifically naive empiricism; historical linguistics, socio-linguistics, typology, and so on, refer not to theories, but to areas of focus, bundles of methods, or to return to the parable, pieces of the elephant. These tools or lens only achieve intellectual legitimacy and coherence—the elephant can only be perceived for what it is—when they are situated within an overt theory of language.

Where does this leave philology in contemporary scholarship? If we equate philology roughly with textual criticism (e.g., Hendel 2016), it has a place in manuscript culture, sometimes now situated within editorial theory. If philology is used in relation to historical linguistics (e.g., Barr 1969), it can be situated as historical study of the language (e.g., ancient Hebrew) within a specific linguistic corpus (e.g., the Bible). In contrast, philology defined so that it applies to any activity related to reading texts has no distinct identity, for it becomes little more than a synonym for interpretation, broadly construed. A glance at the modern university already drives home the deleterious impact of fuzzy scholarship on the humanities; can it survive more, under the guise of a resurrected philology?

(As a footnote to my last sentence, I point to a number of “philology” papers given at SBL the last two years—many have bizarrely little transparent connection to language or close-reading of text. I encourage curious readers to read some of these paper titles and abstracts yourselves. I recommend having a strong drink handy when you do!)

I gave a paper at SBL last year in which I discussed what I thought was the obvious and sensible notion that theory is necessary for analysis. For language, this means some sort of clear and coherent theory of language. The exhumation of philology has problems with theory, but that doesn’t seem to be stopping the grave diggers. Apparently theory avoidance is a malady that is hard to get rid of and spreads easily.

Anyway, I spent about 5 minutes with the announcement, which is also posted here (I shall remain hopeful that it’s not a permanent condition), and had a little Swiftian fun by changing a few words. So read further for some Fall fun.

RENEWED PHRENOLOGY

Cranial Studies for the 21st Century
Renewed Phrenology is an international working group of scholars in cranial studies whose work reflects critically on the intellectual frameworks brought to bear in phrenological practice. The work of this group is diverse in its interests, interdisciplinary in its readings of craniums and the return to pre-scientific world-views, and meta-critical in its practices. We are comprised of a core group of researchers and are supported by a broader advisory board.

Why we started this group
The systematic study of craniums, their fissures and their bumps has been central to education and intellectual activity since the development of the haircut itself. Phrenology has been foundational to approaches to the cranium-mind relationship and the development of ideas about personality traits.

In the history of ideas, the scholarly and pedagogical activities that are constitutive of phrenology have been essential in the formation of our identities through early childhood head shaping and in situating and reintroducing methods sans theory and rigor in the twenty-first century.

In the modern academy, scholars from different disciplines apply similar methodologies on sometimes vastly different craniums. But sometimes ongoing work labors without the benefit of a deep understanding of phrenology’s own history as a discipline and in particular, the inaccuracies its ways of reading craniums have historically produced.

We seek to reshape phrenology’s future through understanding its past. Through our work, we hope to renew the value of careful attention to cranial landscape and the cultural values encoded in cranium shape, size, and terrain and discover forgotten or overlooked cranial patterns within phrenological practices. Faculty and students working in diverse disciplines stand to benefit from learning about each other’s work in phrenology inside and outside the classroom. While our focus is specifically on cranial phrenology and the unique challenges faced to the study of these “head bumps,” our group is engaged with ongoing work in other pseudo-scientific disciplines.

What we are interested in is feeling craniums and drawing unwarranted personality conclusions. We are particularly concerned with what cranianists (and practitioners of other related areas) mean when they speak of “phrenology,” as both a set of cranial reading practices and an integrated theory of craniums. We are interested broadly in interventions in phrenology and its historical-critical posture that reflect critically upon gender, race, class, religious identity, and other subject positions embedded in the ridges, fissures, and bumps of the cranium.

We are interested in posing broad and un-empirical questions of craniums and the environments that produced them and are thus a reflection of a set of personality traits, such as hair production, hygiene, scalp shininess, cranium identity, cranial community, and the cranium in the imagination.

To learn more, see our page on recommended further reading.

What we do
The members of this group are engaged in a number of research projects, devise and infiltrate a number of workshops and conferences under the mantle of Renewed Phrenology.

Core members of this group are founders and chairs of key program units within the Society of Pseudo-Scientific Inquiry, including the program units “Phrenology is a Science, Really!” and “Why Won’t Medical Schools Let Us In?”.

Learn more about us
Photos of recent events
See calendar of upcoming events

*Even my popular essays have footnotes. I will not claim that this is some inherently good thing. As I wrote the essay, I began to wonder if I’d lost the ability to write for those who aren’t specialists in my field. Perhaps someday I’ll write a “Biblical Hebrew Grammar for Dummies” without a single footnote. Speaking of which, a departed friend and colleague once told me that best thing about emeritus status was no longer needing to support every idea in an article with footnotes! He showed me his last article, and it didn’t contain a single one. By the way, if you’re wondering, I wrote this footnote simply for the fun of it and for the “meta-ness” of it all (for inspiration, see below, M. Fox, “On Footnotes, Hebrew Studies 1987).

Fox_OnFootnoes_HS1987

A review of Douglas Petrovich, The World’s Oldest Alphabet: Hebrew as the Language of the Proto-Consonantal Script, with a contribution by Sarah K. Doherty and introduction by Eugene H. Merrill (Jerusalem: Carta, 2016).

Shortly after Douglas Petrovich’s book on the early alphabetic texts appeared at the end of 2016, I was asked to deliver a guest lecture on the topic during my last sabbatical. Before the request, I wasn’t aware of Petrovich’s monograph, though I had noticed various online comments and essays (even interviews) he’d given on the subject.

I will admit that I found it odd that Petrovich should be weighing in on the discussion since he studied no Hebrew grammar or Northwest Semitic epigraphy with me at the University of Toronto, where he wrote his doctoral thesis on Near Eastern archaeology.

And so the lecture was an opportunity to consider in detail the argument Petrovich mounted about the early alphabetic texts, which I’d not previously given a great deal of thought beyond discussing them during the first week in my Hebrew epigraphy seminars.

My initial impressions of Petrovich’s book were threefold. First, as with most Carta products, the production quality was high. The maps are wonderful, the illustrations clear and useful, and the paper is both heavy and high quality. Second, the amount of research that went into this study was striking. Which brings me to the third impression—it was mind-boggling that all this effort was spent on an argument that was explicitly tied to Petrovich’s assumptions about the Bible as an accurate historical resource and his desire to bolster the traditional chronology for the Patriarchs and Moses.

And as I continued the book, I was deeply disappointed that a work with so many flawed notions of Hebrew grammar would have been published. (And, as I soon discovered, the analysis of Egyptian was equally problematic.) Moreover, Petrovich’s writing style often reads more like a set of course lectures and he often cites sources in an exaggerated way if he wants to impress the reader with their authority.

Petrovich’s argument has been responded to by Alan Millard, Christopher Rollston, and on Rollston’s blog also the Egyptologist Thomas Schneider (links provided further below). I will also soon be addressing the topic on another online forum (and when it appears I will insert a link here). But none of these responses address the serious flaws of Hebrew in Petrovich’s analysis. This is what I will cover in this post.

See my review below …

1. Book Summary

In his 2016 volume on the early alphabet epigraphs from Egypt and the Sinai in the second millennium, Douglas Petrovich argues that the language of the “proto-consonantal” script used in a scattering of epigraphs from Wadi el-Ḥôl, Lahun, and Serâbîṭ el-Khâdim is a Hebrew antecedent to the language of the Bible and that this provides linguistic support for the historicity of Moses and his composition of the Pentateuch “by 1406 BC, the year of his death and the beginning of the conquest of Canaan” (195). Beyond the complex issues presented by the text themselves, Petrovich presents his work as the dawning of a new day in biblical studies, as the springboard for a serious challenge to the “critical-scholarly world [that] has cleverly built an impenetrable force field around the evidence” for “the historical credibility of the biblical account of the Israelites in Egypt from Jacob’s day until that of Moses” (188, 186).

Besides an introduction (Chapter 1) and conclusion (Chapter 2), Petrovich’s book consists of two chapters in which he works through the primary artefacts with “proto-consonantal Hebrew” (PCH) writing. In Chapter 2, he covers the inscriptions he assigns to Egypt’s Middle Kingdom: Sinai 115, 376, 377, Wadi el-Ḥôl 1, 2, and the Lahun “bilingual” ostracon. In Chapter 3, he introduces the context of the Sinaitic epigraphs found at Serâbîṭ el-Khâdim and then analyzes Sinai 345a,b, 346a,b, 349, 351, 353, 357, 360, 361, 375a, and 378. After the conclusion, Petrovich includes four appendixes: (1) on the original 22 letters of the PCH alphabet, (2) the “additional” five proto-alphabetic letters, (3) a declining and parsing guide for Middle Egyptian and PCH words, and (4) a chronology of relevant ancient Egyptian dynasties.

Petrovich concludes that the language of the “PCH” script can be confidently identified as ancient Hebrew, for three distinct reasons (191). First, Petrovich has identified the proper noun “Hebrews” in the caption text of Sinai 115. Second, “every single proto-consonantal letter was found to have a ME hieroglyphic exemplar from the ME sign list, and to match with a corresponding Biblical Hebrew (BH) word that is logically and acrophonically connected to the meaning of the pictograph.” And third, in three separate texts Petrovich reads three proper names he identifies as biblical persons:Ahisamach(Sinai 375a; Exod 31:6), Asenath(Sinai 376; Gen 41:45), and Moses(Sinai 361; Exod 2:10).

On paleographic matters, interested readers should visit Christopher Rollston’s blog and his two posts on the topic (here and here). It is also useful to read Alan Millard’s post on the ASOR blog “Ancient Near East Today” (here). Additionally, the guest post on Rollston’s blog by the Egyptologist Thomas Schneider is important (here), as is the fact that Petrovich’s Egyptian teacher, my U of T colleague Prof. Ron Leprohon, has called into question not only Petrovich’s readings, but also the presentation of at least one of the core texts (on Sinai #115, Leprohon pointed out to me how Petrovich cropped the text in a way that favors his reading but does not present the entire object and writing evidence).

Petrovich has responded to each of these reviews with lengthy comments posted on his academia.edu site. In fairness his responses should also be consulted.

In this post, I will will cover a number of issues that are not dealt with in any of the other reviews and not yet re-addressed by Petrovich: his flawed understanding of Hebrew grammar and the implications for his reconstructions.

2. Problems with Hebrew Grammar in Petrovich’s Analysis

At the core of Petrovich’s argument is not just his identification of the scriptas “Hebrew” but the wordsandgrammaras Hebrew, more specifically, a Hebrew that is nearly biblical in shape. For each text, he provides a BH transcription as well as a list of BH words (with biblical references) that he has identified in the text. Some of the claims about Hebrew grammar, however, raise doubts about Petrovich’s knowledge of the language as well as his understanding of the methods of comparative and historical Semitic linguistics. I will illustrate this with two of the more egregious problems.

Gemination

My first concern primarily centers on Petrovich’s understanding of gemination, but this will lead us through a few other issues on the way. For instance, Petrovich reads the sequence of Middle Egyptian signs in Sinai 115 as G-B-Ỉ-T-U, for which he claims that the B was “doubled,” or “written once, but spoken twice,” with one segment ending the first word and the other segment beginning the second word. Based on this “perfectly legitimate Hebrew grammatical convention,” he interprets the sequence as gb [b]ỉtu “earth-god’s house,” which he equates with BH bêt-ʾēl (26-27). The phrase thus represents a Hebraized (note the BH word for “house”) Egyptian phrase (note the possessive noun in the first, bound position) in Egyptian writing. There are multiple issues to address here.

First, Petrovich’s explanation for the use of an Akkadian word /bitu/ instead of the expected Hebrew form /bayit/ is worth noting:

If Abraham truly originated from Ur (Gen 11:31), where he would have spoken Akkadian before departing for Canaan, presupposing Ḫebeded’s [the supposed author of Sinai 115] knowledge of Akkadian would be extremely plausible …. the Hebrews either used bitu for house, following Akkadian rather than Canaanite, or they simple chose Akkadian over Hebrew, since written Hebrew text was unknown to the world up until this point in time …. [If] the former option is correct, the initial vowel-class could have changed from i to e (bit to beṯ). The Hebrew Caption thus would be preserving a pre-formative moment in the Hebrew language, before bitubecame beṯ. (27, 29)

That Abraham spoke Akkadian versus, say, a West Semitic language like Amorite is debatable in the abstract, though the absence of any evidence at all for either view renders it wholly speculative. The equation of Canaanite and Hebrew is inaccurate at best (apart from the decidedly un-Hebrew language found in the second millennium el-Amarna texts, “Canaanite” refers to a first millennium dialect group) and spurious at worst, if intended to put the existence of Hebrew in the second millennium on firm linguistic ground by linking it to the el-Amarna artefacts. Moreover, authors and speakers do not “simply choose” to use one language or another at arbitrary points; research into multilingualism and code-switching indicates that there are clear patterns and motivations behind switching. The burden of proof for such a switch in the Sinai epigraphs lies solely on Petrovich, though none was given.

Also, his suggestion that the change of bitu to beṯcould simply reflect a dropping of the final a and a change of the word-internal vowels betrays an ignorance of both Akkadian phonology (i.e., the rule accounting for the change of *baytu to bitu) and the fact that Hebrew, as a West Semitic language, experienced a different history for the monophthongization of the –ay– in *baytu resulting in beṯ.

I strongly doubt that Petrovich’s complex phrase, which is neither Hebrew nor Egyptian, would have made any sense to either Hebrews or Egyptians in the second millennium. In any case, gemination does not work in the way that Petrovich presents it.

Gemination is not a “doubling” or copying of a sound segment, but the prolonged pronunciation of the segment. Admittedly, this is often obscured in typical Semitic transcription conventions, where, for instance, a geminated Bis written as two graphemes –bb-; the international phonetic alphabet convention more accurately signals that gemination remain a single grapheme by using the sign ː, hence a geminated Bis rendered bː.

There are two contexts for gemination in BH. First, the requirements of a word pattern itself may include gemination, such as with the middle root consonant in the Pi‘el: דִבֵּר dibːer. Second, the cliticization of the article haC– (with its indeterminate final consonant) or preposition min– result in gemination of the initial consonant of the host noun (presumably both due to anticipatory assimilation of the final consonant of the cliticized particle), e.g., *haC-bayit > habːayit ‘the house’ and *min-bêt > mibːêt ‘from the house of’).

The point of this for interacting with Petrovich’s claims is that gemination does notcreate a consonant that “was written once but pronounced twice” (157). Moreover, the lengthening of a consonantal segment occurs entirely within words, not across word boundaries. Even apart from the assertion that a dagesh forte was used in Sinai 345b—two millennia before the Masoretes developed their notational system!—it is simply ad hoc to argue that the “doubling” convention was used across word boundaries as “a much freer use hundreds of years before the oldest extant manuscripts of the Hebrew Bible are attested” (89).

If the phenomenon that Petrovich describes was operative in the phonology and represented in the writing system of those who wrote the early alphabet texts, it would be more accurate to identify it as a type of external sandhi, specifically, anticipatory sound assimilation across a word boundary (of course, sandhi could have been operative in the phonology of the language of these early inscriptions without it being represented in the writing system).

External sandhi, though, is rare in early West Semitic, and when it does occur, it appears to be related to set phrases, names, or titles. Consider, for example, Byblian Phoenician byḥmlk (for *bn yḥmlk) ‘son of Yḥmlk’ (KAI 6.1, 7.3; also KAI 8), Cypriot Phoenician mlkty ‘king of Kition’ (CIS I,11.2, but otherwise always mlk kty), and epigraphic Hebrew ḥyhwh ‘life of Yhwh’ (Lachish 3.9, Arad 21.5, but see Lachish 6.12, 12.3 for ḥy yhwh). Sandhi across word boundaries does not occur in the Hebrew Bible except in two names, יְרֻבַּעַל (<*yarûb-baʿl) in Judg 6:32 (+12x) and מְרִי־בַעַל‎(//מְרִיב בָּעַל‎) in 1 Chr 9:40. In general, see Lipiński 2001: 202-3; for West Semitic onomastica, see O’Connor 2004, esp. 450, n. 41, 465, n. 142; for Phoenician, see Friedrich and Röllig 1999: 56; Krahmalkov 2001: 124; Gzella 2013: 181; for the Canaanite dialects, see Garr 1985: 40-43.

Within the Sinai corpus, Wilson-Wright has argued that mʾhbʿlt in Sinai 345 is an example of sandhi and thus parallel to mʾhb bʿlt ‘beloved of the Lady’ in Sinai 374. Wilson-Wright suggests one additional example, mhbʿlt in Sinai 348, which is the same phrase but missing both the b of the sandhi process as well as the ʾof the root of the first word (2013: 142, n. 16). However, this particular artefact was lost in the early twentieth century and the best line-drawing is based on the squeeze, which casts serious doubt about our ability to use any data from this text (see Albright 1969: 17-18)

Since external sandhi operates on adjacent sound segments (i.e., no other consonant orvowelcan intervene), Wilson-Wright also proposes that this indicates that the language of these early texts reflect the loss of short final vowels, at least on bound words. He notes that this parallels the loss of case vowels on bound words in Old Akkadian but differs from Northwest Semitic. In this branch of Semitic, the final alefsin Ugaritic and the spelling used in early Byblian epigraphs suggest that short final vowels did not quiesce until the early first millennium. I am not convinced by Wilson-Wright’s argument, especially since his other arguments about the language of Sinai 345 draw heavily on parallels to Central Semitic and Northwest Semitic features.

In summary, of the five cases of “gemination” across word boundaries that Petrovich proposes, only three examples could actually qualify, though reconfigured as external sandhi: gbbỉtu in Sinai 115, mʾhb bʿlt in Sinai 345, and šnt tmhn in Sinai 361. The other two, bʾt tzzt =בָּאתָ תֹּזַזְתָּ and bšn nš = בְּשָּׁנָה נְשֹׁה, in Sinai 360 clearly do not qualify, since the consonants Petrovich takes as “doubled” are separated by a vowel that never quiesces in the respective paradigms of ancient Hebrew (2ms perfect verb affix and feminine nominal affix, respectively).

I could also include in this discussion Petrovich’s approach to the “doubling” involved with II-III or geminate root verbs. Thus, in Sinai 377 he claims that the sequence M-L is the verb מָלַל “he has scraped, inscribed,” for which the second L of the root “would have been written once but spoken twice, as practiced typically for geminate verbs during the Bronze Age” (33). This is not an accurate understanding of gemination or of the morphology of geminate verbs. See also similar explanations for verbs based on the root Z-L in Sinai 346a [92, 98] and Sinai 349 [103], S-Bin Sinai 351 [118, 123], Q-M in Sinai 353 [129, 137], ʾ-R,G-N, and ʾ-N in Sinai 357 [145, 151], B-Š in Sinai 361 [160, 169].

Moreover, given the un-Semitic structure of the proposed phrase gb bỉtu in Sinai 115 (with the possessive noun in the first, bound position) as well as the necessary conjecture regarding the loss of short final vowels (which does not occur this early in West Semitic outside of onomastica), I think it highly unlikely that external sandhi was a productive morphophonological rule in ancient West Semitic. It was either limited to names and titles, as it is in Amorite evidence (see O’Connor 2004), or the three possible cases in the Sinai corpus simply represent writing errors.

See also similar explanations Petrovich gives for gemination across a word boundary in Sinai 345a [83, 89]: mʾhb ʿlt (=mʾhb bʿlt מְאֻהָב בַעֲלָת [sic]); Sinai 360 [155, 157]:bʾt zzt (=bʾt tzzt בָּאתָ תֹּזַזְתָּ) and bšn š (= bšn nš בְּשָּׁנָה נְשֹׁה); and Sinai 361 [160, 169]: šnt mhn (=šnt tmhn שְׁנַת תִּמָּהוֹן).

Feminine Noun Morphology

The second grammar issue I will discuss concerns the affix attached to feminine singular nouns. Petrovich assumes that in second millennium Hebrew such nouns already had the inflectional -ā^h ending (with h as a mater lectionis) in the free form. Relatedly, he frequently claims that because matreslectioniswere not used until the first millennium, a number of identified nouns could be feminine in the absence of any inflectional ending (see pp. 43-44, 123, 136, and many other supposedly feminine nouns whose lack of a feminine ending is not explained).

Relatedly, note Petrovich’s reading of the sequence M-R-ʿ-T-W in Sinai 346a as the noun מַרְעִית “pasture” with an attached 3ms possessive pronoun וֹ “his.” What Petrovich seems to miss is that the waw in the BH form of the attached pronoun does not have a consonantal value; it is, rather, a mater lectionis, a place-holder indicating a vowel (which is specified as an /o/ in the later Masoretic notation of the holem over the waw).The problem this raises for Petrovich concerns the history of the writing system, since the use of matres lectionis was not a spelling convention in any NWS language of the second millennium. Even in Ugaritic, towards the end of the second millennium, there is no clear evidence for the use of matres lectionis; see Pardee 2003: 26-35. Though Deitrich and Loretz 2008, followed by Riley 2013, argue that Ugaritic spelling of Akkadian words suggests the use of matres, borrowed words are more often adapted to native spelling conventions than new spelling conventions created to accommodate borrowed words. Thus, the argument for Ugaritic matres remains weak.

While Petrovich is correct that matres lectionis were not employed until the first millennium, it is unlikely that he is correct on any other point. The evidence of Ugaritic, Phoenician, Moabite, and even some early Hebrew epigraphs (e.g., שׁת in the Samaria Ostraca) unequivocally demonstrate that the feminine singular free form ending was –(a)t in Northwest Semitic through the second millennium and into the early first millennium. In fact, even the briefest research into comparative Semitic research would have been sufficient to derail his argument if he had looked; for the interested reader, I suggest starting at page 236 in Lipiński 2001.

The ambiguity that Petrovich proposes does not work for the vast majority of nouns, which are clearly inflected as masculine or feminine. The masculine noun “calf” would have been written ʿGL whereas the feminine “heifer” would have been ʿGLT.

Petrovich also treats the glides Y and W and the related issue of H as a final root letter in a similarly implausible manner. For every sequence that Petrovich identifies as a form of a verb that in BH is a III-H root, such as ʾ-P (supposedly אפה “to bake”) in Sinai 377 and Wadi el-Ḥôl 1, he explains the absence of the root-final H in the alphabetic text as “expected” because writing the H was a “later orthographic convention” (34). The historical development of originally III-Y verbs, which become III-H in BH, is quite complicated and the details of the process are not always clear. While small number of second millennium (Middle Babylonian) words do not exhibit an etymologically-expected final Y or W, the weight of evidence, including attestations in the earliest first millennium Phoenician and Moabite epigraphs, attest the written representation of a final Y or W. We cannot, therefore, simply assume that a proposed second millennium form of Hebrew would have the same morphophonology as its first millennium descendant. Petrovich’s treatment of this complicated issue does not reflect any awareness of the methods, data, or conclusions common to historical and comparative Semitics.

See also these examples adduced by Petrovich: Wadi el-Ḥôl 2:Y-G-K, which is supposedly “your afflicter,” a Qal 3ms participle of יגה with 2ms pronoun; Wadi el-Ḥôl 1: N-G for BH נֹגַהּ ‘light’, which is a true III-H verb in BH, making the absence of the H particularly unexpected; Sinai 349: ʿŠ for BH עָשָׂה, which is a III-Y verb and K-M for כָּמַהּ, but this root is a true III-H root and the final H should not be missing; Sinai 353: ʾ-T for BH אָתָה, which is a III-Y verb; Sinai 357: ʾ-N inflected here as an unattested Pual verb for BH אנה, a III-Y verb; Sinai 361: M-Š for BH מֹשֶׁה, which is related in the biblical text to the root מָשָׁה, a III-Y root; Sinai 375a: G-ʾfor the BH ms participle גֹּאֵה (note Petrovich’s incorrect pointing in the last syllable), a III-Y verb and ʿ-P for BH the II-Y root-based עָיֵף, a form in which the Y is consonantal and would have been written.

Reliance on Rare Hebrew Words

A final grammaticalobservation on Petrovich’s analysis concerns his reliance on biblical hapax legomenaor very rare words, whose meanings even in the biblical text are open to debate. Below I list, with Petrovich’s glosses, words he invokes that occur three times or less in the Hebrew Bible:

Lahun: רֹן “celebration” (1x, Ps 32:7); Sinai 345b: אֱיָל “strength” (1x, Ps 88:5); Sinai 346b: צֹנֶה ‘sheep” (1x, Ps 8:8); Sinai 349: שַׁעַר “measure” (1x, Gen 26:12), כָּמַהּ “to yearn” (1x Ps 63:2) and תְּלִי “quiver” (1x, Gen 27:3); Sinai 351: שֹׁקֶת “watering trough” (2x, Gen 24:20; 30:38) and צָבֶה “swollen” (1x, Num 5:27); Sinai 353: קוֹמְמִיּוּת “erect” (1x, Lev 26:13) and בלם “to restrain” (1x, Ps 32:9); Sinai 357: סָךְ “multitude” (1x, Ps 42:5) and טושׂ “to swoop” (1x, Job 9:26); Sinai 360: אשׁשׁ “to take courage” (1x, Isa 46:8); Sinai 361: בושׁ Polel “to delay” (2x, Exod 32:1; Judg 5:28) and תִּמָּהוֹן “confusion” (1x, Deut 28:28; Zech 12:4); Sinai 376: קלע “to engrave” (3x, 1 Kgs 6:29, 32, 35); Sinai 377: מלל “to inscribe, scrape” (1x, Prov 6:13)

Petrovich has also assigned a number of words lexical meanings derived from contextual use in the biblical text rather than a denotative gloss based on any clear lexicological principles. Consider the examples below:

Sinai 346a: זלל in BH as “to despise” occurs in the Hiphil, not the Qal (nor the Pual, as proposed for Sinai 349) and סוּג does not mean “an apostate” by itself (in Prov 14:14 it occurs within the phrase סוּג לֵב); Sinai 349: תזזin the Qal is not causative in BH; Sinai 351: סבב is never used with a temporal entity as the subject, i.e. “a year turns/changes” in BH; Sinai 353: חמשׁ denotes being grouped in 5 parts or by 50s and only connotes “organize for war” in the context of Exod 13:18; Josh 1:14; 4:12; Judg 7:11 and then only as a Qal passive participle; Sinai 357: מצה does not occur in the Pual (nor the Piel); Sinai 361: תמה does not occur in the Piel in BH and חֹבֵשׁ does not mean “bound servitude” in Isa 3:7; Sinai 376: קלעonly occurs in the Qal in BH, not the Pual (nor Piel); Wadi el-Ḥôl 2: the suggested form for עות “crooked one” fits no paradigm and יגה for causative “to afflict” does not occur in the Qal in BH, but only Piel or Hiphil

3. Final Comment on History and Religion

In summary, Petrovich does not make a compelling case for many of his readings. His morphological analyses wreak havoc with the historical reconstruction of West Semitic grammar. A final example of this comes in his conclusion, where he links the use of the word BʿLT and the image of Hathor in the Sinai alphabetic epigraphs with the Golden Calf episode in Exodus 32 (199-200). Petrovich’s argument is a bit convoluted but can be summarized as follows:

• the Israelite’s use of a calf did not “come out of a hat”;

• the Hebrew word for calf, עֵגֶל (note that in Exod 32:4 the word is bound, עֵ֣גֶל מַסֵּכָ֑ה), is masculine in the Bible, but the feminine word “heifer” would have had the same form in the “pure consonantalism” of the PCH inscriptions;

• since the Pentateuch had to have been written no later than 1406 B.C.E. if Moses is viewed as a historical person “and the biblical chronology is taken as literal and accurate” (199), the form in Exod 32:4 might have originally been feminine, given that the ה mater lectionis marking unbound feminine nouns “would not have been added until at least ca. 850 BC” (200);

• this means that the later Hebrew scribes would have been forced to guess whether the word handed down in the Mosaic consonantal Pentateuch was masculine or feminine; they chose masculine but it could have been a young female cow;

• “therefore, the probability is great that the golden heifer fashioned by Aaron at Mount Sinai in 1446 BC was an image of none other than Hathor, the goddess whom the Israelites who participated in mining expeditions of Serâbîṭ had been worshipping routinely” (200).

Hold on, I need to pick my jaw off the floor again. Even re-reading this argument has me in awe, and not in a good way. I’ll be mercifully brief in refuting this appalling argument.

• First, a feminine noun in the mid-second millennium would have had the final-t morpheme (it is only in the first millennium that the lost of –t begins; see Lipiński 2001: 205, 237).

• Second, the phrase in Exod 32:4,8 is עֵגֶל מַסֵּכָה, in which עגל is bound; if it were feminine, the bound form would never lose the final -t in Hebrew regardless the era!

• Third, the עגל is unambiguously masculine in 32:24, where the demonstrative follows is: ‏וַיֵּצֵ֖א הָעֵ֥גֶל הַזֶּֽה.

Finally, there is a severe dissonance for someone who adheres to a naive reading of the Bible as a historical source to suggest that the later (monarchic era!) scribes could and would alter the meaning.

4. Concluding Thoughts

As I mentioned at the outset in this post that the rhetorical style of Petrovich’s book troubled me and with my concluding thoughts I will explain why. First, there is a pugilistic stance towards contemporary scholars when their conclusions do not match Petrovich’s or serve his goals. For example, in the preface the monographs of Benjamin Sass and Gordon Hamilton are said to have “glaring weaknesses” because they refrain from offering translations of the early alphabetic texts (xii). This is an unfair characterization, since Sass’ and Hamilton’s topic was the origins of the alphabet, not the texts themselves. Moreover, one could also argue that the lack of translations is a strength of both works since it is quite likely (Petrovich’s attempts notwithstanding) that the sequences of extant letters do not present us with interpretable texts.

Petrovich’s language towards many contemporary scholars stands in significant contrast with strangely fawning references to early twentieth century scholars, such as Hubert Grimme and Alan Gardiner, whom he identifies as “one of the twentieth century’s greatest linguists of the ancient Egyptian language” and “a world-class Egyptologist” (9). Such descriptive ornamentation, either positive or negative, goes beyond the dispassionate consideration of ideas; whether intentional or not, it serves to discredit some and present others as authorities. Moreover, while we should recognize the scholarly achievement of past generations, there is a clear argument flowing beneath Petrovich’s discussion of previous scholarship, namely, that earlier scholarship that was “friendlier” to a traditional chronology of the Bible is better and more believable.

Second, Petrovich attempts to head off any criticism of his provocative work by telling his reader that he has no doubts his work will be prejudicially criticized but that his interdisciplinary training made it incumbent upon him to publish his results (xiii). (In the introduction [really a foreword], Eugene Merrill similarly warns that while Petrovich’s work will likely be objected to “because of his ideological and/or theological predilections,” his extensive research, fastidious attention to detail, and his acclaimed expertise in every relevant discipline should comfort the reader [vi]). Based on his self-described “exacting” research, Petrovich contends that judgments concerning his conclusions should notbe “determined hastily”; rather, his arguments should be allowed to age for “three, four, or five decades.” Citing Schopenhauer in his preface (xiii), he implies an equation of his thesis with “truth,” which must endure ridicule before ultimate acceptance. And he claims that his interpretations are free of “personal bias,” even though he narrates how his study of the early alphabetic epigraphs rose out of his research into biblical chronology, resulting in his forthcoming volume, New Evidence of Israelites in Egypt from Joseph to the Exodus. Identifying two letters as משׁ and then equating them with the personal name of the biblical Moses is hardly free of bias.

The inverse relationship between specialization and dilettantism is no stranger to biblical studies. As our various disciplines and sub-disciplines require increasing sophistication in theory and method, it is not easy to acquire adequate training and skills in related areas of inquiry. More than once among a group of peers have I heard a wistful (and naive) lament about how easy scholars had it a century ago, when one could “master the entire field” and work on multiple, nearly unrelated topics.The antidote, of course, is coupling time with intellectual curiosity, leading one into new areas over the course of a career. Tragedies of dilettantism may occur when either element is missing and the necessary expertise and maturity of method are absent. Tragically, Douglas Petrovich’s monograph on the early alphabetic epigraphs from Egypt and the Sinai epitomizes what may happen without adequate training or the time-tempered maturation of intellect and thesis.

The World’s Oldest Alphabetis, without a doubt, a creative, detailed, and passionate investigation of the early alphabetic texts from the Sinai and Egypt. It is a topic that requires one to interact with archaeology, epigraphy, paleography, writing as both a technology and a social convention, the grammars of ancient and biblical Hebrew and Egyptian, not to mention the known details of the history during the second millennium in Egypt and the Levant. Such a task is not for the faint hearted and Petrovich should be applauded for having great courage. With that said, Petrovich’s arguments, conclusions, and general rhetorical stance invite to serious challenge and the readers should be aware that this volume must be used very carefully.

Sources Cited

Albright, William Foxwell. 1969. The Proto-Sinaitic Inscriptions and their Decipherment. Cambridge: Harvard University Press.

Deitrich, Manfried and Oswald Loretz. 2008. “Vokalbuchstaben” im Keilalphabet von Ugarit und im griechischen Alphabet in historischer Betrachtung. Pp. 239-260 in Orbis Ugariticus: Ausgewählte Beiträge von Manfried Dietrich und Oswald Loretz zu Fest- und Gedenkschriften. Ed. M. Dietrich. AOAT 343. Münster: Ugarit-Verlag.

Friedrich, Johannes, and Wolfgang Röllig. 1999. Phönizisch-Punische Grammatik. 3 Auflage, neu bearbeitet von M.G. Amadasi Guzzo unter Mitarbeit von W. R. Mayer ed. Analecta Orientalia 55. Rome: Pontifical Biblical Institute.

Garr, W. Randall. 1985. Dialect Geography of Syria-Palestine, 1000-586 B.C.E. Philadelphia: University of Pennsylvania Press.

Gzella, Holger. 2013. Pp. 84-110 in Linguistic Studies in Phoenician Grammar: In Memory of J. Brian Peckham, ed. Robert D. Holmstedt, and Aaron Schade. Winona Lake, IN: Eisenbrauns.

Krahmalkov, Charles R. 2001. A Phoenician-Punic Grammar. Handbuch der Orientalistik, Erste Abteilung, Nahe und der Mittlere Osten 54. Leiden: Brill.

Lipiński, Edward. 2001. Semitic Languages: Outline of a Comparative Grammar. 2nd ed. Orientalia Lovaniensia Analecta 80. Leuven: Peeters Publishers.

O’Connor, Michael. 2004. The Onomastic Evidence for Bronze-Age West Semitic.Journal of the American Oriental Society 124 (3): 439–470.

Pardee, Dennis. 2003. Review of Ugaritische Grammatik, by Josef Tropper (AOAT 273; Münster, Ugarit-Verlag, 2000). Archiv für Orientforschung 50: 1-404.

Riley, Jason A. 2013. “Why, O -y”: The 1cs Suffix in Ugaritic and Its Bearing on the Case of the Vocative. Ugarit Forschungen 44: 261-284.

Wilson-Wright, Aren. 2013. Interpreting the Sinaitic Inscriptions in Context: A New Reading of Sinai 345. Hebrew Bible and Ancient Israel 2: 136-148.

First, from a religious perspective, the only reason for reconstructing some previous version of the text is if a community’s doctrine of Scripture places the authority in the “autographs”. But for those communities that consider tradition authoritative, there is really no religious motivation for finding an earlier layer, since it would deny the authority of the development. So does all this effort derive from serving the conservative Protestant community?

Second, from an academic historical perspective, producing an eclectic text boils down to an interesting intellectual exercise. But what more? Only the most arrogant would present their reconstructions are sure (and these I would run from the fastest). For the rest of us, eclectic texts present few research uses. I’ve already made the linguist’s argument. And the historian’s argument would differ only in a few particulars. And even then, the thought of a responsible historian using an eclectic text is a horror.

I think of a student’s accepted JBL article on the use of בית ישׂראל in Ezekiel. He notes that the 11 occurrences of the slightly different phrase בני ישׂראל in Ezekiel are typically understood to be the product of later redaction. Now, if this is accurate, it helps the historian reconstruct the development of ideas in the reception and transmission of the Book of Ezekiel. But if, in the pursuit of some earlier stage of the text of Ezekiel, a text critic were to deem these insertions late enough to not be part of the “the earliest inferable textual state of a book” (Hendel 2016: 50) and so omit them in the reconstructed text, because they are, in fact, later “convolutions” to the text (and especially if an eclectic text were removed from the extensive commentary), then the historian using said reconstructed text is at risk of missing important historical information.

And to anticipate the refrain that all reconstructions will be clearly marked, if text critics cannot trust the non-specialist to avoid naivete with regard to the historical sources, then they certainly can’t be trusted to note the in-text markings of reconstructions and look down or around for the critical notes. Consider the horrible ways that BHK and BHS have been used by students and scholars alike! If the users are dumb, they’re dumb regardless of the tool. We should simply have higher expectations for scholarly work. I deplore dumbing things down. I, for one, refuse to have any text critic treat me like a dumb user.

So, to bring it back around: aside from being an interesting intellectual exercise (such things can indeed be justified), what other point does a reconstructed “text edition” serve?

But the negative assessment of this post surprised me (even after it was edited to remove some of the more egregious language). Initially I considered responding to the points seriatim, but I waited and slept on it (some wisdom does come with age) and concluded that it wasn’t a wise use of my time, since the post reflects a knee-jerk reaction, not a substantive engagement with my ideas. (And only later, with a little searching, did I discover that the author had written an MA thesis on the feasibility of eclectic editions of the Hebrew Bible, so in hindsight his reaction is not surprising.) I would ask other readers to step outside their well worn paths of thought and engage with the principles of my essay.

I do, however, want to point out one element in the critical blog post that is worth considering: the attitude of superiority that I’ve seen among some text critics, that the text critic is a “special sort of person.” It is presumptuous and arrogant to imply that “advanced competence in 5-10 ancient languages” makes such a scholar any more philologically rigorous or sophisticated than one who has achieved expertise in 1, 2, or 4 languages. It is also dangerously false to assume that learning numerous languages entails methodological and theoretical sophistication. Finally, “linguistically competent” is such a fuzzy phrase, it should be avoided. The blogger’s use can only refer to being able to “access” ancient languages (at some varying level of competence). He cannot (accurately, at least) be referring to the education it requires to engage the field of linguistics and carry out informed linguistic analysis.

Given that my essay began life as one section in an article about James Barr’s legacy, I think it fitting to conclude with a few a propos thoughts from his Comparative Philology [note that I have inserted “biblical” for “Semitic” in every case below]:

Our arguments here have some effect on priorities in education for biblical scholarship. The strong influence of comparative philological method may have produced an unfortunate overemphasis on comparative study in the training of students. The prestige and fashionableness of the philological approach often cause students to study a larger number of [biblical] languages than they can master. These languages are not mastered properly and all the effort does not lead to a thorough knowledge of the texts. … To observe this, unfortunately, is not enough to put a stop to the tendency. The intellectual prestige of the philological approach is reinforced by the apparent social prestige of linguistic polymathy. It continues to be widely supposed that study of a large number of [biblical] languages is the gateway to competence in biblical studies.

… In spite of our debt to comparative philology, Hebrew does remain a teachable subject in its own right; and, while the student must now always be aware of the contributions of cognate languages, he will, unless he is ready to study these languages thoroughly, be best employed not in gaining a smattering of them but in learning how to evaluate, in relation to his Hebrew knowledge, the suggestions made on the basis of them. This means that eventually adequate modes of communication and co-operation have to be built up between two kinds of scholar: (a) those who really know the cognate languages or some of them (can any now really know them all?) and ( b) those who only assimilate this knowledge within their own grasp of Hebrew. But we can at least do something to depreciate the false prestige which has attached to the polyglot ideal, and rebuild the picture of the Hebraist. The polyglot ideal, we may remind ourselves, by no means obtains in the Indo-European field; no one supposes that to appreciate Greek literature one must study all the Indo-European languages. (1968: 295-296, 298, emphases added)

Barr is here typically acerbic. And while his point was aimed at what he saw as the undue emphasis on comparative Semitic education in Hebrew studies, I obviously think the principles apply to all language study associated with biblical studies. Knowing one thing well is preferable to knowing many things moderately. And even knowing many things well does not mark a particular scholar as someone better or “more special” than those who have chosen to focus their time and skills elsewhere, such as in the equally complex task of reconstructing ancient Israelite history, or the bewildering complex world of archaeology, with its ever increasing technological sophistication. Regardless the number of languages or non-language research tools learnt, the point should always be the methodological rigor, the theoretical sophistication, and the consistent application of logic.

Note that I do not think that this criticism applies to those whose works (and persons) I have personally encountered, such as Michael Fox, whose Proverbs HBCE volume I have unfortunately criticized (I wish other volumes had been released!). It has not been my experience that Michael, for instance, ever considered himself superior to scholars working along different intellectual paths, such as historians of ancient Israel,  BH linguists, or literary critics. He simply set out do perform the highest level of scholarship with the tools he already had or those he committed himself to add to his arsenal. May Michael’s sense of humour and humility, combined with a commitment to the best possible scholarship, be an example to us all.

Part 4: Specific Objections, Part Two (and a Conclusion)
Unfortunately, Fox’s HBCE commentary leaves unaddressed at least two significant questions that arise about how the decision to emend was reached. First, the poetic judgment against repetition is not justified (by either Fox or Hendel). Two poetic lines in Proverbs often have overlapping or even identical items; for instance, verbs identical in root if not also inflection, are used in enough cases to establish it as acceptable poetic style (see, for example, see Prov 8:5 [הָבִינוּ//הָבינוּ‎], 11:7 [תֹּאבַד//אָבְדָה‎]‎, 11:16 [תִּתְמֹךְ//יִתְמְכוּ]‎, 18:20 [תִּשְׂבַּע//יִשְׂבָּע)]). If we added examples of other parts of speech as well as the other poetic corpora, it would become clear that ancient Hebrew poets were not so averse to what modern scholars may consider to be “pointless” or “banal” repetition between poetic line pairs (on the essentially “repetitive” syntax and semantics of poetic line relations, see Holmstedt f.c).

The second question raised by Fox’s choice to emend is whether an error of dittography is really suggested by the Septuagint’s κάθῃ. A brief survey of the Greek evidence in Proverbs will illustrate the issue. Throughout the Septuagint, the Hebrew שׁכב is overwhelmingly rendered by Greek κοιμάομαι (middle of κοιμάω) ‘to fall asleep, go or lie abed’. But this Greek verb is used only once in Proverbs, in 4:16, where the MT does not have שׁכב but כשׁל.FN1 Thus, there is already a departure by the Greek Proverbs translator from patterns established outside Proverbs. Moreover, of the eight times that שׁכב is used in Proverbs, the Septuagint translates with an interesting variety of terms: κάθῃ (> κάθημαι ‘to sit, sit down, sit quiet, lie’) in 3:24a for תִּשְׁכַּ֥ב; καθεύδῃς (> καθεύδω ‘to lie down to sleep’) in 3:24b for וְשָׁכַבְתָּ֗; κατάκεισαι (> κατάκειμαι ‘to lie down’) in 6:9 for תִּשְׁכָּ֑ב; καθεύδῃς (> καθεύδω ‘to lie down to sleep’) in 6:22 for the infinitive in ‏בְּֽשָׁכְבְּךָ; κατακείσῃ (> κατάκειμαι ‘to lie down’) in 23:34a for שֹׁכֵ֣ב; and for the forms of שׁכב in 6:10, 23:34b, and 24:33, the Septuagint offers no gloss and appears to render the text quite differently.2 Concerning the proposed change to ישׁב, the Greek translator is more consistent: in all but one of its occurrences κάθημαι renders ישׁב (with the sole exception outside Proverbs: the use of a form of κάθημαι to translate the participle רֹכֵ֨ב in Isa 19:1).

To what end does this brief study bring us? Fox himself characterizes the translation technique of the Greek Proverbs translator as “flexible” and he favorably quotes Peter Gentry’s assessment that “The problem of the relationship between LXX and MT Proverbs is notorious and vexing” (2015: 36). Fox is certain that the Septuagint Proverbs reflects a different edition of Proverbs than the Masoretic edition and avers that, “[s]ometimes [the Septuagint translator] maps his source closely, sometimes paraphrases, sometimes expands the quantitative representation of Hebrew words, sometimes reduces it, and sometimes just guesses at meaning” (40). He goes on to say that the lack of an isomorphic translation at one place simply illustrates “the kinds of things the translator can do.” In light of both the general character of the Septuagint Provers, as described by Fox, and the variety of glosses given to Hebrew שׁכב in Proverbs, it becomes clear that the case of תשׁכב in 3:24 is not nearly as simple as Fox (and Hendel) suggest. On the one hand, the poetically acceptable use of repetition of verbal roots in associated poetic lines and, and other hand, the reasonable possibility that the Greek translator paraphrased the תשׁכב (perhaps due to the same impulse that motivated Fox’s emendation, to see a sequence of actions) or the translator used a Hebrew text that had תשׁב (which also may have reflected an inner Hebrew change from תשׁכב for the impulse already mention) together suggest that that the Septuagintal evidence does not strongly support emendation.

Conclusion
The point of these two examples is that decisions to emend must reflect a deep sophistication that is often beyond the time and abilities of a single scholar (as I have admitted for myself in handling the Septuagint). For 5:22, any statement about the “syntactic integration” (or lack thereof) must be grounded in both a thorough knowledge of Hebrew grammar and an awareness of how the Versions, especially the non-Semitic Septuagint, deals with acceptable but uncommon Hebrew syntax. Fox’s own argument for the similar syntax of Prov 13:4 and the omission by the Greek of נפשׁו provides a reasonable argument for keeping the את הרשׁע of MT 5:22a. For Prov 3:24, we have seen that the Greek does not present a rigid profile for rendering the Hebrew שׁכב and the juxtaposition of poetic lines with the same root or form is not uncommon in Hebrew poetic style. Thus, two other options are linguistically just as likely as Fox’s choice to emend: the Greek translator may have rendered תִּשְׁכַּ֥ב with a minority but semantically acceptable choice, or a copyist wrote תשׁב for תשׁכב by accidentally skipping the כ (haplography).

This discussion is in no way meant to detract from the scholarship of Michael Fox, who was one of my mentors. Rather, the problems I have noted in the first volume of HBCE, produced by a scholar of high repute, who has spent an entire career working on Proverbs, suggest that if such a text-critical project is even feasible the volumes cannot be accomplished by individuals. In the absence of the kind of extraordinary skill set rarely found among scholars in our age of compartmentalization, a highly reconstructive project requires a text-critical and linguistic partnership for each volume, with both kinds of scholars in continuous conversation as they reconstruct the philological text and linguistic text.

Scholarly partnerships aside, in light of a reconstructed text’s (lack of) linguistic value, a text critical project like the HBCE should always and only be considered a commentary series, never a text edition. Such an eclectic text (or better, in Fox’s words, a “construct”) should never be used as a classroom text or for exegetical scholarship. It has no historical reality and, as such, has no direct value for historical or linguistic research. Unfortunately, since Fox’s Proverbs text is printed a second time at the back of the volume—apart from the commentary— it appears that eventually creating a separate, stand-alone volume of the reconstructed texts is precisely what is intended. But hiving off the text part of the commentary for an eclectic text edition would undermine Hendel’s statement that the “decisions and analyses will then be available for discussion, refinement, and refutation—the normal process of scholarship” (2016:16). Divorcing the final product apart from the commentary would not only gut the reconstructed text of any pedagogical value, it would cast a darkly elitist shadow over Hendel’s argument that the HBCE project will protecting those “who may be innocent of the discipline of textual criticism,” that is, “those least qualified, ” from making “important text-critical judgments” (ibid).

Notes:
1. Fox (2015:111) suggests that either “a copyist duplicated ישׁנו from 4:16a” which the Septuagint translator rendered differently in 16b for the sake of variety (16a renders ישׁנו by ὑπνώσωσιν]), or a scribe accidentally flipped the שׁכ to כשׁ (metathesis), which the Greek translator then, presumably, read as שׁכב (how the ל of כשׁל was read as a ב in שׁכב is not explained).

2. Strangely, Fox does not discuss the Septuagint’s omission of לִשְׁכָּֽב in 6:10 (130-131) or 24:33 (330-31). And he describes the lack of a gloss for כְשֹׁכֵ֗ב in 23:34b as “semantically superfluous” (319).

Works Cited

Brooke, George J.
2013 The Qumran Scrolls and the Demise of the Distinction between Higher and Lower Criticism. Pp. 1-17 in Reading the Dead Sea Scrolls: Essays in Method, EJL 39; Atlanta, GA: Society of Biblical Literature.

Brotzman, Ellis R. and Eric J. Tully
2016 Old Testament Textual Criticism: A Practical Introduction. 2nd ed. Grand Rapids, MI: Baker Academic.

Fox, Michael V.
2000 Proverbs 1-9: A New Translations with Introduction and Commentary. AB 18A. New York: Doubleday.
2009 Proverbs 10-31: A New Translations with Introduction and Commentary. AB 18B. New Haven, CT: Yale University Press.
2015 Proverbs: An Eclectic Edition with Introduction and Textual Commentary. HBCE 1. Atlanta, GA: SBL Press.

Friedman, Matti
2012 The Aleppo Codex: A True Story of Obsession, Faith, and the Pursuit of an Ancient Bible. Chapel Hill, NC: Algonquin Books.

Goshen-Gottstein, Moshe H.
1979 The Aleppo Codex and the Rise of the Massoretic Bible Text. BA 42 (3): 145–63.

Hendel, Ronald S.
1998 The Text of Genesis 1-11: Textual Studies and Critical Edition. Oxford: Oxford University Press.
2008 The Oxford Hebrew Bible: Prologue to a New Critical Edition. VT 58: 324-351.
2016 Steps to a New Edition of the Hebrew Bible. Text Critical Studies 10. Atlanta, GA: SBL Press.

Holmstedt, Robert D.
2013 The Nexus between Text Criticism and Linguistics: A Case Study from Leviticus. JBL 132 (3): 473-94.
f.c. Biblical Hebrew: and the Appositive Style: ‘Parallelism’, requiescat in pace. To appear in Vetus Testamentum

Holmstedt, Robert D. and Andrew R. Jones
2017 Apposition in Biblical Hebrew—Its Structure and Function.” KUSATU 22: 21-51

Tov, Emanuel
2000 The Textual Basis of Modern Translations of the Hebrew Bible: The Argument against Eclecticism. Textus 20: 193 – 211. (Revised version, Pp. 92-106 in E. Tov, Hebrew Bible, Greek Bible, and Qumran: Collected Essays; Tübingen: Mohr Siebeck, 2008)
2006 Hebrew Scripture Editions: Philosophy and Praxis. Pp. 281-312 in From 4QMMT to Resurrection: Mélanges qumraniens en hommage à Émile Puech; ed. F. García Martínez et al.; STDJ 61; Leiden: Brill. (Revised version, Pp. 247-70 E. Tov, Hebrew Bible, Greek Bible, and Qumran: Collected Essays; Tübingen: Mohr Siebeck, 2008)
2011 Textual Criticism of the Hebrew Bible. 3rd ed., revised and expanded. Minneapolis, MN: Fortress.
2014 New Editions of the Hebrew Scriptures: A Response. HeBAI 3: 375 – 383.

Williamson, Hugh G.M.
2009 Do We Need A New Bible? Reflections on the Proposed Oxford Hebrew Bible. Biblica 90: 153-175.

Part 3: Specific Objections, Part A
Beyond the principled objection to it given above, a more practical objection against the project is the lack of any theoretically-oriented linguists involved. The enterprise of fully reconstructing a text (beyond “simple” scribal errors, however we define these) certainly requires a deep knowledge of the available artifactual evidence and the plausible histories of transmission. Yet it is also a fundamentally linguistic endeavor and therefore requires a high sensitivity to the likely linguistic changes that may lie behind textual changes that are more than merely slips of the pen.1 Two brief examples out of Fox’s HBCE volume will suffice to illustrate the linguist’s concerns about decisions made by non-linguists.

The first example is one that Fox highlights in the introduction to his HBCE volume: the status of את הרשׁע in the MT of Prov 5:22, provided in (1).

(1) Prov 5:22: עַֽווֹנוֹתָ֗יו יִלְכְּדֻנ֥וֹ אֶת־הָרָשָׁ֑ע וּבְחַבְלֵ֥י חַ֝טָּאת֗וֹ יִתָּמֵֽךְ׃

Fox notes that the phrase את הרשׁע in the first half of 5:22 is not represented in the Septuagint (G) or Peshitta (S) and is “not integrated into the Hebrew syntax” (5). He considers the phrase therefore to be “an epexegetical gloss clarifying the object of ילכדנו” and as such “is not really necessary” (123; see also Fox 2009: 204-5). In response, the linguist would note that, strictly speaking, many types of modification in language are “not really necessary.” But the desire for perspicuity in the use of language for the communication of ideas leads to a great deal of “unnecessary” clarification. The goal of clarification lies behind the use of any nonrestrictive relative clause or appositive, both of which are abundantly attested in the Bible. Indeed, apposition is the syntax behind the “synonymous parallelism” that lies at the heart of Hebrew poetic style: taking the idea of one line and reformulating in a second line in order to clarify the desired proposition or image (see Holmstedt f.c.).

Therefore, while Fox is accurate in identifying את הרשׁע as a phrase used to clarify the object attached to the verb in ילכדנו, to say that it is not syntactically integrated is simply mistaken. What occurs in Prov 5:22a is what is called in traditional grammatical descriptions “prolepsis” of the object (see, e.g., Rendsburg 1990: 125-32; Joüon and Muraoka 2006: §146e) and what would be analyzed in linguistic terms as either as apposition with a pronominal anchor or right-dislocation (see Holmstedt 2014 on right-dislocation; see Holmstedt and Jones 2017 on apposition).2 To bypass the technical linguistic details, the basic communicative result for either analysis is to clarify and/or highlight the precise referent (הרשׁע) of the anchor (the 3ms pronoun attached to the verb in ילכדנו). Interestingly, Fox notes this very syntax, though with the pronoun attached to a noun, in Prov 13:4a:

(2) מִתְאַוָּ֣ה וָ֭אַיִן נַפְשׁ֣וֹ עָצֵ֑ל
‘craving (but nothing!) is his appetite, the sluggard’ (Prov. 13:4a)

In his Anchor Bible commentary on Proverbs 10-31, Fox notes the syntax of the “anticipatory suffix” as support against emending נפשׁו to נפשׁ, i.e., omitting the pronoun (2009: 562). The obvious question, apart from any Septuagint evidence is, Why is the syntax acceptable in 13:4a and not in 5:22a? And perhaps the lack of את הרשׁע in 5:22 of the Septuagint should receive a similar explanation as Fox gives for the lack of נפשׁו in 13:4 in the Septuagint, Vulgate, Peshitta, and Symmachus: “Since there is no good explanation for the loss of this word, it was probably present in their source texts (contrary to BHQ) but considered as adequately implied by the notion of desiring. The difficulty of the syntax may have motivated this approach” (207).
The second example I will discuss concerns Prov 3:24, given in (3):

(3) Prov 3:24
MT (L): אִם־תִּשְׁכַּ֥ב לֹֽא־תִפְחָ֑ד וְ֝שָׁכַבְתָּ֗ וְֽעָרְבָ֥ה שְׁנָתֶֽךָ׃
LXX (G): ἐὰν γὰρ κάθῃ, ἄφοβος ἔσῃ, ἐὰν δὲ καθεύδῃς, ἡδέως ὑπνώσεις·

The issue seems straightforward: Fox emends תִּשְׁכַּ֥ב to תֵּשֵׁ֥ב, based on the Septuagint’s κάθῃ and the Syro-Hexapla’s ܬܬܒ. Fox argues that the MT’s תשׁכב resulted from the scribal error of “near dittog[raphy] ב → ‎כב‎” (103; also Fox 2000: 162-63). To support this change, he argues that the use of ישׁב in the first half “fits into a sequence of actions that represent the totality of a day’s activities: walking (3:23), sitting down (3:24a), going to sleep (3:24b).” Fox also asserts that the MT’s use of שׁכב “in both stichoi is pointlessly repetitious.” This example is highlighted by Hendel, who suggests that the supposed dittography is “motivated by the scribe’s anticipation of the verb in the second half of the verse” (2016: 156); Hendel also suggests that the result in the MT yields a “banal parallelism.”

In the fourth and final post, I will wrap up my criticism of eclecticism.

Notes:
1. Holmstedt 2013 for an illustration of this with regard to a הוּא‎ and הִיא‎ variation in the witnesses to Lev 1:17 and 25:33
2. Prov 5:22 is listed among the examples of the “anticipatory pronominal suffix” in Rendsburg 1990: 125-32.

Part 2: General Objections to Eclecticism

If Hendel effectively counters the objections to an eclectic Hebrew text, why would (or should) anyone continue to oppose it? The Hebrew linguist must dwell on two nagging problems. First, a reconstructed text is not a historical artifact.1 Fox is admirably candid about this:

I wish to be clear that the text I have produced, however successful, never had physical existence. It is a construct. It can be defined as the proto-M as it should have been, the text the authors and editors wanted us to read. (2015: 5, italics in original)

And yet, Fox’s clear expertise in Hebrew and the book of Proverbs notwithstanding, his Book of Proverbs “construct” does not constitute a primary historical or linguistic source. It is a modern text with a modern author, Michael Fox. But evidence localized in a historical artifact—primary source data—is precisely what a linguist depends on. The judgment of certain linguistic items in the historical artifacts to be ill-formed, due to some vagary of the scribal process, should certainly indicated in critical editions of texts; but for the linguist studying the language data, whether for synchronic or diachronic purposes, reconstructed items are unusable. Furthermore, reconstructions placed in the text actually obscure the historical data and, due to the need to identify and set them aside, become time-consuming obstacles to linguistic analysis.2

In the third post, I will continue with specific objections.

Notes:

1. Tov 2014: 378, n. 13; also note Brooke 2013: 13 in reference to eclectic editions of the New Testament.

2. It is also worth noting that, Hendel’s many comments notwithstanding, in textual criticism outside biblical studies, the pendulum is swinging (or has swung) away from producing eclectic critical editions. Indeed, in the recent words of one of my non-biblical studies colleagues, the notion of producing an eclectic text is “barbaric.”

I suspect that most readers will consider the former topic to make some sense for me to address, while the latter topic makes very little. I have never claimed to be a text critic. And yet, I do continue to teach and carry out research on the Hebrew Bible, so it has been an issue bouncing around the hollows of my head for some time. Moreover, I have had to face the issues of diplomatic-vs-eclectic text more directly with my research into Ge’ez and the Abba Garima Gospels. So it was a good opportunity to sort my thoughts out conquering the Hebrew Bible: A Critical Edition (HBCE) project.

In this and 3 subsequent posts (2, 3, 4), I will present my case against eclecticism from the perspective of a linguist.

Part 1: Background

In contrast to almost all other subfields in biblical studies—whether concerning the New Testament, the Septuagint, the Peshitta, or even the Ethiopic Bible—the modern study of the Hebrew text of the Old Testament has begun with scholarly editions that are diplomatic in nature. That is, the text presented is that of a single historical witness with any variants or critical notes placed in marginal apparatuses. Previous to 1937 and the publication of the third edition of Rudolf Kittel’s Biblica Hebraica (BH3), printed editions of the Hebrew Bible used the second Rabbinic Bible, which had been the textus receptus almost since Daniel Bomberg printed it between 1524 and 1525 (Tov 2011: 70-73, 341-46; Brotzman and Tully 2016: 129-47). In the third edition (BH3), fourth edition (BH4, 1983), also known as Biblica Hebraica Stuttgartensia (BHS), and the most recent edition, Biblica Hebraica Quinta (BHQ, 2004-), which is still in-progress, the text presented was that of the Leningrad Codex (AD 1008). The Hebrew University Bible Project (HUBP, 1955-) is the other major text edition and has been in progress since 1955, with only three volumes so far published (Isaiah, Jeremiah, and Ezekiel); the HUBP uses the once complete but now damaged Aleppo Codex (AD 925) as its base text.1
Though such an endeavor is standard in many other biblical studies subfields, an eclectic text of the Hebrew Bible—an edition in which the text is a reconstruction aimed at presenting some earlier historical stage of the text—is a concept that has never gained serious traction. As Williamson summarizes,

In these cases [of the Greek New Testament, Septuagint, etc.] it has long been standard practice for the editor to gather all the evidence available to him or her, such as different manuscripts, citations in other works and so on, and then to produce what he or she regards as the probable original form of the text—a process which may well also include some conjectural emendation of passages which are deemed to be corrupt but for which no reading has survived that seems to give a satisfactory solution. The apparatus in such an edition documents the evidence from all the available sources while the printed text does not represent any one of those sources in its entirety. What is more, in the case of classical texts, it is far from unknown for the editor to incorporate decisions about later editorial activity and so to omit sections which are deemed not to derive from the original author. The result is known as an eclectic text, whereas in the case of the standard Hebrew Bible editions it is known as a diplomatic text. (2009: 157)

Though there have been previous experiments with an eclectic Hebrew text (Tov 2006: 291), the seeds of the latest push for an eclectic Hebrew text were sown in Hendel 1998, which presents a reconstructed text for Genesis 1-11 and seems to have been an early proof of concept for the Oxford Hebrew Bible (Hendel 2008), later renamed the Hebrew Bible: A Critical Edition (HBCE). In this new eclectic edition, the first volume, on Proverbs, has been published (Fox 2015) as has a thorough description and apology for the project (Hendel 2016). Hendel’s apologia is particularly necessary in the face of incisive criticisms the project has faced (see, e.g., Tov 2000, 2006, 2011, 2014, Williamson 2009, Brooke 2013). Hendel summarizes the raison d’être of an eclectic edition succinctly: “if an eclectic edition is done well, it approximates a particular manuscript, the archetype, though it also reaches behind the archetype when it detects and corrects its scribal errors. An eclectic edition aims at the earliest inferable textual state of a book, which is an empirical and justifiable goal” (2016: 50).

There are numerous principles or details of practice concerning the actual HBCE project, and they are critiqued ably by Tov, Williamson, Brooke, among others. Examples of these run from large questions about what the aimed-at “archetype” actually is and about whether any set of analytical criteria can raise such a project above the charges of subjectivity to arguably smaller (and more easily adjusted) questions about the project’s use of the Leningrad Codex as its base text and the employment of features specific to the Masoretic tradition, such as vowel pointing and cantillation accents, when the goal of the “earliest inferable textual state” of any biblical book presumably predates the Masoretic features by centuries (Williamson 2009: 164). In his prolix apologia,2 Hendel (2016) directly addresses each objection and anticipates others, effectively clearing the way for an eclectic text project. Moreover, the two strongest points in favor of Hendel’s position have, in my opinion, largely avoided challenge. First, as Hendel rightly notes, “from a historical perspective it is more correct to regard the manuscripts as eclectic and the critical text as an attempt to reverse the eclectic agglomeration of primary and secondary readings” (Hendel 1998: 115). This certainly seems to be the case for the Leningrad Codex, which Goshen-Gottstein calls “a none-too-successful effort to adapt a manuscript of a different Tiberian subgroup to a Ben Asher Codex” (1979: 150).

The second point in favor of an eclectic edition is that in practice eclecticism already dominates the field. Almost all modern critical commentaries reconstruct some text behind the Leningrad Codex and present a translation based on that reconstruction (cf. Williamson 2009: 158, n. 10); similarly, as Tov (2000) notes, most modern translations reflect an eclectic approach to the Hebrew text. Thus, eclecticism, whether recognized or not, is the widespread modus operandi in Old Testament studies (Hendel 2016: 20).

In the next post, I will turn to my deep concerns about the project.

Notes:
1. According to most accounts, the Aleppo Codex was damaged in a fire due to anti-Jewish riots over the creation of the State of Israel in 1947. However, Friedman has recently questioned the fire story, asserting that the evidence supports a mostly complete codex leaving Aleppo (Friedman 2012). Regardless when and how the current form of the codex was established, the result was the loss of most of the beginning and ending of the codex, including the Pentateuch, small portions of 2 Kings, Jeremiah, Minor Prophets, Chronicles, and Psalms, and most of Song of Songs, Lamentations, Esther, Daniel, Ezra, and Nehemiah (Goshen-Gottstein 1979: 149).

2. I confess to seeing little relevance in Hendel devoting an entire chapter (chapter 10) to Frank Moore Cross or to the mostly pretentious essay (chapter 12) on the “untimeliness” of philology.