Belléli vs. Hesseling

I was curious when I opened up Dirk’s edition, and closed it with “a certain feeling of being cheated”.

Lazare has written a rather extensive discussion, including (as he is quite right in saying) “a small chapter in the history of Biblical exegesis”. I trust it was only a concern for brevity that led him to enumerate only the bad stuff about my book—and to bypass not only my bibliographical research and my studies on the vocab, but also my conclusions on the translation’s language.

If you’re going to publish this kind of text, you need to know both the Hebrew of the original, and the Greek of the translation. Not just Greek in theory, but as it’s actually spoken nowadays.

I’m not going to abuse the journal’s hospitality to defend myself against the omissions (?) of my good stuff in Lazare’s review. I admit he’s right about my Hebrew sucking. I said as much in the book. Does that mean I should not have even tried to put out the edition? Maybe, but this book was too important to wait for the improbable arrival of some Hellenist who knows Hebrew backwards. And you know what Lazare? From what I’ve seen of your transcriptions in the Revue des Études grecs, you’re not that Hellenist: you are clueless about Early Modern Greek. Chatzidakis has already sonned you on your knowledge of linguistics in Athena 3:625ff anyway.

As the following comments make clear, Belléli was adamant that the 1547 Torah should be in the same language as spoken in 1897 Athens (and Corfu), and did not allow for the fact that Greek, even highly vernacular Greek, had changed in the past five centuries.

I have been “placed in unfavourable circumstances in Corfu, far from libraries and great centres of studies.” My Dutch colleague has been luckier.

Lazare has spent years on the text, but he could not get a publisher. Shortly after he got to Paris, I heard it on the grapevine that circumstances would not allow him to complete the work—and he pretty much admits as much. I was already under contract, and I figured my work (“conceived with laudable intentions”—that’d be just about the only good thing Lazare had to say about it) was too useful to Greek Studies to keep out of print. Even if my Hebrew does suck. And printing this thing was a bitch (my printer was low on Greek characters), which is why my intro has several last minute fixes.

Having an idiot printer does not compare in the hardship stakes with a pogrom. And running out of Greek characters? The thing did get published fascicle by fascicle, and Dirk admits that he changed his mind on how to transcribe as the fascicles came out.

There’s several bits of the text that don’t make sense unless you know Hebrew. At the least, you’d expect that the Hebrew names would be rendered faithfully. But Dirk has voiced the sch’vas of Μααλαλεελ, Μιεταβααλ (?), Νιμεροδ, Σινεαβ, Θιδεαλ, Μαμερε. He mutes the second segol in two-segol words, giving Ιεφθ, Πελγ, Περζ. He comes up with Ισασκαρ, λεβιμ, and (straight out of the Septuagint) Νεφθαλι. He always renders kaf as k, giving Χανοκ, Ερκ, Βεκρ. How can we trust his rendering of Greek when he gets the Hebrew script for Hebrew names wrong?

It’s defensible if Dirk was doing a strict transcription; but Dirk doesn’t leap to defend himself here.

A lot of what looks odd in the Greek of the translation is due to the transcription in Hebrew script. There was a “tacit convention” allowing the transcription to be just approximate, so it shouldn’t be taken too seriously, at the expense of positing impossible Greek.

Despite what Lazare says, I still believe the translator knew the Greek alphabet, even if he’d never heard of ancient grammar and literature. His transcription is not as phonetic as it would be if he’d never heard of the alphabet; why else would someone write [timboli] as /tin poli/?

That’s the problem with writing about linguistics in 1897: people still haven’t worked out what phonology is. Lazare is right here (especially as he tries to carry the analogy to other languages), and Dirk is wrong: people have an intuitive understanding of phonology, and they don’t need to be trained in a particular alphabet to suppress allophones in their writing. A Romaniote who’d never seen την πόλιν could still write טן פלי <tin poli> in Hebrew.

As everyone knows, Sephardim don’t distinguish tav with and without a dagesh (ת /θ/, תּ /t/). The Romaniotes traditionally did distinguish the two, but the distinction is dying out because of Sephardim taking over the teaching of Hebrew: the only Romaniotes preserving it in Corfu are recent arrivals from Epirus, and even their children are dropping it. The Torah was published by Sephardim, so we would expect them to confuse ת and תּ on occasion in the Judaeo-Greek.

So Romaniote Hebrew, like Yemenite, preserved pronunciation distinctions with the dagesh that Ashkenaz and Sepharad Jews didn’t. Ashkenazi ת is /s/.

So it’s absurd for Dirk to think (p. xl, xlvi) that θ in the Judaeo-Greek μουσθάκι, εδαπανεύθην proves familiarity with learnèd Greek. I think he would welcome the explanation that θ is confused with τ: not only does it explain the erroneous Φιλισθιμ, Ναφθαλι, but also impossible Greek forms like θέλειος, θελειώνω, ξεθελειώνω, δεκαθέσσερεις, είκοσι θέσσαρες; and τάφειο should be read for θάφειου.

Yup, those thetas are looking pretty indefensible.

Oh, and defending the intermittent reading χαθώς for καθώς (כּ /k/, כ /x/) is embarrassing.

μετά did remain as μετ’ before pronouns in particular, and in fact was extended to other prepositions by analogy: γιατ’ εμένα; αντιτά in the Torah. This doesn’t prove familiarity with learnèd Greek either: in fact the Torah line-breaks as both μετά σεν and με τασέν, so they clearly don’t recognise μετά as the original form of the preposition in /metasen/. And μετά is never used before nouns.

Lazar’s right again: nothing necessarily learnèd about a pronominal μετά σεν.

Other syllable breaks of compound verbs are on ancient prepositions, e.g. προς-φορά; but that doesn’t prove much. Any unlettered peasant in Germany would do the same kind of syllable break. And if the translator had any exposure to learnèd [Christian] Greek, he’d have followed the Septuagint expressions that had become commonplace: he wouldn’t have come up with δαρμοί for πληγαί του Φαραώ, σκήφτρα for φυλαί του Ισραήλ, σύτροφο for αγάπα τον πλησίον σου, πουλιά for χερουβίμ.

Was the translation intended for liturgical or educational use? We can’t be sure, but the sabbath pericopes at least are read in Ottoman and Greek synagogues in the vernacular, in the midday or later services at least: Ruth, pirke abot, Lamentations, allegorical commentary on Song of Songs (the Jews fleeing Arta from Corfu in the latest war recited it in synagogue), and certain haftarot: Jonah in particular is known to us in Judaeo-Greek from several manuscripts. And these were all chanted; but the edition of the Torah has virtually no cantillation marks, other the marking the ends of verses. (So the translation was not intended for liturgical use.) Dirk is clueless about all of this, as his punctuation shows, disfiguring the meaning of passages.

Had Dirk known Hebrew better, he could have also dealt with the bad typography of the Torah, such as confusions of ב /b/ and כ /x/ or ד /d/ and ר /r/. So Num 30 has εμπόδισεν not εχώρησεν. (Btw, the Hebrew has /ebodisen/, so I’d rather transcribe it as ε(μ)πόδισεν.) Dirk faulted me doing this kind of phonetic transcription: but by writing φε(γ)γίτες, I was clearly intending the reading /feɡites/, not /feʝites/.

That’s the problem with trying to stick to historical orthography while doing phonetic transcription. Greek does not distinguish with μπ between [b], [mb], and [mp]; they’re not phonologically distinct ([b] ~ [mb] is an isogloss dialectally, and a sociolinguistic distinction in the standard language); though words may be borrowed with [mb] or [mp], spoken use tends to merge them. Cretan Renaissance poetry, written in Latin script, distinguished between <mp> and <mb>; I’ve seen an edition in Greek script use a diacritic on the pi to differentiate them. I will say though, Lazare’s transcription is awkward.

Dirk confuses clitic pronouns with the homonymous articles (του, της, τους): more care in “decomposing the Hebrew words into their constituent parts” would have addressed this.

I don’t see how it would. They’re both clitics, I can’t see how the Hebrew would tell you which is the proclitic and which the enclitic.

Dirk makes up a verb πειάτε Gen 20:13; this is merely πε γιατέ εμέν.

More clearly, πε για-τ’ εμέν.

In “a complete lack of attention by the editor”, Dirk confuses the tribe of “Dan” with δεν το, δεν την in Gen 49:46-47. And the misreading makes no sense.

στώφλια Gen 6:46? What’s that supposed to mean? It’s just στο πλάγι της.

σκατηλάχω Num 11:23? That’s σε καταλάχη.

I admitted that I could make no sense of this reading in the Wroclaw edition. Whether by guessing or a clearer copy, Lazar comes up with σε καταλάχη. Better, though there is the slight problem that there has never been such a verb as καταλαγχάνω in Greek. I’m pretty confident about emending that to σε καταλάβη, swapping כ /x/ with ב /v/.

There are too instances of καταλαγχάνω, including one in Digenes Acrites Escorial 1407. The verse reads (in Modern understanding of the text, NIV) “You will now see whether or not what I say will come true for you.” The Kriaras glosses for καταλαγχάνω are, “meet someone”, “happen to be somewhere”, and… “realise, find out” (διαπιστώνω). Which matches “see whether something comes true.” And the passage demonstrating this sense? Judaeo-Greek Elegies 164 εγώ είδα κ’ εκατάλαχα τους ξένους πώς τους κλαίνε “I have seen and realised how they lament for strangers”. (Papageorgios, S. 1901. Εβραιο-ελληνικαί ελεγείαι. Παρνασσός 5: 157-174.) Lazare’s right again. We don’t know if it was a lucky guess, or whether this was a peculiarly Judaeo-Greek sense of the word that Lazare knew (and Dirk didn’t).

A verb πικέρνει Gen 40:21? The noun πικέρνης is already used all over the chapter.

Pretty indefensible, that.

Because of the similarity of פּ /p/ and פ /f/ (again, with and without dagesh), επήγεν and έφυγαν, and πάη and φάη, are often mixed up.

βογγίζει “moans” should be βογίζει = βοΐζει < βοή “shouts”.

The editor should be making the text readable. There are a lot of Hebraicisms in the text, but the majority of the target audience for the edition doesn’t know Hebrew. Dirk should have highlighted the Hebraicisms in each chapter, and listed them. “Without some device of that sort, the text often presents a mass of unintelligible words, like minerals in a mine shaft which have not yet had the attention of a metallurgist.”

True, but I’m sympathetic to Dirk here: that would have been too long a wait.

Dirk does not sift through the Hebraicisms, and dismisses as Hebraicisms phenomena which can be explained within Greek. He thinks Gen 50:47 συ(μ)πάθησε το φταίσιμο σκλάβους θεού του πατρός σου has σκλάβους in the accusative, and this somehow reflects the construct genitive. But only the nominative is used for the construct genitive. Anyway, the genitive has been dying out in Modern Greek.

The accusative in Gen 50:47 could be under the influence of the verb, so it looks like an object. But the text is full of genitive plurals anyway; the singular genitive is alive and well in the Modern language, with only the plural genitive retreating; and the translator had no problem with chains of genitive like τα λόγια του Εσαυ του υιού της του μεγάλου Gen 27:42. Nope, these are Hebraicisms.

Greek dialect does do this kind of thing (in fact I’ve corresponded about this kind of thing with a doctoral student with regard to the Walad text). But without a definite article, it’s harder for me to read it as genitive.

Apposition does not prove Hebraicism either. He emends Ex 6:13 τον Φαραώ βασιλιάς της Αίγυφτος (as I read the London copy) to βασιλιάς, and the nominative is a modernism, but the Hebrew doesn’t differentiate between them (eth- is not regular and does not affect the form of the noun).

There’s dozens of these appositions, and they’re Hebraicisms too. From Lazare’s illustration of the apposition (Τον είδες το Τζώρτζη; “Did you see George [accausative]?” Ποιος Τζώρτζης “Which George? [nominative]”), I don’t think he quite gets what an apposition is. Actual appositions have to agree in case by definition.

I doubt this is true by definition; but given how Greek works, I think Dirk’s right.

The translation is literal enough to preserve the distinction between the Kal and Hiphil conjugations. That’s why the future is να + subjunctive. (I’ve changed my mind since my 1890 Revue des Études grecques paper.) As Psichari found, all Greek futures are based on θέλω; the translator has elided out the θέ in θε να and kept the να as closer to the verb, and getting rid of the “verbal debris” of θε. Dirk has cluelessly attributed to the tense a recent interpretation of the tempus imperfectum, when the original sense of the tense was merely future.

Right. Like someone who, like Lazar says, “has never picked up a book in Greek”, can tell what verbal debris is.

As Dirk had already pointed out in his book, να-futures were the original Byzantine form: they were the norm until 1400, and I’m not surprised to see them in this text. (That doesn’t necessarily prove the text is substantially older than 1547. But I’m not going to the garage to check Panayiotis Pappas’ work on the future particle, it’s already 1 AM on a schoolnight.)

Dirk should also have checked with a Hebraicist about του ειπεί, του ερτεί. Indirect objects are preceded by le “to”; the translator is merely rendering le in the same way when it precedes an infinitive (given that the genitive was used for indirect objects in Greek). If there is another particle instead of le in the original, you get forms like από του ειπεί, preserving the dative/genitive article. (Dirk said prepositions always take the accusative, but ignores that in his own transcriptions. But this του is a dative, not a genitive.)

This is a known archaism in Byzantine Greek (where the genitive really is genitive), but the translationism makes sense as an explanation.

Not convinced by the examples of accusative indirect objects: they are pronouns without a distinct genitive, such as ανάγγειλες εμέν, επήρα αυτήν εμέν για γεναίκα. If the genitive εμενός wasn’t a one-off in the text, you’d see it used in this context.

The indirect object test works for με vs. μου, not εμέν(α). If that’s all the examples Dirk can find, then Lazare’s right, the text is clearly genitive.

The Hebrew cannot render palatalisation by inserting /i/. The translator does write κεφαλιτίκια with two /k/s, “but he was not late in perceiving the inexactitude of this transliteration. This concern tormented him for a long time”, so you’ll see transcriptions like απλίσεψε for απλίκεψε, παραζειλιά for παραγγειλιά. That makes things even worse. So he goes back to a [phonemic] transcription, which is exactly the “tacit understanding” I mentioned before. But Dirk fails to explain several forms in that way (i.e. with <k> and <x> standing for both /k, x/ and /kj, xj/ = [c, ç]), leading to δικό for δίκιο, φτωχά for φτώχια, κουφό for κούφιο, βρακόνι for βραχιόνι, πλάκα for πλακί. Ex 15:27 should not be classical φοίνικες, but Modern φοινικιές. Also παρα(γ)γειλιά, μεριά, not παραγγειλά, μερά; the [j] is not always distinctive in pronunciation.

Maybe I was wrong in places; but better to overtranscribe than to undertranscribe, and lose linguistically interesting phenomena. I’ll concede I was wrong about /t/ for /θ/ being linguistically interesting; but Lazare would level: χλόγιη, χλογίση to χλόη, χλοΐση Gen 1:11, ορανού to ουρανού Gen 1:17, πετάει to πετά Gen 1:20, να ιδή to να δγή Gen 2:19, συγκολληθή to συκολληθή Gen 2:24, είπεν to είπε Gen 3:1, εμέν to εμέ Gen 3:12, έγνεψεν to έγνεψε Gen 4:5, &c. The whole point of our linguistic interest is in the phonetics. And some of the forms Lazare thinks impossible in Modern Greek show up in the current language, per Hatzidakis—and even in his own transcriptions: χωρίζει Gen 1:6, σερπετεύγει Gen 1:21, ζωγής Gen 2:7, η-γι-αδερφή Gen 4:22.

Clearly in previous work Lazare was as bad as Valetas in insisting on no final /n/, and no epenthetic [j]. In other words, expecting that 1547 Greek should be identical to 1897 Greek. That alone vindicates Dirk.

Accents are handled arbitrarily: όχτω and έκατο almost always, ασπρός, no distinction between αρχός and άρχος. Πόλεμου, σύντροφου, άθρωπου should have been accented, as nouns, on the antepenult (πολέμου, συ(ν)τρόφου, αθρώπου).

Lazarus is right about some of the misaccentuations, which are my stupid printer’s fault. But Hatzidakis has shown there’s nothing wrong with πόλεμου, σύντροφου, άθρωπου.

The text does not present the spellings βασιλεάς, αετός, γονεών, ελαιές; the translator knows no etymology, and these should just have been written βασιλιάς, αϊτός, γονιών, ελιές.

Both of them are doing historical orthography, and historical orthography had not yet settled on how to deal with /e/ > [j]. The modern orthography has gone with Lazare.

The origin of έκατσε and έτσι are obscure; we don’t see them at all in the first three books, but they are used regularly in the last two. I think this was because the Spanish publisher was unfamiliar with /ts/ from Spanish: the translator went along with this in the beginning, and avoided /ts/, but then insisted, and indeed started hypercorrecting /s/ next to /ts/: hence κάτσητς instead of κάτσης.

“Anakites” in Num 13:22, 33, Deut 2:10, 11 is rendered as Ελλήνου, Ελλήνων. Why “Hellenes”? Maybe from a midrash. Num 13:22 says that Talmai was the offspring of Anak, and the Rabbinic literature associated Talmai with the Greek Ptolemies.

The translator understands the word for “spring, first months of the year” as somehow relating to the word for “father”, and renders it as πρώιμο. Dirk gives it in Ex 34:18 and elsewhere as το μήνα των πρώιμων, which he understands as “the month of early fruits”. But “early fruits” are rendered in Ex 34:22 as πρωιμάδια.

As Lazare argues elsewhere, the translator is bloodymindedly literal enough that he always renders the same Hebrew word with the same Greek word. But I don’t see what πρώιμος means if not “early (as of fruit)”: I’m not seeing evidence of the translator’s misderivation of the Hebrew word from “father”.

Dirk’s α(μ)πελές in Gen 25:16 is his hurried reading of εις τις αυλές τους. Ex 27:8 κουφοπλακώνει is just κούφ(ι)ο πλακ(ι)ώ, and adjective and a genitive plural noun. Num 5:27 χτίστε should be να χτιστή. αλογή “aloe” should be read in Deut 29:17. να άρχουνται Num 5:29 should be να αρνούνται. κουκουτσίνα μη φάη Num 6:4 is not a new-fangled noun, but κουκούτσι να μη φάη. Num 11:26 is ανατέθην, transposing the /t/ and /θ/ of the Constantinople edition; Dirk couldn’t cope with /ain^aaθ^etⁱin/, and invents απλίκεψεν. (I’m not sure what that first /i/ is doing there either.)

Editing such a text for the first time is like editing a manuscript for the first time: there’ll be lots of bad first readings. Such as Perles, Fürst, and for that matter Belléli. I see my mistaken αμπελές, and raise you your nonsensical και γλαυβράδα Gen 3:24. Only I’m not concluding you were “hasty”; an easy blunder to make, which I fixed to κ’ εγλαμπράδα ( < έκλαμπρος). Which matches the Hebrew exactly.

Dirk records the reading αναθέτην in the Paris copy; he didn’t just invent απλίκεψεν, and would have seen it in the Wroclaw copy (which Lazare did not see). So this is bona fide textual variation. I’m not convinced the emendation of αναθέτην is necessary, given the uncertainty the modern language has had about τίθημι.

Lev 25:47 άρριζον is literally a “rootless” newcomer. I don’t know where Dirk gets his reading έρριζον from, the British Museum copy is clear.

The translator does not realise that eth- can indicate accompaniment, as well as the accusative. He translates Gen 26:10 as επλάγιαζεν… με τη γεναίκα σου, but Num 5:19 as αν δεν επλάγιασεν ανήρ εσέν—as if πλαγιάζω is a transitive verb.

Kriaras records transitive πλαγιάζω as “capsize”, “lay someone down to sleep”, and “have sex with”; but all the examples of the last are from the Torah. It’s so ill-fitting in Greek already, that I don’t doubt it is a literal translation of what looks in Hebrew like a transitive. Of course, lay in that sense in English is transitive too, so we needn’t conclude “lay” in Hebrew isn’t transitive. The point is though, this is a translationism.

Gen 8:21 ingeniously uses φύση; Dirk uningeniously reads this as ποίση, but there’s no such noun, and learnèd ποίηση does not count.

Dirk did intend a noun by this; he’d have written the verb as ποίσῃ. NIV has “every inclination of [man’s] heart is evil from childhood”; Lazare brings up the fact that the translator was guessing at the “inclination” word, and came up with “nature”. Dirk came up with ποίση “creation”, which is awkward (ότι ποίση καρδι̯ά του άθρωπου κακή [Paris: κακό] από τα παλληκαράτα του).

Deut 18:27 has the correct ληνό, a press, which the Dutch editor misconstrues as λινό, linen.

“violate, desecrate” is often rendered by a verb λιτώνω, which I had derived from λίθος in 1890, and Dirk had gone along with me. I now think it’s λυτώνω, from λυτός “loose”; cf. γλυτώνω.

Kriaras has λιτώνω, etymology unknown. This book review is not in Kriaras’ bibliography—which is otherwise all-encompassing; and the post-1997 hiatus volumes of Kriaras no longer have corrections to earlier entries. There’s a reason I’m blogging this stuff.

Apart from ολιγώτερο Deut 7:7, there are no comparatives in the text; the translator is happy to use a periphrastic construction in Greek (adjective + από) to match the Hebrew literally. The exception was to avoid the misconstrual of εσείς το ολιγώτερο από όλα τα έθνη, if it were rendered as εσείς το ολίγο από όλα τα έθνη. κάλλιο is no longer apparent as a comparative, and Dirk’s examples of πλια Ex 10:29 and Lev 27:26 are temporal adverbs, not comparisons.

What, to avoid the reading “you, slightest of all nations” as “you, a little bit of all nations?” I know the latter reading would sit ill with the Ancient Hebrews’ panic over miscegenation, but would the expression λίγο απ’ όλα “a bit of everything, a mix” been current yet? Lazar’s right on πλια (Contemporary πια), probably right on κάλλιο.

But the linguistic value of this ‘monument of the word’ is incontestable, as it has the merit of showing us a language which has in no way undergone the influence of learnèd efforts. Noone will any longer believe, I hope, that our translator had been educated with anything even slightly literary in his Greek. Like his coreligionists, he had a shut-in life in the recesses of the ghetto, without ever coming into contact with the Christian Greeks living in Constantinople. Nonetheless, the Greek usage of the first Jewish inhabitants of Constantinople was quite old: it dated from a time when persecution and mutual distrust had not yet become a daily fact. Except for some religious terms which they had to have as their own—just like the Christians did—the Jews of that region spoke the language common to all the people before entering the Byzantine stage. They underwent with the others the same phonetic changes and evolutions in morphology and syntax. Though the new segregation laws separated the two races, isolation was neither strict enough nor continuous enough to decide different linguistic developments.

We have tried to show in the first part of this study that in phonetics and morphology, there are no phenomena distinct to this text; we will do an even better job of this when we come back to this text, in studies which will treat in more detail questions of direct interest to linguistics. Our task here has been to deal with issues more specific to Hebrew. For syntax, we were the first to warn Hellenists that there is much of the original language left, but we cannot at all agree with Dr Hesseling in dismissing as Hebraicisms any fact which needs some effort to be explained. Particularly the issue of the genitive plural, which tends to be represented as an accusative, deserves more of our attention. Certain irregularities—and unfortunately they are not a few—are only apparent, and are merely due to the haste with which the Dutch editor went about his work. With more time and circumspection, many difficulties would have been made clear without too much effort. Dr Hesseling made his edition without become familiar beforehand with the language of the original: he has seen chasms where they’re aren’t any, but there are also bumps in the ground which he hasn’t realised, and he has run up against them—not without damage. And concerning Greek itself, he has not always been right: within him theory has not been accompanied by practical knowledge of the language, which would have clarified obscure points and removed so many doubts.

In brief, his work, though conceived with laudable intentions, has not reached its goal, and realises in only a very incomplete fashion the desire of so many scholars to finally have a readable edition of the Greek Torah of Constantinople.

Such examples as these show that arriving at a perfect understanding of such a text as this requires capabilities which are rarely sufficient in a single person. For my part, I confess that reading Mr Belléli’s not exactly benevolent article has provided me with clarifications on several points. And even though in his critique he does not seem to have gleaned anything good from my work, for my part I have no hesitation in recognising that I am indebted to Mr Belléli’s study for precious information.