Friday, December 07, 2018

John’s Bible Version in John 19:37?

See the update to this post below.
I continue my series of highlighting places where a NT author cites the Old Testament but does not use the Old Greek/Septuagint (see 2 Cor 11:3; 1 Cor 15:54). In addition, I would propose that the NT author in these cases probably does not give his own ad hoc rendering of the Hebrew, since there was a perfectly good revision of the older Greek translation at his disposal. My assumption, therefore, is that the NT author simply used and modified an already existing Greek translation of which he and his audience were aware. Here, I list the Hebrew, OG, and the readings of the Three for the part of Zechariah 12:10 that John quotes in 19:37 including some context:
Hebrew: וְהִבִּ֥יטוּ אֵלַ֖י אֵ֣ת אֲשֶׁר־דָּקָ֑רוּ וְסָפְד֣וּ עָלָ֗יו
“and they will look to me whom they pierced and they shall mourn for him.”
Greek: καὶ ἐπιβλέψονται πρός με ἀνθ᾽ ὧν κατωρχήσαντο καὶ κόψονται ἐπ᾽ αὐτόν
“and they will look to me because they danced triumphantly, and they will mourn over him.”
John 19:37: καὶ πάλιν ἑτέρα γραφὴ λέγει· ὄψονται εἰς ὃν ἐξεκέντησαν 
“and again another scripture says, ‘They look at whom they pierced‘.”
Aquila: α’ σύν ᾧ ἐξεκέντησαν καὶ κόψονται αὐτόν
“[they will look to me(?)] whom they pierced, and will mourn for him.”
Theodotion: θ’ ...πρός με ὃν ἐξεκέντησαν καὶ κόψονται αὐτόν
“[they will look] to me whom they pierced, and they will mourn for him.”
Symmachus: σ’ ἔμπροσθεν ἐπεξεκέντησαν καὶ κόψονται αὐτόν
“[they will look to me(?)] before/in whose presence they pierced, and they will mourn for him.”
The OG’s κατορχεῖσθαι “to dance triumphantly” is a hapax legomenon in the Greek Old Testament’s corpus and probably resulted from reading a form of the verb רקד “to leap about, dance” Piel, which metathesized ד and ר due either to ד/ר confusion or exegesis.

The readings of the Three were originally incorporated into Origen’s Hexapla but come down to us via Ra 86 (image from DigiVatLib). John clearly depends on Theodotion’s version for his quotation of Zechariah 12:10, not the Old Greek. However, John has also modified it slightly by using a different preposition than Theodotion (but see the Syrohexapla for the Th fragment which could be retroverted as εἰς which would mean that Th’s version equal’s John’s form of the quotation in this respect). In any case, John has certainly not read with the Old Greek in this place but rather the revision of it.

The apostles (at least Paul, John, and Matthew) were aware of not only the older Greek version but also other forms of the Greek scriptures, for they cite and quote them too. What factors led to their choice? The Hebrew source? The texts at their disposal in any given situation? We don’t know. But what seems clear is that these Jewish followers of Jesus had not declared an exclusive preference for the older Greek version. At one point, they are quoting from the ‘LXX’ and at another point they are quoting from one of its revisions. We would do well to bear this phenomenon in mind as we continue to read the NT’s use of the OT and also how these matters develop in the second century and beyond.

UPDATE 12/8/2018

I’ve now had the chance to look at Syrohexapla (Syh fol. 112r) for the Theodotion reading in Zach 12:10 (ܒܗܘ), and no doubt, the translator rendered an equivalent for a Greek preposition before the relative pronoun ὅν. The beth is often used for εἰς in this tradition thus Ziegler’s εἰς in the second apparatus is probably correct.

In my mind, then, John removes πρός με (contextual to be sure) and modifies the verb from ἐπιβλέψονται to ὄψονται. But εἰς ὃν ἐξεκέντησαν seems to be the original reading of Theodotion and that would give three words of correspondence. Even Ra 86 agrees with John on the choice of the relative pronoun for two words of correspondence.

When compared with Aq, Sym, and OG, we see that not all come to the same rendering of the Hebrew which makes agreements between Th and John all the more interesting. Lastly, for the key word “pierce,” John had several lexical options in Greek but landed on Theodotion’s equivalent. It could be coincidence. But presuming that version is already around, I don’t think we need to argue along those lines in this case.


  1. It does not seem in itself to be to be a very strong case for John's dependence on a textualised form of Theodotion since the citation is so short (and well suited to its Johannine context), and the problem with the OG so widely known.

    1. Right, Pete. And yet, it is a citation (γραφή = textualized?) of four words agreeing more with Th, than OG, Aq, or Sym, especially if Syh's Vorlage had εἰς. The four Greek versions show that even the rendering of five Hebrew words can lead to some different possibilities, even when Aquila would have had OG and Th before him, thus dependency and not an ad hoc rendering of the Hebrew seems likely. Even if John is citing this text from memory it would still be based upon Theodotion not the OG, right?

      How widely known was the problem with the OG in the first century? I haven't looked to see where else this text is cited. Do you know? Thanks for your comment, Pete. I was expecting some push back here :-).

  2. Are there other words for "pierce" that could plausibly be used, or is εξεκεντησαν the only plausible rendering assuming one is looking at or aware of דקרו?

    1. There is ἀποκεντέω for דקר in Num 25:8, διελαύνω in Judg 5:26, τετραίνω in 2 Kings 18:21. I only looked quickly at these. Numbers 25:8 confirmed that different complex forms were used for the same Hebrew word. But to your point, Stephen, if John were only remembering this text why did he use the word in Theodotion and not any of the other words sharing a similar semantic range? Good question. Thanks.

    2. Just for the sake of clarity, Jdg 5:26 is rendering מחץ and 2 Kings 18:21 is rendering נקב. But your point remains: There were other alternatives.

      Since there were other alternatives, does this suggest (and how strong is the suggestion) some inter-dependence among the Three regarding this passage?

    3. Thanks for clarifying the Hebrew in those places, Stephen.

      As you know, I think Th is the earliest of the Three and it does appear in many instances that Aquila used his version with some modification. The current text is a good example of Aquila's dependence on Th with some modification. Make sense?

    4. Yes, thanks. Thanks for posting this. Makes me want to go open up some books and bone up on the Three...

    5. Stephen, then it's working :-). Thanks for your interactions on here.

  3. Thanks for this post, John. Two comments:
    (1) this is one of the passages Jerome pointed to as demonstrating that the apostles did not use the LXX but rather the Hebrew text. See, e.g., his Preface to the Pentateuch.
    (2) since Justin Martyr is the first person we know of who included in the LXX more than just the Pentateuch, I assume the apostles would not have identified one of the available versions of the Twelve as the LXX version. I bring this up only to say that even if there is some sort of loyalty to the LXX—as there certainly is for some Jews at this time, e.g., Philo—that loyalty to a particular text form does not extend beyond the Pentateuch. I'm not suggesting that the author of the Fourth Gospel held any loyalty to the LXX, but I am saying that he would not even have known that there was a LXX version of Zechariah. Would he have recognized one version as traditional? I don't know. If he was aware of different Greek texts, I imagine he did not think of any as more official than another.

    1. To the first point, that makes curious whether Jerome elsewhere demonstrates awareness of the Three?

    2. Thanks for your comments, Ed. Two comments/questions in reply:
      (1) Does Jerome envision them using the Hebrew per se or only a Greek version close to it? As you know, "the Hebrew" can mean different things in the fourth century.
      (2) Quite right, Ed, which is why my whole post used Old Greek until the last instance of 'LXX' in quotes. Here, I'm only talking about a popular (mis)conception of the whole issue, which you helpfully point out. Even though the Jews probably would not have thought in terms of the LXX outside of the Pentateuch, the Minor Prophets scroll from Nahal Hever does show that at least some Jewish groups probably thought in terms of older and revised/newer translations. At least, that's how the matter is put. My point is only to show that several of the NT authors used these versions--not the older one. Thanks again.

    3. Jerome seems to think the apostles used the Hebrew directly. In his mind, they were Jews and therefore had no trouble consulting the Hebrew directly, and they would have wanted to consult the Hebrew directly because of the inaccurate translations available.

      On the second point, I guess Theodotion's version may have circulated under his name (already in the first century)? I'm just trying to think of how readers might (or might not) have recognized different available versions. I guess the OG translations circulated under no particular names, though, which allowed for the adoption of them into the LXX (attributed as a block to the Seventy translators).

    4. Yes, Jerome makes use of Origen's Hexapla, and he is fully aware of the Three, and makes substantial use of them in his translations and commentaries.

    5. That's very interesting about Jerome's view of the apostles' use of the Hebrew in those instances. Does he ever put that view into close connection with the versions of the Three? That is, does he somehow make a distinction between the apostles' ad hoc renderings and the version of Theodotion? He couldn't somehow say Theodotion's translation of Isa 25:8 was a mistranslation because Paul used it in 1 Cor 15:54, for example. This is what makes me think at times he and others simply collapse the readings of the Three into the category of "the Hebrew."

      I'm not sure whether Th's version would have circulated under his name but I suppose it's possible. I would think some, especially like Paul, would know that Jews in diaspora are reading something closer to the OG than the revisions coming out of Palestine, even if they would not necessarily describe the situation that way.

  4. Thanks for your work on this, John; I've enjoyed chatting with you privately, and I thought I'd add to our discussion publicly. Basically I agree with Peter Head’s initial comment that the case for John’s dependence on Theodotion is not very strong here. My further comments refer specifically to kaige, which is part of the Theodotion tradition but is not identical to Theodotion’s recension in the Minor Prophets.

    I’m not as inclined to grant εἰς as the preposition here. You point out that “beth is often used for εἰς in this tradition,” but this is somewhat misleading or overstated. Correct me if I’m wrong, but the only instances of kaige *changing* an OG preposition to εἰς for ב occur in Mic 5:4, 5, where kaige twice changes OG ‘to (ἐπί; Heb. ב) the land’ to ‘into (εἰς) the land,’ which is more precise grammatically. It would be even more surprising for kaige to render אלas εἰς, I think.

    Even more rare would be a verb of seeing + εἰς. It makes sense in John’s context (as Peter Head also noted), since Jesus’s side is pierced and onlookers could literally see inside him. Otherwise there are expressions that make sense spatially, such as βλέπω/ὁράω + εἰς to look into heaven (Susanna 1:9), the height of heaven (Isa 38:14 = Odes of Solomon 11:14), or the depths of Hades (Prov 16:25); there is also the expression to look into the face of someone, in the sense of showing partiality (1 Sam 16:7; Matt 22:16//Mark 12:14).

    However, it is rather rare to use εἰς for ‘at;’ we see it in Sirach 40:29 to look at (εἰς) another person’s table, but—in addition to 19:37—John does use it when the disciples look at (εἰς) one another (13:22), so in my estimation it suits John’s style more than kaige’s. I also think it would be a stretch to expect something other than κεντέω with a prepositional prefix, since—again, correct me if I’m wrong—ἀπό & ἐκ account for all but one of the OG renderings of Heb. דקר; the exception is the odd feet binding in Zech 13:3.

    It’s by no means impossible that John knew kaige/Theodotion, but to me the relative pronoun + *κεντέω does not suffice to prove literary dependence. The γραφή John references is Zechariah, and his use of ‘pierce’ in the allusion/quotation aligns with the Hebrew against the Old Greek, but I don’t think John is necessarily consulting a recension. For the Minor Prophets, John’s only other quotation is of Zech 9:9, and there I’m convinced that John relies on Matthew’s narrative rather than any attested Greek version.

    [As I’ve observed elsewhere (“The Equivalence of Kaige and Quinta in the Dodekapropheton,” in Found in Translation: Essays on Jewish Biblical Translation in Honor of Leonard J. Greenspoon, ed. James W. Barker, Anthony Le Donne, and Joel N. Lohr [West Lafayette: Purdue University Press, 2018] 127–52, here 152 n. 61), “The relation between kaige/Quinta and Theodotion in the Dodekapropheton needs further study;” yet it is clear that “Theodotion readings in the Minor Prophets are distinguishable from kaige/Quinta.”]

    [Re Zech 9:9, see my John’s Use of Matthew (Minneapolis: Fortress Press, 2015), ch. 4, where I rehabilitate Edwin Freed’s 1961 argument in JBL.]

    1. James, thanks so much for your comment. I've enjoyed the offline discussion and am glad you have commented here.

      Probably, I wasn't as clear as I needed to be in my UPDATE. My point there was that the Theodotion fragment in Syh has a beth preposition in Syriac which often renders εἰς in its Greek source. If that's true here, then Theodotion in this source would be: ἐπιβλέψονται πρός με εἰς ὃν ἐξεκέντησαν. This is why we need a critical edition of these fragments of the Hexapla because this reading is largely lost in Ziegler's second apparatus. This would now mean that Theodotion's rendering has quantitative agreement with his Hebrew source even if εἰς for את seems a bit strange to me at present.

      This reading could alleviate some of the issues you raised about the valency of εἰς with a verb of seeing. Theodotion's reading originally had πρός with the verb of seeing. John had no use for πρός με but used the second phrase instead which worked better for his context, perhaps.

      I still struggle to see what use John would have for the Greek equivalent of Hebrew דקר, if he's citing the Hebrew into Greek from memory. Couldn't he have used any Greek verb with the sense of the Hebrew "pierce"? At the point at which he's not using OG or recension, how does he land on Theodotion's equivalent and not some other? There were other Greek words for "pierce", even other compounds with *κεντέω, yet he does not use them. It could be coincidence. But if Syh preserves the original Theodotion reading, I'm inclined to say that John knew of and modified an existing revision rather than his own ad hoc translation that comes awfully close to an existing version.

      I could be wrong here. There are more pieces to this puzzle that need to be solved before a firm conclusion can be made. (1) We need critical editions of the hexaplaric fragments. (2) We still need a better way of talking about kaige and the members within its tradition. (3) We still need to determine what access NT authors had to these early revisions/texts and what access they would have to the Hebrew texts, either orally or in written form. But I'm thankful for the exchange, James.

  5. Thanks for all of this John. Are we assuming the version of John's Gospel we have was not finalised until after Theodotion's revision was finished? Would this put Jn not being finished till after, say AD 150 ?

    1. Hi Stephen, I and some others date historical Theodotion to the first century. The order Aquila, Symmachus, and Theodotion comes from the order they appear in the Hexapla, probably not chronology. I can’t defend that here but follow the link in the post more on Theodotion’s version.

  6. Thanks for this post. As little as we know about the Three, it's exciting to see all three of them used for source criticism of NT quotations. I have a question: why do you renderκαὶ κόψονται ἐπ᾽ αὐτόν as "and mourned over him" in OG and as "and [they] will mourn for him" in all the other translations?

    1. Thank you for this comment. I have corrected the translation of the OG. They should all be future tense.

  7. Hello Mr. Meade. Where could you read the entire verse 10 of Zechariah 12 of the three versions it alludes to in Greek? Thanks, a little late, but I found this site today.