HistoricalChristian.Faith

Isaiah 28:11–12 at 1 Corinthians 14:21: “with men of other tongues” — did Paul quote a revised, Hebraizing Greek (Aquila-like), or render the Hebrew himself?

Type: supporting argument (off-timeline) · Assessment: contested — the catalog’s neither / free tag on 1 Corinthians 14:21 is secure at its core: Paul’s “with men of other tongues and other lips will I speak unto this people” reproduces neither the “Old Greek” Septuagint nor the Masoretic Hebrew verbatim, and it stands far closer to the Hebrew’s sense than to the very differently-worded Old Greek. What is genuinely disputed is the mechanism: whether the citation witnesses a pre-existing revised, Hebraizing Greek text of Isaiah (of the type the second-century reviser Aquila later codified) — so that even Paul’s Isaiah was textually plural — or whether it is a free, argument-driven Pauline rendering from memory, recasting the verse toward the Hebrew for his own purpose. Origen already reported that Paul’s wording matched Aquila, and one word (ἑτερογλώσσοις) is non-Septuagintal; but the divergences that most look “Hebraizing” — the shift to the first person “I will speak,” the added “not even so,” the appended “me,” the future tense, the wholesale omission of the verse’s positive half — are exactly what Paul’s glossolalia argument needs, and the citation matches no known text word-for-word. So the two readings stand roughly level, strong by different lights, with no decisive datum either way.

See also: the catalog’s tight revised-Greek case — where the composite match with the later revisers resists coincidence — is Paul’s Isaiah 25:8 at 1 Corinthians 15:54 (assessment sound); this page is its looser, harder-to-call twin. The mechanism question here mirrors the one weighed for Matthew’s formula quotations (a fresh Hebrew rendering, a Hebraizing/revised Greek, or free adaptation all fit the same non-Septuagintal data). And 1 Corinthians 14:21 is one of Paul’s standing MT-leaning counterweights to his usual Septuagintal lean, alongside two catalog rows not yet worked as their own pages — Job 5:13 at 1 Corinthians 3:19 and Job 41:11 at Romans 11:35, both closer to the Hebrew than to the wholly different Greek of Job.

The claim

“Brethren, be not children in understanding: howbeit in malice be ye children, but in understanding be men. In the law it is written, With men of other tongues and other lips will I speak unto this people; and yet for all that will they not hear me, saith the Lord. Wherefore tongues are for a sign, not to them that believe, but to them that believe not…”

— 1 Corinthians 14:20–22 (KJV), c. AD 55, citing Isaiah 28:11–12 as “the law” (Paul’s regular name for the whole Old Testament).¹ The interpretation this argument serves is the catalog’s classification of the row — that Paul’s wording follows neither received text, leaning toward the Hebrew against the Old Greek — sharpened to its strongest, most interesting form: that the citation is a genuine window onto a revised, Hebraizing Greek of Isaiah already in circulation in Paul’s day, one that later fed Aquila, so that the apostolic “Septuagint” was not a single fixed book but a plural family of texts. The steelman is weighed against its live rival: that the form is Paul’s own free adaptation of the Hebrew, shaped by the argument he is making, with no lost Vorlage behind it.

Who advances it

The observation that Paul’s citation is not the Old Greek is undisputed; the debate is over what stands behind it. The oldest witness is the strongest: Origen (c. 240), commenting on the verse, states flatly that Paul “is quoting Isaiah according to Aquila’s translation, and not according to the Septuagint,” and in the Philocalia he spells it out: “For I found the equivalent of this phrase lying in the translation of Aquila” (ἐν τῇ τοῦ Ἀκύλου ἑρμηνείᾳ). Jerome (c. 410), expounding Isaiah 28, reaches the same conclusion from the other side: he judges Paul’s “in other tongues and with other lips” to be “taken from the present chapter according to the Hebrew,” and takes it as a type-case of his own rule that when the New Testament quotes the Old “we do not follow the Septuagint, but the Hebrew.” So two ancient readers, independently, saw that Paul’s Greek matched neither the plain Old Greek nor a wooden Hebrew, and reached for a Hebraizing text (Origen) or a direct rendering of the Hebrew (Jerome) to explain it.²

Among moderns the reading has real support. Otto Michel argued that Paul’s ἑτερογλώσσοις (“of other tongues”), a word the Old Greek of Isaiah does not use, “must derive directly from Aquila’s Vorlage.” Jean Héring allowed that Paul is either quoting from memory or “dependent upon an ancient translation of Isaiah in Greek which Aquila used also,” and Hans Conzelmann leaned to the latter — warning expressly against concluding that Paul is “simply altering the text freely” — while conceding that some of the wording is Paul’s own. The catalog’s tag (“neither / free”) and the standard surveys (Archer & Chirichigno) file the row among the citations that follow a Greek diverging from the received Septuagint.³

The rival — that the form is Paul’s own adaptive rendering, not a copied text — is at least as well represented. Christopher D. Stanley, in the fullest study of Pauline citation technique, makes 1 Corinthians 14:21 a showcase of the phenomenon he traces to “interpretive renderings” produced by Paul himself, calling the relationship between Paul’s wording and the Septuagint “one of the greatest challenges in the entire corpus of Pauline citations” precisely because the wording stands so far from both the Old Greek and the Hebrew. Gordon D. Fee and the interpretive-paraphrase reading of the passage treat the citation as Paul’s own combination of exegesis and application; already H. B. Swete supposed Paul was “probably… quoting from memory… the Apostle’s knowledge of the original” letting him “improve upon the faulty rendering of the LXX.”⁴

The evidence for

Paul’s citation is demonstrably not the Old Greek — and it undoes the Old Greek’s reversal of the Hebrew. The Masoretic Isaiah 28:11 reads כִּי בְּלַעֲגֵי שָׂפָה וּבְלָשׁוֹן אַחֶרֶת יְדַבֵּר אֶל־הָעָם הַזֶּה (“For with stammering lips and another tongue he [the LORD] will speak to this people”) — God the speaker, a judgment on Israel who would not hear the prophet. The Old Greek turns the sentence around: διὰ φαυλισμὸν χειλέων, διὰ γλώσσης ἑτέρας, ὅτι λαλήσουσιν τῷ λαῷ τούτῳ (“through the contempt of lips, through another tongue, because they will speak to this people”) — reading the Hebrew laʿag not as “stammering” but as “mocking, disparagement,” and making the verb third-person plural, so that the heathen speak and the sense becomes brave endurance under foreign taunts. Paul writes ἐν ἑτερογλώσσοις καὶ ἐν χείλεσιν ἑτέρων λαλήσω τῷ λαῷ τούτῳ, καὶ οὐδ᾽ οὕτως εἰσακούσονταί μου, λέγει κύριος — the verb back in the first-person singular (λαλήσω, “I will speak”), God the speaker again, exactly as in the Hebrew’s yĕdabbēr. A writer reproducing the Old Greek would have inherited its plural subject and its “contempt”; Paul did not. His citation restores the Hebrew’s direction while remaining in Greek.¹

One word is positively non-Septuagintal, and an ancient reader traced it to a Hebraizing Greek. Where the Old Greek has the abstract φαυλισμὸν χειλέων (“contempt of lips”) and γλώσσης ἑτέρας (“of another tongue”), Paul has the concrete personal ἑτερογλώσσοις (“men of other tongues”) and χείλεσιν ἑτέρων (“lips of others”) — a wording that renders the Hebrew’s בְּלַעֲגֵי שָׂפָה (“stammerers of lip”) more literally than the Old Greek does. This is the datum Michel pressed: ἑτερογλώσσοις is not the Septuagint’s word here, so Paul’s Greek looks like a different Greek — a Hebraizing one. And it is precisely the resemblance Origen certified from the manuscripts he had: that Paul’s form “lies in the translation of Aquila,” the most literal of the later Greek revisers. On this reading Paul’s Isaiah is not freely paraphrased but quoted — from a Greek already corrected toward the Hebrew.²

“Aquila-like” need not mean “after Paul”: the revision-tradition is older than the reviser. The historical Aquila is dated c. AD 130, three generations after Paul, so “Paul quotes Aquila” cannot be taken literally. But the same objection was dissolved for Paul’s Isaiah 25:8: the Hebraizing revision of the Greek Bible — the kaige stream that Aquila later carried to its extreme — is attested before Paul (the kaige Minor-Prophets scroll from Naḥal Ḥever is palaeographically pre-Christian). A Greek text of Isaiah already pulled toward the Hebrew could therefore stand behind both Paul (c. AD 54) and Aquila (c. AD 130) — a common Hebraizing tradition, not a borrowing across the century. Read this way, 1 Corinthians 14:21 is a second instance of the same larger truth the resurrection-taunt row shows: the New Testament catching the Greek Bible mid-revision.⁵

Adversarial assessment

What holds. The core of the catalog’s tag is beyond dispute and is a genuine, honest counterweight to Paul’s usual Septuagintal lean. Paul’s 1 Corinthians 14:21 does not reproduce the Old Greek — it reverses the Old Greek’s reversal, restoring God as the speaker and a judgment on Israel — and in that sense it “leans toward the Hebrew.” That much was seen by Origen and Jerome and is the settled modern verdict; the row belongs with Paul’s other MT-leaning citations, not with his Septuagintal ones. The strong form of the reading — that a distinct Hebraizing Greek stands behind the wording — is coherent, anciently attested, and has a real lexical anchor in ἑτερογλώσσοις; it is not a fringe conjecture.

What must be refined, so the “revised Greek” reading is not over-read.

1. Origen’s “it matches Aquila” shows resemblance, not dependence — and can cut the other way. Because Aquila post-dates Paul by three generations, Origen’s datum cannot mean Paul used Aquila. It admits three readings, and only one favours a fixed Vorlage: a common pre-Aquila Hebraizing tradition behind both (the steelman); or Paul and Aquila independently pulling the same Hebrew toward the same literal Greek (in which case there is no shared text, only shared method); or — the mirror of the caveat that dogs Isaiah 25:8’s uncial-Q gloss — a later assimilation of Aquila’s transmitted text to the famous Pauline verse, which would make the “match” partly circular. Origen’s testimony is a real weight, but it establishes that Paul’s Greek looks Aquiline, not that Paul copied a document.²

2. The changes that most look “Hebraizing” are exactly what Paul’s argument requires — and are attributed to Paul’s own hand. Paul is arguing that tongues are a sign of judgment on unbelief, so he needs God, not the heathen, as the speaker, and he needs the accent to fall on Israel’s refusal. Precisely those features are his departures from both text-forms: the first-person λαλήσω (“I will speak”), the emphatic οὐδ᾽ οὕτως (“not even so”) for the Hebrew’s bare “yet,” the appended μου (“hear me”), the future tense for the Hebrew’s past “they would not hear,” and above all the wholesale omission of the verse’s positive half (“This is the rest…”), irrelevant to his point. Conzelmann, who otherwise leans to a pre-existing text, concedes the first-person shift and the “even so” are Paul’s. Read as rhetoric, these are not text-critical evidence at all — they are what a preacher does with a verse.⁴

3. The citation matches no known text word-for-word — the lexical anchor is one word, and Michel’s inference from it is an argument from silence. This is where the row differs sharply from Isaiah 25:8, whose case rests on a composite agreement (direction, voice, and the rare word νῖκος) with two revisers independently. Here the only hard non-Septuagintal lexeme is ἑτερογλώσσοις, and no manuscript of a revised Greek Isaiah 28:11 survives to confirm it: Michel’s “must derive from Aquila’s Vorlage” over-reads a single word into a whole lost text, an inference from silence. Stanley’s point stands — the wording is far from the Old Greek and far from the Hebrew — which is easier to explain if Paul is rendering and adapting than if he is transcribing a fixed Vorlage.⁴

4. The proven claim is the text-form, not the mechanism — as with Matthew. What the evidence establishes is that Paul’s Isaiah 28 here is non-Septuagintal and Hebrew-leaning. Whether that arose from a Hebraizing Greek Vorlage, from Paul’s own rendering of the Hebrew, or from a free citation improved from memory by a man who knew the original, is exactly the undecided question that attends Matthew’s formula quotations. Agreement with the Hebrew is, here too, agreement with the literal revisers, so “Paul used a revised Greek” and “Paul translated the Hebrew” predict nearly the same wording and cannot be cleanly separated by the text alone.⁵

Assessment

Contested. Two questions have two answers. First: is Paul’s 1 Corinthians 14:21 a genuine non-Septuagintal, Hebrew-leaning citation — a real counterweight to his Septuagintal lean? Yes, securely — the Old Greek reverses the verse (the heathen the speakers, “contempt of lips”), and Paul restores the Hebrew’s direction (God the speaker, judgment on Israel), so the catalog’s “neither / free” tag holds and the row sits with Paul’s other MT-leaning texts (Job 5:13, Job 41:11). Second: does the wording specifically witness a revised Greek text (of the Aquila type) that Paul quoted, rather than his own free rendering of the Hebrew? Genuinely disputed, with no clear winner. For the revised-text reading: Origen’s ancient report that Paul’s form “lies in the translation of Aquila,” the non-Septuagintal ἑτερογλώσσοις, and the fact — established at Isaiah 25:8 — that a pre-Aquila Hebraizing revision demonstrably existed for Paul to draw on. Against it: the divergences that carry the “Hebraizing” impression (first person, “not even so,” “me,” the future, the omitted positive half) are exactly what Paul’s sign-of-judgment argument needs and are credited by Conzelmann to Paul himself; the citation matches no surviving text word-for-word; and the single non-LXX lexeme cannot bear Michel’s “must derive from Aquila’s Vorlage” without an argument from silence. So the reading is strong text-critically (Origen; one hard lexeme) and weak by the lights of citation-technique analysis (Stanley; Conzelmann’s Pauline redaction) — the signature of a contested case, unlike the tighter, composite match that makes Isaiah 25:8 sound. What to trust is the non-Septuagintal, Hebrew-leaning text-form; what to hold loosely is the specific claim that Paul copied a revised Greek Vorlage here — a possibility Origen and the non-LXX word keep live, but which the evidence does not settle against the equally live alternative that this is Paul’s own reading of the Hebrew, bent to his argument.

Sources