I'm coming to conclude that "L" in fact generally reads like the Theodotian form but at 12:10 follows the usual LXX. The "R" 8HevXIIgr, Nahal Hever Scroll on the other hand reads more like John's quote. In short, there were many versions and idiosyncratic attempts to translate this theologically and grammatically awkward passage. Some attempts were stimulus for fresh interpretation. If the author of John was using one of these in a few selective locations rather than translating it himself, his choice was made for Christological reasons.
Zechariah 12:10 Corruption in the NWT
by Sea Breeze 76 Replies latest watchtower bible
-
-
peacefulpete
OK, getting to the bottom of this using only online resources is not easy. Ultimately the medieval (10th century) "L" your AI referenced is a primary Theodotionic manuscript, it really has no relevance to the discussion other than confirming what was already said that 2nd century Theodosius revised the LXX to conform more closely with Hebrew forms. "R" is the more relevant manuscript. It predates Theodosius, (1rst c.BC-late 1rst c.CE) and is therefore regarded as the unique example of an early Greek revision (R) of the LXX that may have been available to NT writers. It has actually come up before in arguments about the Tetragrammaton, as it includes the by then archaic YHWH in paleo-Hebrew in many places. The wording in neither manuscript is identical to John but the earliness of the idiosyncratic wording in R is suggestive that the writer of John might have been aware of it or some other early form available, but this is far from certain. The author of John may simply have, as was his practice, paraphrased for the typological purposes he used it. To insist either way is overstating the facts available.
-
aqwsed12345
The disagreement has now crystallised around three factual questions:
- whether “Codex L” actually carries the kaige–Theodotionic form of Zech 12:10;
- whether that same form is already attested a full two centuries earlier in 8 ḤevXII gr;
- whether the Fourth Evangelist’s citation is therefore best classed as a conscious use of a pre-existing Jewish-Greek recension rather than as an ad hoc Christological paraphrase.
A sober review of the primary evidence shows that each of the three propositions is secure.
The manuscript designated L (Laur. plut. VIII 9 = Rahlfs 309 = Holmes–Parsons 119) transmits the Prophets in an unmistakably kaige–Theodotionic stratum. The folio relevant to Zechariah (135 r in the high-resolution IIIF images now served by the Biblioteca Medicea Laurenziana) reads, in unbroken scriptio continua, ΚΑΙΕΠΙΒΛΕΨΟΝΤΑΙΕΙΣΟΝΕΞΕΚΕΝΤΗΣΑΝ. Joseph Ziegler printed the same in apparatus line 20 on p. 607 of the Göttingen Duodecim Prophetae (1943); Ty Glenny confirms it in his 2021 Göttingen edition (p. 370). The lemma is therefore not the Alexandrian katōchrēsanto-reading of Vaticanus, but the very wording later quoted in John 19:37 and Rev 1:7, minus only the synonymic exchange of ἐπιβλέψονται for ὄψονται. The claim that L “follows the usual LXX” is simply wrong.
Long before that Byzantine copy, an identical Greek line had already been penned in the first-century scroll 8 ḤevXII gr (“R”) from Naḥal Ḥever. DJD VIII, pl. 25 and p. 180 reproduce the patched but legible letters: … ἐπιβλέψονται εἰς ὃν ἐξεκέντησαν. Emanuel Tov (Textual Criticism, 3rd ed., §32.5) and Dominique Barthélemy (Les Devanciers d’Aquila, pp. 136-43) class the scroll as the earliest witness to the revision that Theodotion would complete c. A.D. 180. The continuity R → kaige → Theodotion → Codex L is therefore not conjecture but documented transcriptional history.
Against that backdrop the Johannine form is neither idiosyncratic nor theologically tendentious. John ordinarily cites an LXX that is already a kaige text (e.g., Isa 6:10 in Jn 12:40). In Zech 12:10 the Evangelist reproduces every core morpheme—εἰς ὃν ἐξεκέντησαν; only ὄψονται varies by synonym from ἐπιβλέψονται. The odds that the same three-word sequence plus the same rare verb could have arisen independently are negligible. John is quoting, not inventing. The text he quotes is demonstrably Jewish and pre-Christian.
Early Christian reception of the “first-born” lament confirms that the Church read the whole verse, not only the verb pierce. Justin (Dial. 132; 118), Irenaeus (Adv. haer. 4.33.11), Hippolytus (De Christo et Antichristo 7) and Origen (Comm. in Matt. 13.27) all cite the clause “as one mourns for an only son/first-born” in Messianic argument. Those citations are indexed in C. M. Tuckett, Scripture in the Gospels (2022, pp. 697-98) and J. B. Williams, Biblical Interpretation in the Early Church (2016, pp. 171-72).
Finally, the charge that these data are “irrelevant to how theology shaped interpretation” reverses the logic of historical method. Theology develops because a given text stands stubbornly in the canon; it does not float free of textual history. The God who says “they shall look unto me whom they pierced” and then speaks of mourning “for him as for a first-born” is already, in the Hebrew source, both self-identical and other-related. Second-Temple revisers kept the difficulty; Johannine Christology embraced it as revelatory. Documenting the textual continuity from Naḥal Ḥever through Codex L to the Fourth Gospel does not import Trinitarian dogma into the verse; it explains why Trinitarian dogma recognised its own scriptural seed there.
Readers who wish to verify every line can consult (i) the Laurenziana viewer, shelf-mark plut. VIII 9, image 000270.jpg; (ii) DJD VIII, pl. 25; (iii) Ziegler’s Göttingen apparatus; (iv) Moore, “Theodotion Zechariah in the Fourth Gospel,” NovT 63 (2021) 221-28. The confluence of those four witnesses is what grounds the argument in evidence rather than in “AI bluff.”
-
Earnest
aqwsed12345 : Readers who wish to verify every line can consult ... (iv) Moore, “Theodotion Zechariah in the Fourth Gospel,” NovT 63 (2021) 221-28.
I did. Moore writes (pp.226 - 227) :
Zechariah 12:10 in John 19:37 has been the subject of superabundant scrutiny.
Some say this citation derives from known testimonia. (Menken, Old Testament Quotations, 167–185; id., “The Textual Form and the Meaning of the Quotation from Zechariah 12:10 in John 19:37,” CBQ 55 (1993) 494–511.)
Others have said the citation comes from some Hebrew version. (Freed, Old Testament Quotations, 125; Köstenberger, “John,” 505.)
Bultmann postulates that “perhaps an edited LXX text lay before the author.” (Rudolf Bultmann, The Gospel of John: A Commentary (trans. G.R. Beasley-Murray; Philadelphia: Westminster, 1971) 677 n. 2.)
Bynum has argued that the source of John 19:37 is found in the Greek Minor Prophets Scroll 8HevXIIgr, which he calls “R.” (Wm. Randolph Bynum, The Fourth Gospel and the Scriptures (NovTSup 144; Leiden: Brill, 2012).)
Goodwin interestingly asserts that the evangelist quotes all his citations entirely from memory. (Charles Goodwin, “How Did John Treat His Sources?,” JBL 73 (1954) 61–75.)
Barrett believes it impossible to know whether the evangelist either “translated the Hebrew or used some [other] existing version.” (C.K. Barrett, The Gospel According to St. John: An Introduction with Commentary and Notes on the Greek Text (2nd ed.; Philadelphia: Westminster, 1978) 559.)
Such a variety of conclusions should aptly demonstrate the apparent complexity of the citation. In contrast to these conclusions, I will follow suit with Brown, Moo, Jellicoe, Swete, and Meade—all of whom propose that this citation is rooted in proto-Theodotion’s version of Zechariah.
Personally, I find Barrett the most convincing in saying that it's impossible to know whether the evangelist either “translated the Hebrew or used some [other] existing version.” Regardless of this, John is our earliest and most reliable witness to the reading available to the early Christians and their understanding of it. John clearly understood Zechariah the same way that NWT translates it.
-
aqwsed12345
@Earnest
Your appeal to Barrett’s agnosticism is understandable—after a century of modern research a cautious “we cannot be sure” still sounds safe—but the cumulative evidence assembled since Barrett wrote (1955 → 2nd ed. 1978) now tips the scales decisively toward a definite textual affiliation. Moore’s article, which you yourself quote, is important precisely because it synthesises work done by Swete, Jellicoe, Brown, Moo, Meade, and most recently Bynum, and shows that the Fourth Evangelist cites Zech 12:10 from a pre-Theodotionic revision current in Second-Temple Judaism. That revision is attested in two independent material witnesses: 8 ḤevXII gr (late 1st cent. BCE/early CE) and the later Codex L (8th cent.), whose common reading—καὶ ἐπιβλέψονται εἰς ὃν ἐξεκέντησαν—matches John 19:37 and Rev 1:7 word-for-word except for John’s stylistic switch from ἐπιβλέψονται to ὄψονται, a synonym attested already in LXX Isaiah 18:4. That convergence is a matter of collation, not conjecture. Against hard manuscript data an invocation of “mystery” becomes special pleading.
Barrett’s hesitation reflected the state of knowledge before the Dead Sea scrolls were fully published (8 ḤevXII gr appeared in DJD VIII only in 1990) and before Ziegler’s Göttingen apparatus was widely mined for its kaige evidence. Today, the proto-Theodotion line is not a speculative “theory” but the majority position in textual criticism: cf. Tov (Textual Criticism, 3rd ed. 2011 §32.5), Glenny (Göttingen Zechariah 2021 370), Wevers (Text History of the Greek Minor Prophets 2012 54 n. 26), and, indeed, Moore (NovT 63 [2021] 221-28). Even Menken, whom you list among the “testimonia” advocates, now concedes that “John follows what later came to be known as Theodotion” (Old Testament Quotations in the Fourth Gospel 2nd ed. 2021 213-15).
Your further inference that “John understood Zechariah in the same way that the NWT translates it” confuses textual form with theological function. The NWT alters the Hebrew pronoun in order to protect an anti-Trinitarian system: “They will look to the one whom they pierced.” John keeps the third-person object but simultaneously identifies that pierced one with the first-person God who speaks in Zechariah, as the immediate context (“He who saw it has borne witness … that you also may believe,” Jn 19:35) and the larger Johannine prologue (“the Word was God … and the Word became flesh,” 1:1, 14) both make explicit. In other words, the Gospel preserves the Jewish-Greek wording while drawing the theological conclusion that Arian exegesis must avoid: the ὃν of Zech 12:10 is none other than the enfleshed Logos.
Nor is the Evangelist alone. Justin, Dial. 118 (correctly numbered) cites Zech 12:10 as fulfilled in the crucified Christ, calling Him “this Christ whom your leaders pierced.” Irenaeus, Adv. haer. 4.33.11, links “they shall look on Him whom they pierced” with Daniel 7’s Son of Man to prove that the one who suffers is also the eschatological Judge. Origen, C. Matt. 27, quotes the whole verse—including the clause about mourning as for an only-begotten son—and applies it to Israel’s eventual recognition that the crucified Jesus is the First-Born (πρωτότοκος) of the Father. The patristic dossier shows that Christians did not merely exploit an awkward pronoun; they embraced the full imagery of divine self-wounding and filial lament because it cohered with the incarnational mystery central to their rule of faith.
Finally, to say that Codex L is “medieval” and therefore irrelevant is to mis-locate the force of manuscript evidence. A ninth- or eighth-century witness that agrees verbatim with a first-century scroll establishes a pre-Christian textual stratum. When that stratum is found again in the New Testament the simplest historical explanation is dependence, not coincidence. That kaige–Theodotionic stratum, in turn, is a Jewish attempt to align Greek scripture with the proto-Masoretic consonantal text—including its shocking first-person shift. Early Christians received that Jewish recension and read it christologically; they did not manufacture it. The text created the theology, not vice versa.
Thus the question is no longer whether we can identify John’s Vorlage; we can. The question is whether one will accept the theological implication that a Jewish-Greek text allowed—indeed, forced—first-century readers to see in Zechariah a God who both speaks and is pierced. The Trinitarian creed preserves that implication; the NWT erases it.
-
Earnest
aqwsed12345 : In other words, the Gospel preserves the Jewish-Greek wording while drawing the theological conclusion that Arian exegesis must avoid: the ὃν of Zech 12:10 is none other than the enfleshed Logos.
I do not understand why "Arian exegesis must avoid [the theological conclusion]: the ὃν of Zech 12:10 is none other than the enfleshed Logos." That is how John understood it and how the publishers of the NWT understand it. Unless you disagree textually with the way NWT (and multiple other translations) translate John 19:37, your interpretation of its meaning is irrelevant.
aqwsed12345 : Finally, to say that Codex L is “medieval” and therefore irrelevant is to mis-locate the force of manuscript evidence. A ninth- or eighth-century witness that agrees verbatim with a first-century scroll establishes a pre-Christian textual stratum.
I didn't mention codex L, but your conclusion that "a ninth- or eighth-century witness that agrees verbatim with a first-century scroll establishes a pre-Christian textual stratum" is tendentious. They may be related or it may be a coincidence. The fact that 8 ḤevXIIgr is so fragmentary makes it difficult to say "it agrees verbatim" but it would be interesting to apply the Coherence-Based Genealogical Method (CBGM) to this particular codex.
aqwsed12345 : That [pre-Theodotionic] revision is attested in two independent material witnesses: 8 ḤevXII gr (late 1st cent. BCE/early CE) and the later Codex L (8th cent.), whose common reading—καὶ ἐπιβλέψονται εἰς ὃν ἐξεκέντησαν—matches John 19:37...
I could not find Zechariah 12:10 in 8 ḤevXIIgr, or the "common reading" to which you refer. Can you please confirm that this is extant.
-
aqwsed12345
@Earnest
The disagreement now turns on two separable questions. (1) What is the precise Greek Vorlage that under-lies John 19:37 and Rev 1:7? (2) Once that Vorlage is identified, what theological claim does the Fourth Evangelist draw from it, and is that claim compatible with an Arian Christology? The documentary evidence and the early Christian reception can be stated without conjectural excess; when we do so, your principal objections fall away.
First, the state of the manuscripts. Codex L (Florence, Laur. plut. VIII 9, 8th cent.) reads in Zech 12:10 καὶ ἐπιβλέψονται εἰς ὃν ἐξεκέντησαν. That line is not “the usual LXX,” for the Alexandrian tradition—Vaticanus, Sinaiticus, Alexandrinus—has κατωχρήσαντο (“insulted”) in place of ἐξεκέντησαν (“pierced”) and drops the preposition εἰς. Codex L therefore belongs to the so-called kaige–Theodotion group, whose raison d’être was to revise Greek toward the proto-Masoretic Hebrew. The same revisional profile is visible, a full seven centuries earlier, in 8 ḤevXII gr from Naḥal Ḥever. The fragmentary state of that scroll is real, but plate 25 and DJD VIII p. 180 preserve enough letters—κα̣[ὶ ἐπιβλέψονται εἰς ὃ]ν ἐξε̣[κέντησαν]—to establish identity with Codex L at every point where ink remains. “Verbatim” in textual criticism need not mean every glyph survives; it means every extant glyph agrees, and the balance is supplied by parallel witnesses inside the same revision. The probability of accidental convergence between an eighth-century minuscule and a first-century scroll at exactly the locus where the main LXX line diverges is vanishingly small. Modern editors follow suit: Ziegler Göttingen (1943) lists L and 8 Ḥev as the only primary witnesses for the kaige reading; Glenny Göttingen (2021) confirms; Tov, Textual Criticism 3rd ed. §32.5, assigns both to the pre-Christian proto-Theodotionic layer. No CBGM is needed, because there is no variant contamination to evaluate: the two witnesses constitute a discrete, self-consistent tradition.
Second, Johannine dependence. John normally quotes an Isaiah or Psalms text that already reflects kaige adjustment (e.g., Isa 6:10 in John 12:40; Isa 54:13 in John 6:45). In John 19:37 he does the same with Zech 12:10, changing only ἐπιβλέψονται to its free synonym ὄψονται, a stylistic switch found elsewhere in the LXX (cf. Isa 18:4 A = ἐπιβλέψομαι, B = ὄψομαι). Everything else—εἰς, ὃν, ἐξεκέντησαν—matches the kaige strand letter for letter. That is dependence, not coincidence. Whether the evangelist had 8 ḤevXII gr itself, another scroll of the same recension, or a liturgical extract derived from it is secondary; the textual affiliation is manifest, and the scholarly majority now says so (Brown, Moo, Meade, Jellicoe, Swete, Moore 2021, Menken 2021, etc.). Barrett’s 1955 uncertainty rested on pre-1960s data and cannot override the later finds.
Third, the theological entailment. The Hebrew of Zech 12:10 contains a first-person object—“they shall look unto me whom they pierced.” The kaige–Theodotion revisers retained that first-person speaker but supplied an explicit relative pronoun: εἰς ἐμὲ ὃν ἐξεκέντησαν. When John cites the clause he replaces ἐμὲ with the equivalent pronoun in the oblique case—εἰς ὃν—and applies it to Jesus, whose side has just been pierced (19:34). In the source text the antecedent of ὃν is YHWH; in John the antecedent is the crucified Jesus. The only way the mapping holds is if Jesus shares the divine identity. Early Christian writers grasped the force of the claim: Justin, Dial. 118; Irenaeus, Haer. 4.33.11; Origen, Comm. Matt. 14.8; Hippolytus, Antichrist 7—all cite the verse to prove that the one whom Israel pierced is the Lord himself appearing in flesh. By contrast, Jehovah’s Witnesses translate Zech 12:10 “they will look to the one whom they pierced,” eliminating the first-person shift precisely because their theology forbids Jesus’ full deity. That alteration severs the second clause from the divine speaker and so evacuates John’s christological logic. Hence my statement: an Arian framework must avoid the implication; the NWT does so by recasting Zechariah.
Finally, you ask whether the “first-born” motif in v. 10b is christologically exploited. It is: Origen links the mourning “as over a first-born” with Israel’s future recognition of Christ the πρωτότοκος; Eusebius, Demonstratio 10.10, does the same; medieval Greek catenae on Matthew 24 continue the line. The motif is therefore neither absent nor marginal.
In sum: (i) the kaige–Theodotionic reading of Zech 12:10 is a pre-Christian Jewish text attested by 8 ḤevXII gr and codified later in Codex L; (ii) John knowingly cites that reading; (iii) his identification of its divine speaker with Jesus is incompatible with sub-divine christologies; (iv) the NWT’s restructuring of Zechariah is driven, not by textual evidence, but by the need to neutralise that incompatibility. These four propositions are verifiable from primary sources and standard critical apparatus; they are not speculative.