“He will see light” in Isaiah 53:11

Over on Southern Equip today, I published a piece on Easter and textual criticism by commenting on a few of the crucial textual variants within Isaiah 53. I hope the post introduces more folks to some of the issues that have been long recognized, even if some of our translations are slow to incorporate them.

Isaiah 53:11 contains a significant problem in the text. I provide the main witnesses and some additional commentary on the reading in Ziegler’s II App containing the readings of the Three.

The Key Witnesses to “He will see light” in Isaiah 53:11

1QIsaa
יראה אור וישבע “He will see light and be satisfied”

1QIsaa (four lines up; image from Digital Dead Sea Scrolls)

1QIsab
יראה אור י[שבע “He will see light, he will [be satisfied]”

4QIsad
יראה או[ר ] ושבע “He will see lig[ht] and be satisfied”

4QIsad (2nd line from top; image from Leon Levy Library)

G
δεῖξαι αὐτῷ φῶς “to show him light”

In each witness, “light” is the direct object of the verb “to see.” This is a fairly common idiom, even in Isaiah (cf. Isaiah 9:2). G probably read the same Hebrew consonantal text but as a causative instead of the simple transitive.

The Key Witnesses to Isaiah 53:11 “He will see”

MT (= Vulgate, Peshitta, Targum?)
יִרְאֶה יִשְׂבָּע “He will see, he will be satisfied”

Theodotion
ὄψεται ἐμπλησθήσεται “He will see, he will be filled”

Aquila
ὄψεται ἐμπλησθήσεται “He will see, he will be filled”

Symmachus
ὄψεται χορτασθήσεται “He will see, he will be filled“

The readings of ‘the Three’ are found in Ra 86 and Q as the Edition makes plain below. But I want to comment here on a peculiarity in the Edition and an uncharacteristic infelicity in the second apparatus. Below at v. 11, Ziegler broke up the marginal note in Ra 86 (pictured below) across two separate readings, giving the impression that there were two separate fragments in the margin of the manuscript: (1) ὄψεται and (2) ἐμπλησθήσεται ἐν τῇ γνώσει αὐτοῦ. The obvious problem here is that the reader cannot tell whether the Three had the crucial word φῶς or not. One must actually consult the MS, which is easier now than ever, to see that there is one continuous fragment for the Three which omits φῶς, and therefore, the Three agree with MT. In fact, the Three are the earliest witnesses to the shorter text.

Ziegler’s Göttingen Isaias

Ra 86 (Barb. 549, f. 112v; image from Digivatlib)

Our reading is the top note in the margin with an index to δεῖξαι in the bible text. Clearly, the scribe copied one continuous note for the Three, and we do not have to wonder whether the Three had the longer or shorter reading. In the new critical edition of the Hexaplaric fragments, the one continuous reading of Ra 86 will be supplied so that future researchers will be able to access the correct reading more easily.

Thus, the main witnesses to the proto-Masoretic Text all attest the shorter text, the text without “light.”

Conclusion

A couple of factors probably decide in favor of the longer reading, “he will see light.” First and most significantly, different Qumran texts and G agree. All agree that 1QIsab is a very good representative of proto-MT and it has the longer text against MT. That 1QIsaa (a text not as close to proto-MT), 4QIsad, and G agree with 1QIsab probably shows the independence of the longer reading across witnesses. Second, it’s probable that אור was omitted because אור looks similar to the אה, and thus homoioteleuton accounts for an accidental omission of אור just before the time of the Jewish revisers.

I’m happy to see that many English translations have already adopted this reading (e.g. NIV CSB). I wonder what it would take for them all to adopt it.

Why Does Michelangelo’s Moses Look Like That?

Translations have consequences. In Exodus 34:29, there is a fascinating example of the tension between the formal and functional renderings of the Hebrew text in the history of its translation. Here are the relevant texts on which I want to focus, but if your translation offers some interesting insight, indicate so in the comments.

The Versions

MT: ֹוּמֹשֶׁה לֹֽא־יָדַע כִּי קָרַן עוֹר פָּנָיו בְּדַבְּרוֹ אִתּו
Now, Moses did not know that the skin of his face qāran [when he spoke with him].

OG: Μωυσῆς οὐκ ᾔδει ὅτι δεδόξασται ἡ ὄψις τοῦ χρώματος τοῦ προσώπου αὐτοῦ ἐν τῷ λαλεῖν αὐτὸν αὐτῷ.
Moses was not knowing that the appearance of his face’s skin was magnified [while he spoke with him].

Aquila (apud Jerome Am III 6.13): et Moyses nesciebat quia cornuta erat species uultus eius.
And Moses was not knowing that the appearance of his face was horned.

Vulgate: et ignorabat quod cornuta esset facies sua ex consortio sermonis Dei
And he was ignorant that his face was horned [from conversing with God].

Aquila’s Version

The first issue to sort out is the Greek text of Aquila. In his commentary on Amos 6:13 (PL 25:1067), Jerome is commenting on the noun Karnaim (קַרְנָיִם) “horns,” and among other texts he appeals to the Hebrew and Aquila’s edition of Exod 34:29 for an understanding of a person with horns. He does not provide the Greek reading of Aquila (which has not been preserved), but Aquila used κέρας and derivatives systematically for קרן and derivatives so we can reconstruct his version with some probability. Jerome used the adjective cornutu “horned,” so probably Aquila had something like ὅτι κερατώδης ἦν ἡ ὄψις τοῦ δέρματος τοῦ προσώπου αὐτοῦ. I doubt whether Aquila would have used ὄψις and not the nominative δέρμα “skin”, which would agree more closely with the Hebrew, but that’s another question for another day.

Hebrew Meaning

The reason for the differences in translation comes from whether to render the Hebrew text formally or functionally. Qāran is from qeren “horn,” which is often times an image of strength in the ancient world and the Hebrew Bible, specifically the strength of a king (e.g. Deut 33:17; 1 Sam 2:1, 10). In the ancient world, gods and kings were often described as horned as a measure of their great or superior status, and perhaps the horns were a symbol of the deification of the king. Thus the denominative verb “is horned” (cp. the Hiph stem in Ps 69:32 of a bull “sprouting horns”) could symbolize Moses’ strength as Israel’s leader (cf. Exod 4:16).

On the other hand, there are other references to horns in the Hebrew Bible such as the horns on the altar (many places in Exod and Lev), which would probably not symbolize superiority, but atonement and meeting with God. As mediator, Moses’ horns would perhaps fit with this background as well.

Conclusions

OG-Exod interpreted Moses’ horns with a metaphorical rendering by assigning shining, glory, or magnificence to Moses’ face in the presence of Yahweh (cp. Targ and Pesh). Aquila revised the text according to etymology or ultra-literalism, and thus restored the ancient picture of a leader or mediator with horns. Jerome continued this tradition in his Vulgate, which must have also impacted the interpretation of Moses’ appearance by the time Michelangelo put chisel to marble to sculpt his Moses.

Thus Jerome did not mistranslate the Hebrew (neither did the LXX for that matter). But he did borrow the ultra-literal translation of the Hebrew that Aquila had already supplied. And it is this rendering that explains why Michelangelo’s Moses looks the way it does.

Textual Examples Wherein MT and the Jewish Revisions Differ

In this post, I give a few examples wherein Theodotion, Aquila, or Symmachus reflect a different vocalization of the consonantal text than what the later Masoretes recorded as the traditional reading. The issue is this: how closely do the Three (1–2 century Jewish revisers of the Greek Jewish Scriptures) mirror the Masoretic Text (9–10 century)? Of course, the general answer is that they followed the proto-MT closely, but that is different than saying they agree with the MT perfectly. As a caveat, textual criticism focuses on the differences between texts (which is what I'm about to do), but let's not let these relative few, but important differences, distort our view of the overwhelming agreement between MT and the Three. It's difficult to overstate the Three's close agreement with MT, which is why it would be easy to gloss over places where they disagree. My examples come from Job and Isaiah, and they could be multiplied.

Job 34:6a

MT: עַל־מִשְׁפָּטִי אֲכַזֵּב
Concerning my judgment/right, I lie

OG: ἐψεύσατο δὲ τῷ κρίματί μου
He lied about my judgment/right

Theodotion and Aquila: περὶ τὴν κρίσιν μου ψεῦσμα
There is a falsehood/lie concerning my right/judgment

Comment: The Old Greek read אכזב as a verb similar to later MT, while Th and Aq read it as אַכְזָב, an adjectival/nominal "false" or "falsehood." They read the same consonants with different vocalizations. As an aside, HALOT's entry of אַכְזָב probably could have cited the readings of Theodotion and Aquila here in support of this rarely attested Hebrew lexeme (cf. HALOT 1:45).

Job 35:9a

MT: מַרֹב עֲשׁוּקִים יַזְעִיקוּ
Because of the multitude of oppressions they cry out

Theodotion: ἀπὸ πλήθους συκοφαντούμενοι κεκράξονται
From a multitude those oppressors/those being oppressed will cry out.

Symmachus: συκοφαντιῶν
Of oppressions

Comment: There is no Old Greek for this verse, the Greek line in our MSS coming from Theodotion. Sym agrees with the vocalization of MT, “oppressions”  (cp. Ecclesiastes 4:1), while Th read עָשׁוֹקִים “oppressors” or עֲשׁוּקִים “the oppressed” (the latter option may equal the vocalization of MT but indicates a different derivative, the pl. pass. ptc.). In any case, Theodotion and MT attest to the same consonantal text but different vocalizations or understandings of those consonants. Or, if we want to read MT as the pass. ptc., then Symmachus has the different reading or understanding.

Isaiah 3:12a

MT: ֹוְנָשִׁים מָשְׁלוּ בו
And women rule him

OG: καὶ οἱ ἀπαιτοῦντες κυριεύουσιν ὑμῶν
And creditors rule you

Theodotion: δανεισται
Creditors

Aquila: ἀπαιτοῦντες
Creditors

Symmachus: γυναικες
Women

Comment: Symmachus agrees with MT in his reading of נשים "women." But Theodotion and Aquila read נשים as נֹשִים "creditors" from I נשׁא/II נשׁה “to lend” (the analogous formation of III-ה and III-א verbs). Therefore, this is another example of some of the Jewish revisers agreeing with the Old Greek's reading of the consonantal text where the Masoretes preserved a different vocalization, still an ancient reading as Symmachus confirms.

Isaiah 53:8b

MT: ֹמִפֶּשַׁע עַמִּי נֶגַע לָמו
Because of the transgression of my people, the strike was to them.

OG: ἀπὸ τῶν ἀνομιῶν τοῦ λαοῦ μου ἤχθη εἰς θάνατον
Because of the lawless deeds of my people, he was led to death.

Theodotion: ἀπὸ ἀθεσίας τοῦ λαοῦ μου ἥψατο αὐτῶν
Because of the faithlessness of my people, he struck them.

Aquila: ἀπὸ ἀθεσίας τοῦ λαοῦ μου ἥψατο αὐτῶν
Because of the faithlessness of my people, he struck them.

Symmachus: διὰ τὴν ἀδικίαν τοῦ λαοῦ μου πληγὴ αὐτοῖς
On account of my people's unrighteousness, the strike was to them.

Comment: The purpose of this example is not to engage the textual issue between the OG and MT (as fun and interesting as that one is). More modestly, today, I want to point readers to the fact that MT vocalized נגע as a noun (cp. Symmachus), while Theodotion and Aquila rendered the same consonants as a verb (cp. Jerome's Vulgate: percussit eos/eum).

Conclusion

There are some large-scale differences between the readings of Theodotion and proto-MT (e.g. parts of Theodotion Daniel and the longer ending of Theodotion Job). But most readings of the Three are of the kind surveyed in this post. These readings, preserved for us in Origen's Hexapla and its subsequent Christian reception, give evidence for the history of the Hebrew Bible and also for Jewish exegetical approaches to their texts around the turn of the era and into the Rabbinic period. They provide a link between the texts of the Dead Sea Scrolls and the Old Greek on the one hand and the later Medieval Hebrew MSS on the other. Thus we would do well to pay attention to them.

Paul’s Bible Version in 2 Corinthians 11:3?

The New Testament books were written at a significant time in Jewish Greek literary history. Most of the Hebrew Scriptures had been translated at various times and in various places by 132 BC (cf. the Prologue to Sirach’s reference to the Law and the Prophets and the other ancestral books as translated into Greek) with a book or two perhaps translated in the first century AD. To complicate matters, before the turn of the era, some Jews began to revise older Greek translations to bring them into greater alignment with what is now known as the proto-MT or the conservatively copied text and also these same Jews undertook new translations of some books (Lamentations, Song of Songs, Ecclesiastes and perhaps Ruth). This literary event has been called the kaige tradition (it was also called the kaige revision/recension, but that name is passing away) because the revisions and new translations often times translated Hebrew גם/וגם with καίγε (e.g. 8HevXII). Really, the hallmark of the tradition was greater fidelity to the proto-MT than their predecessors (shown in greater quantitative alignment and isomorphic renderings). This tradition culminated in the ultra-literal translation of Aquila (flourished ca. 130). Origen then incorporated Theodotion, Aquila, and Symmachus into his Hexapla along with the Hebrew and the Seventy.

But before the Hexapla, the NT authors were caught in the midst of this grand transition. Not only did they cite, allude to, and use the language of the Old Greek (commonly called the Septuagint), they also used other Greek versions of the Hebrew Scriptures circulating during the first century. Since the kaige tradition had commenced by the first century, we should expect to see some evidence of it in the NT. Formerly, scholars would refer to the kaige-Theodotion version or a Proto-Theodotion to explain this phenomenon. But these terms were introduced to explain the alleged discrepancy between a presumed historical Theodotion of the second century AD and a Greek version of the Hebrew scriptures that was extant in the first century. A better way forward is to suggest that historical Theodotion actually lived and worked in the early part of the first century and reinterpret the patristic statements that presumably locate him in the second century, but I can’t defend that thesis in this post. But what happens when we make this shift? Let’s look at Paul’s allusion to Genesis 3:1 (see edition by Field for the texts given below) in 2 Corinthians 11:3.

2 Corinthians 11:3: φοβοῦμαι δὲ μή πως, ὡς ὁ ὄφις ἐξηπάτησεν Εὕαν ἐν τῇ πανουργίᾳ αὐτοῦ...
But I fear lest somehow, like the serpent deceived Eve with his cunning, your thoughts might be led astray...

Genesis 3:1 (Old Greek): ῾Ο δὲ ὄφις ἦν φρονιμώτατος πάντων τῶν θηρίων τῶν ἐπὶ τῆς γῆς...
Now the serpent was most sagacious of all the beasts upon the earth...

Genesis 3:1 (Aquila): καὶ ὁ ὄφις ἦν πανοῦργος (Theodotion: πανοῦργος; Symmachus: πανουργότερος) ἀπὸ παντὸς ζώου τῆς χώρας...
And the serpent was crafty from every living creature of the field...

Clearly, Paul’s word choice more closely parallels the language of the Three revisers in Genesis 3:1 than the LXX. That is, Paul used the NKJV, not the KJV. The Three chose a word that was a bit more negative than the OG’s “most sagacious” to render Hebrew עָרוּם. Since the version of Theodotion was extant in the early part of the first century, Paul probably knew of it and alluded to it when making the comparison between the serpent’s craftiness and the teaching of the false apostles. 

At 2 Corinthians 11:3, NA 27/28 notes the allusion to Genesis 3:1 but not the specific version. Should NA indicate to which version the NT author alludes? It seems this could be helpful to the reader so that we don’t assume the NT author is always referring to the Hebrew or the Old Greek. What do you think?

Furthermore, we should continue to pay attention to the hexaplaric remains when we interpret the citations, allusions to, and language of the Greek scriptures in the New Testament. Dirk Jongkind has made a similar point on Matthew 2:15 here.