Paul has just said two awkward things in a row: the law was added because of transgressions (v. 19) and it came through a mediator, which the promise did not (vv. 19-20). The obvious objection follows at once, and he puts it himself: if the law was added later, through intermediaries, on account of transgressions — is it then working against God’s promises?

The answer is the strongest negative in his vocabulary, and then a reason. The reason is a counterfactual: if a law had been given with power to make alive, righteousness really would have come out of law. The grammar of the sentence (second-class condition, ἄν in the apodosis) states that the premise did not hold. No such law was given. Therefore righteousness does not come out of law — and therefore the law is not a competitor to the promise at all, because it was never in that line of work.

The force of the verse depends on noticing what Paul does not say. He does not say the law is bad, or that it failed at its job. He says it was never given the capacity that would have made it a rival. Three features carry the argument: the adversarial κατά + genitive; the contrary-to-fact condition; and the verb ζῳοποιῆσαι, which belongs to God and the Spirit, never to a statute.

Proposed — pending reviewF. F. Bruce, The Epistle to the Galatians: A Commentary on the Greek Text, New International Greek Testament Commentary (Grand Rapids: Eerdmans, 1982)

Ὁ οὖν νόμος κατὰ τῶν ἐπαγγελιῶν τοῦ θεοῦ; has no verb. A copula is supplied in English and bracketed in the literal tier, per the corpus policy on verbless clauses.

Two further points must be reported rather than smoothed:

1. The question mark is editorial. The earliest majuscules are unpunctuated, and nothing in the Greek marks this as interrogative except discourse structure — the inferential οὖν, the abrupt verblessness, and the μὴ γένοιτο that answers it. The μὴ γένοιτο is decisive: Paul never uses the formula except to repudiate a question or inference just put (Rom 3:4, 6, 31; 6:2, 15; 7:7, 13; 9:14; 11:1, 11; Gal 2:17; 3:21). But the reader should know the punctuation rests on that argument, not on a manuscript.
2. The question is framed with neither οὐ (expecting yes) nor μή (expecting no). Paul puts it bare. That is characteristic of the diatribe at the point where an objection is genuinely dangerous and the author wants it fully stated before he answers it.

The inferential οὖν is load-bearing and must be rendered (§10): it ties the objection specifically to the concession just made in vv. 19-20, rather than floating it as a general question about law.

Proposed — pending reviewHans Dieter Betz, Galatians: A Commentary on Paul's Letter to the Churches in Galatia, Hermeneia (Philadelphia: Fortress, 1979)

κατά with the genitive is adversarial — drawn up against, in opposition to (BDAG κατά A.2.b; Matt 12:30; Acts 19:16; Gal 5:17, 23). This must stay sharply distinct in English from the several κατά + accusative uses already fixed in this corpus (norm, credential-list, concrete target, quantity-noun, personal pattern). Galatians switches between the two constructions within a few verses — κατὰ σάρκα at 4:23, 29 is normative; κατὰ τῶν ἐπαγγελιῶν here is hostile — and the English must switch with it.

The question is therefore not whether the law differs from the promises or is additional to them, but whether it is ranged against them. Renderings like “contrary to the promises” are acceptable; “inconsistent with the promises” is not, since it reduces a military-forensic opposition to a logical mismatch.

Reaffirming the standing entry: optative of wishing with μή, the strongest repudiation Greek has, and the LXX equivalent of Hebrew ḥālîlâ (Gen 44:7, 17; Josh 22:29; 1 Kgdms 20:2). Fourteen of fifteen NT occurrences are Pauline, ten of them in Romans.

The rendering “May it never happen!” keeps the optative mood visible. “God forbid” supplies a divine subject the Greek does not have and is confined to declared liturgical register (§11). “Certainly not” and “by no means” drop the wish-mood entirely and flatten the strongest thing Paul can say into an ordinary denial.

Note what the formula rejects. It does not reject the idea that the law was added, or came through a mediator, or concerned transgressions — Paul has just asserted all three. It rejects only the inference that these facts put law and promise in competition.

εἰ … ἐδόθη (aorist indicative) with ἄν … ἦν in the apodosis is a second-class condition (BDF §360). The form itself asserts that the protasis does not hold: no such law was given.

Three consequences for the rendering:

• English requires “if … had been given … would have been.” A present-tense rendering (“if a law is able to give life”) converts a counterfactual into an open condition and destroys the argument.
• ἄν is indispensable and must not be dropped; where witnesses omit it (as happens elsewhere with unreal apodoses in Koine, cf. Gal 4:15), they are simplifying.
• The mixed tenses are normal: an aorist in the protasis with an imperfect plus ἄν in the apodosis gives a past unreal protasis and a present-or-general unreal apodosis — “if a law had been given, righteousness would (as things now stand) be out of law.”

The conditional is what makes the verse an argument rather than an assertion. Paul does not say “the law cannot give life.” He says that had it been able to, the conclusion his opponents want would genuinely follow — and the form of the sentence states that it was not.

Proposed — pending reviewRichard N. Longenecker, Galatians, Word Biblical Commentary 41 (Dallas: Word, 1990)

A divine passive with the agent unexpressed. This belongs to a chain running through the chapter: προσετέθη (v. 19, “was added”), διαταγείς (v. 19, “ordained”), ἐδόθη here, δοθῇ (v. 22, of the promise given to those who believe). Paul consistently declines to name the giver of the law in this passage, having named him freely as the giver of the promise (v. 18, κεχάρισται ὁ θεός).

The reticence must be preserved. Renderings supplying “if God had given a law” write in what the passive withholds, and at precisely the point where the relation of God to the law is under pressure (vv. 19-20). The note may observe that Paul nowhere denies divine origin — Rom 7:12 settles it — but the rendering must not supply it here.

Note also that the subject νόμος is anarthrous: a law of such-and-such a kind, not the law. The question is about a category, not about the Torah’s record.

νόμος ὁ δυνάμενος ζῳοποιῆσαι — an anarthrous noun followed by an articular participle. This is the construction already fixed in the corpus for the doubled article (Phil 4:17; Jas 1:9; Gal 2:9), applied where the head noun carries no article of its own. The article makes the participle a defining modifier: a law, namely the kind able to make alive.

Rendering with an English relative clause (“a law that was able to give life”) is permitted in reading tiers; the literal tier retains “a law, the [one] able to make alive,” so the construction stays visible. What must be avoided is a prenominal adjective (“a life-giving law”), which both loses the verbal aspect and makes the capacity sound like a property of law as such rather than a specification of a hypothetical kind.

Compound of ζῳός + ποιέω: make alive, give life. Eleven NT occurrences, and in every one the agent is God, the Spirit, or Christ — raising the dead (John 5:21 twice; Rom 4:17; 1 Cor 15:22, 36, 45; 2 Cor 3:6; 1 Pet 3:18). The LXX usage is the same: God kills and makes alive (Deut 32:39; 1 Kgdms 2:6; 4 Kgdms 5:7; Neh 9:6).

This distribution is the point of the verse. The verb Paul chooses to name what law cannot do is a verb that in biblical Greek belongs to God alone. To ask whether a statute could make alive is already to have answered.

The ζωή root must stay audible in English: “make alive” or “give life,” never “impart spiritual life” (adds an adjective the Greek withholds), never “save” (that is σῴζω, a distinct lexeme already fixed), never “bring righteousness” (imports the apodosis into the protasis and makes the sentence circular).

Note also the deliberate contact with 2 Cor 3:6, where γράμμα kills and τὸ πνεῦμα ζῳοποιεῖ — and with Gal 3:2-5, where the Spirit is the thing the Galatians received not out of works of law.

Proposed — pending reviewHenry George Liddell, Robert Scott, and Henry Stuart Jones, A Greek-English Lexicon, 9th ed. with revised supplement (Oxford: Clarendon, 1996)

Possessed capacity, not exercised action (BDAG; Luke 14:31; Rom 4:21; Titus 1:9; Jas 4:12). The claim is not that a law failed to give life, but that a law with the power to do so was never given. The distinction is exegetically consequential: on the first reading law is a failed instrument, on the second it is an instrument never designed for the purpose.

Keep lexically distinct from ἰσχύω (“have strength for,” already fixed) and δυνατός + infinitive (“able, competent to”), both of which occur in this corpus. The aorist infinitive is complexive — the life-giving viewed as one whole act, not as an ongoing supply.

Adverb formed from the participle of εἰμί: really, actually, as a matter of fact (Luke 23:47; 24:34; 1 Cor 14:25; 1 Tim 5:3, 5, 16; 6:19). It asserts correspondence to reality.

Its position in a contrary-to-fact apodosis is pointed. Paul is conceding that the inference would genuinely hold — the opponents’ logic is not faulty, only their premise. “Really” and “truly” both carry this; “indeed” is weaker and reads as a mere intensifier; “certainly” tips toward assertion of the conclusion itself, which the conditional denies.

The adverb is fronted before the prepositional phrase, which keeps the emphasis on the genuineness of the entailment rather than on the source. Preserve the position in English where possible.

ἐκ marks source, per the standing corpus entry. Galatians 3 sets a series of ἐκ-phrases against each other as the spine of its argument: ἐξ ἔργων νόμου (vv. 2, 5, 10), ἐξ ἀκοῆς πίστεως (vv. 2, 5), ἐκ πίστεως (vv. 7, 8, 9, 11, 12, 22, 24), ἐν νόμῳ (v. 11), ἐκ νόμου (vv. 18, 21), ἐξ ἐπαγγελίας (v. 18). One English preposition — “out of” in the literal tier — must serve every occurrence, or a deliberate refrain becomes invisible.

νόμου is anarthrous and qualitative, so the article is withheld in English even where it reads stiffly: law as a mode or principle of righteousness, not the Mosaic document named. Supplying “the law” narrows a categorical claim.

Note the near-identical phrasing at v. 18: εἰ γὰρ ἐκ νόμου ἡ κληρονομία. Verses 18 and 21 are the same move made twice, once about inheritance and once about righteousness.

The noun occurs three times in this verse, articular once and anarthrous twice:

• ὁ νόμος (opening question) — anaphoric, the law under discussion since v. 19.
• νόμος (protasis) — anarthrous and qualitative, a law of a specified kind.
• ἐκ νόμου (apodosis) — anarthrous, law as source-principle.

The alternation must be tracked rather than levelled. Reading all three as “the law” (the near-universal English practice, and the only option available to the Latin, which has no article) converts a graded argument into a flat series of statements about the Torah. Reading all three as anarthrous loses the anaphora of the opening question.

This extends the standing anarthrous-νόμος policy (ἐκ νόμου, ὑπὸ νόμον, ἐν νόμῳ, οὐκ ἔστιν νόμος) by making explicit that the policy is applied per occurrence, not per verse.

The article is anaphoric: the righteousness that has been the subject since 2:21 (εἰ γὰρ διὰ νόμου δικαιοσύνη, ἄρα Χριστὸς δωρεὰν ἀπέθανεν) and through 3:6 (Abraham’s reckoned righteousness), 3:11 (ὁ δίκαιος ἐκ πίστεως ζήσεται), and the δικαιόω chain of vv. 8, 11, 24.

Retain the article: “the righteousness” in the literal tier, “righteousness” permitted in reading tiers with the compression noted. Per the standing entry, one English root across δίκαιος, δικαιοσύνη, δικαιόω, and καταδικάζω so the family stays traceable; “justification” is excluded as base gloss because it would sever the noun from δίκαιος at v. 11.

Note the logical link to 2:21 and to the ζωή vocabulary: righteousness and life are coupled in v. 11’s Hab 2:4 citation (ὁ δίκαιος ἐκ πίστεως ζήσεται) and in v. 12’s Lev 18:5 (ὁ ποιήσας αὐτὰ ζήσεται ἐν αὐτοῖς). What the law promises in Lev 18:5 is precisely life on condition of doing — and v. 21 says that no law was given with the power to deliver it.

Three points of variation, none affecting the sense materially.

τοῦ θεοῦ after τῶν ἐπαγγελιῶν is read by the great majority of witnesses and is printed without brackets in NA28. A handful of witnesses omit, and the omission is easily explained as scribal haplography or abbreviation; the phrase is also Pauline habit.

ζῳοποιῆσαι is the reading of the critical text. Some witnesses read the present ζῳοποιεῖν, levelling the complexive aorist toward a durative; the aorist is the harder reading and is to be retained.

ἄν is present in the apodosis in the main tradition, and its position (ὄντως ἐκ νόμου ἂν ἦν) varies slightly among witnesses. Where any witness omits it, that is simplification of an unreal apodosis and must not be followed — Koine does occasionally drop ἄν (cf. Gal 4:15), but the scribal direction is toward removal of difficulty, not creation of it.

None of these touches the shape of the argument.

The counterfactual is sharpened by what the LXX does with ζῳοποιέω and its near relatives.

Deut 32:39 (LXX): ἐγὼ ἀποκτενῶ καὶ ζῆν ποιήσω — “I will kill and I will make alive.” The verb-phrase is the divine prerogative stated as the ground of God’s uniqueness.

1 Kgdms 2:6 (= 1 Sam 2:6): κύριος θανατοῖ καὶ ζωογονεῖ, κατάγει εἰς ᾅδου καὶ ἀνάγει.

4 Kgdms 5:7 (= 2 Kgs 5:7): the king of Israel tears his clothes at being asked to heal Naaman — μὴ θεὸς ἐγώ εἰμι τοῦ θανατῶσαι καὶ ζωοποιῆσαι? The very verb, in a rhetorical question whose answer is obvious.

Neh 9:6: σὺ ζωοποιεῖς τὰ πάντα.

Against this background, Gal 3:21’s protasis is not a cautious concession but something close to an absurdity entertained for the sake of argument. A law that could ζῳοποιῆσαι would have to be God.

The other LXX text in play is Lev 18:5, cited two verses earlier (v. 12): ὁ ποιήσας αὐτὰ ζήσεται ἐν αὐτοῖς. The law does speak of life. Paul’s counterfactual concerns not what the law says but what it was given power to effect.

Chrysostom (Hom. in Gal. 3) reads the verse as removing a competition the objector invented: the law and the promise aim at the same end, and the law’s inability to give life is not a defect in it but the reason a further provision was made. He is careful that Paul does not disparage the law — ὁ νόμος οὐκ ἐναντίος, ἀλλὰ ἀσθενέστερος.

Jerome (Comm. in Gal.) presses the counterfactual grammatically, noting that the form of the condition is itself the denial: si data esset lex says that none was.

Augustine reads v. 21 with Rom 8:3 and takes the inability as located in the flesh rather than in the law: the law is holy, but it commands what the unaided will cannot perform. His reading made this verse a standard text in the Pelagian controversy. It is a legitimate construal but it is a synthesis with Romans, not an exegesis of Galatians, where the explanation offered is the law’s commission (v. 19) rather than human weakness.

Luther (Lectures on Galatians, 1535) makes the verse the hinge of the law/gospel distinction: law was never given ad vivificandum but ad ostendendum peccatum, and the whole error of his opponents is to ask of the law an office it never held.

The Reformation reading and the Augustinian reading diverge over the same question flagged in the contested block above — whether the law’s inability is a weakness or a job description — and the verse does not settle it.

Proposed — pending reviewJohn Chrysostom, Commentary on the Epistle to the Galatians, in Nicene and Post-Nicene Fathers, First Series, vol. 13 (Edinburgh: T&T Clark, repr. 1994) · Jerome, Commentarii in Epistulam Pauli ad Galatas, trans. Andrew Cain, Fathers of the Church 121 (Washington, DC: Catholic University of America Press, 2010) · Martin Luther, Lectures on Galatians (1535), Luther's Works vols. 26-27, ed. Jaroslav Pelikan (St. Louis: Concordia, 1963-64)

The structural claim of the verse is easy to state and easy to lose. Paul is not saying that law and promise are two routes to the same destination, of which one works and one does not. He is saying that only one of them was ever a route.

That is why the question can be answered with μὴ γένοιτο rather than with a concession. A thing is κατά — ranged against — another only if the two are contending for the same ground. The law was given for the sake of transgressions, until the seed should come (v. 19). It is not a failed promise; it is not a promise at all. The whole force of vv. 15-18 is that the promise is a διαθήκη ratified four hundred and thirty years before the law existed, and a later addition cannot annul it.

What follows in v. 22 is the positive account: Scripture shut up everything under sin, precisely in order that the promise might be given out of faith to those who believe. The law’s negative function serves the promise rather than opposing it. The counterfactual of v. 21 clears the ground for that by establishing what the law was never commissioned to do.

Two guardrails follow for the rendering. First, nothing in the English may make the law sound defective in itself — Paul does not say so here and says the opposite at Rom 7:12. Second, nothing may soften the counterfactual into a possibility: righteousness is not hard to obtain out of law, it is not available out of law, and the grammar says so.

Proposed — pending reviewJ. Louis Martyn, Galatians: A New Translation with Introduction and Commentary, Anchor Bible 33A (New York: Doubleday, 1997)