This is the hinge of Galatians, and arguably of Paul. The sentence that began at v. 15 — spoken still, formally, to Cephas at Antioch — turns from a shared Jewish premise to a shared Jewish conclusion. Three times in one verse Paul writes ἐξ ἔργων νόμου, and three times he negates it. Twice he writes πίστις with a genitive of Christ. The verse ends by grounding the whole in Psalm 143:2.

Four things carry the weight:

1. δικαιόω is forensic-declarative: God pronounces in the right. The passives leave him unnamed.
2. ἐξ ἔργων νόμου is anarthrous and marks a source or principle, not a list of achievements. What it names — the whole Torah-regime, or the boundary-markers that separated Jew from Gentile at the Antioch table — is the central dispute of the last half-century of Pauline scholarship, and it is not settled here.
3. πίστεως Ἰησοῦ Χριστοῦ is a bare genitive. It may mean faith directed toward Christ (objective) or Christ’s own faithfulness (subjective). Both readings are seriously held; this apparatus renders the bare genitive and refuses to decide.
4. ἐὰν μή is formally an exception-formula (“except through”), which on its face would make faith a qualification of law-works rather than their alternative. Paul’s own οὐκ … ἀλλά elsewhere pushes against that. The seam is real and is marked, not smoothed.

The middle clause is the surprise: having stated the principle, Paul does not argue for it — he reports that we too believed. The Jewish apostles’ own conversion is the evidence.

The sentence from v. 15 has no finite verb until ἐπιστεύσαμεν. εἰδότες is a nominative plural participle whose adverbial relation to that verb Greek leaves unmarked, exactly as with ὑπάρχων (Phil 2:6) and ὤν (Gal 4:1).

Three readings are live:

• Concessive: “Jews by nature … though we know …, we too believed.” This sets the knowledge against the ethnic privilege of v. 15.
• Causal: “because we know …, we believed.” The knowledge is the ground of the act.
• Circumstantial: simply attendant.

The δέ favors some contrast with v. 15, which is why the reading text uses “yet.” But the literal tier keeps the bare participle, per standing policy: where the relation is grammatically unmarked and exegetically consequential, English supplies the connective only in smooth registers and flags it.

Note also what εἰδότες claims. Per the corpus entry, οἶδα denotes settled knowledge, not a conclusion just reached — and the first person plural includes Cephas. Paul is not teaching Peter something; he is holding him to what Peter already knows.

Built onbdf · na28

Proposed — pending reviewH. D. Betz, Galatians: A Commentary on Paul's Letter to the Churches in Galatia, Hermeneia, Philadelphia: Fortress, 1979.

Three forms in one verse: present passive δικαιοῦται, aorist passive subjunctive δικαιωθῶμεν, future passive δικαιωθήσεται. One English root throughout, and one shared with δικαιοσύνη (“righteousness”) and δίκαιος.

The LXX usage is decisive for the sense. Rendering Hebrew צדק in the hiphil, δικαιόω regularly means “declare in the right” in a judicial setting: Exod 23:7 (“you shall not justify the wicked”), Deut 25:1 (the judges “justify the righteous and condemn the guilty”), Isa 5:23, Ps 142:2 LXX — the verse Paul cites here. The verb does not mean “make righteous” in these contexts; it names a verdict. The Latin iustificare (iustum facere) and the long Western argument over whether justification is declarative or transformative owe more to the Latin morphology than to the Greek.

All three forms here are passive with the agent unexpressed. This is a divine passive; God is the one who pronounces, and the rendering must not supply him. “God does not justify anyone on the basis of works of law” says more than the Greek.

δικαιοῦται is a gnomic present — what is characteristically the case, not what is currently in progress. Render with a simple English present, never a progressive.

Built onbdag · muraoka-lxx

Three occurrences in twenty-eight words, each anarthrous, each governed by ἐκ. The repetition is the rhetoric and must survive in English without variation.

ἐκ is source, kept distinct from διά (means) in the same sentence — Paul alternates them deliberately and the English must too. “By works of law” blurs the two.

νόμου is anarthrous and qualitative: law presented as a regime or principle rather than as the named Mosaic document. Per standing corpus policy the article is withheld in every tier even where English reads stiffly, because supplying “the law” narrows a categorical claim.

ἔργα νόμου is the contested phrase. Three positions:

Built onbdag · bdf

Proposed — pending reviewJ. D. G. Dunn, The Epistle to the Galatians, Black's New Testament Commentaries, London: A. & C. Black, 1993.

The verse contains the construction twice: διὰ πίστεως Ἰησοῦ Χριστοῦ and ἐκ πίστεως Χριστοῦ. Per standing corpus policy the bare genitive is retained in literal and reading tiers, because every fluent English rendering adopts one of two readings.

Two further data worth recording. First, Paul varies the formula within the sentence: full Ἰησοῦ Χριστοῦ at the first occurrence, bare Χριστοῦ at the second, with the accusative order reversed in between (Χριστὸν Ἰησοῦν). Second, Greek has no article on πίστεως at either point, so the English article in reading tiers is supplied for idiom and does not reflect a definite Greek form.

Built onbdag · bdf

Proposed — pending reviewR. B. Hays, The Faith of Jesus Christ: The Narrative Substructure of Galatians 3:1–4:11, 2nd ed., Grand Rapids: Eerdmans, 2002. · J. D. G. Dunn, The Theology of Paul the Apostle, Grand Rapids: Eerdmans, 1998.

ἐὰν μή is a fixed exceptive idiom, never a live conditional (standing corpus entry: “except / but only,” and never “if not”). The difficulty is what it excepts from.

Read strictly, the clause carves faith out of the preceding negation: a man is not justified out of works of law except through faith of Jesus Christ — which could be heard as “law-works do not justify unless accompanied by faith.” That is not what Paul argues anywhere, and the rest of the verse (καὶ οὐκ ἐξ ἔργων νόμου) excludes it.

The rendering keeps the bare “except,” which is as open in English as the Greek, and marks the seam with a dash rather than resolving it with “but only.”

Built onbdf · bdag

Proposed — pending reviewR. N. Longenecker, Galatians, Word Biblical Commentary 41, Dallas: Word, 1990.

Paul writes διὰ πίστεως in the first clause and ἐκ πίστεως in the third, within one sentence, while using ἐκ for ἔργων νόμου at all three points. Per standing corpus policy the prepositions must receive distinct English words wherever an author sets them in one argument: διά is intermediate means (“through”), ἐκ is source (“out of”).

Whether Paul intends a real distinction here is a separate question. Rom 3:30 sets ἐκ πίστεως and διὰ τῆς πίστεως side by side of the circumcised and the uncircumcised respectively with no evident difference, and most read the Galatians variation as stylistic. But the variation is data, and levelling both to “by faith” — the universal practice of English versions — hides it. The apparatus records the variation; the rendering preserves it without pressing it.

One consequence is worth noting: ἐκ πίστεως exactly matches ἐξ ἔργων νόμου in construction. The two phrases are set as competing sources, which is the sharpest form of the antithesis and is visible only if both receive “out of.”

Built onbdf · bdag

The main clause at last. Three marked elements, all of which must survive:

καί is adjunctive, attaching to the pronoun: “we also.” Per standing policy adjunctive καί before a pronoun is always rendered. Dropping it removes the claim that the Jewish apostles joined a company of believers already constituted — Gentiles among them — rather than founding one.

ἡμεῖς is an expressed nominative, redundant to the verb ending and therefore emphatic. It resumes the ἡμεῖς that opened v. 15 and closes the bracket: we, Jews by nature … we too believed.

ἐπιστεύσαμεν is a complexive aorist — the act viewed as a single completed whole, not a process. English simple past or present perfect; not “have come to believe,” which imports a process the aorist does not describe.

The rhetorical move is the thing to notice. Having stated the principle, Paul offers no argument for it in this clause. He offers a fact: men who held every advantage v. 15 names nevertheless believed. Their own conversion is the premise’s proof.

Built onbdf

πιστεύω εἰς + accusative is a distinctively early Christian construction with no real classical or LXX precedent; Greek normally takes πιστεύω with a dative of the person believed or ὅτι of the proposition. The directional preposition signals personal committal rather than propositional assent.

Per standing corpus policy the wooden tier keeps “believe into” so the idiom stays visible, and “believe in” is used in smooth registers. Keep distinct from πιστεύω + dative (“credit someone’s word,” Jas 2:23 / Gen 15:6) and πιστεύω ὅτι (“believe that,” Jas 2:19), both of which Paul and James use elsewhere.

Note the word order: Χριστὸν Ἰησοῦν in the accusative, against Ἰησοῦ Χριστοῦ in the genitive earlier in the same verse. Per standing policy the order is retained as written at each point rather than normalized.

Built onbdag · bdf

Telic ἵνα with the aorist passive subjunctive: the purpose of the believing. The aorist is complexive, not iterative.

The clause is worth pausing over because of what it does to the sequence. Paul says they believed in order that they might be justified — which is to say, the verdict is sought through the believing and not through the law-observance they already possessed. The purpose clause thereby makes explicit what the opening participle only asserted.

Per standing policy, keep the passive passive with the agent unexpressed, and render ἵνα “so that,” which preserves the telic force without over-specifying.

Built onbdf

The final clause is not Paul’s own formulation but a reworking of Ps 142:2 LXX (= Ps 143:2 MT):

ὅτι οὐ δικαιωθήσεται ἐνώπιόν σου πᾶς ζῶν “for no one living will be justified before you”

Hebrew: כִּי לֹא־יִצְדַּק לְפָנֶיךָ כָל־חָי.

Paul’s changes are three and all three are deliberate:

• ἐνώπιόν σου (“before you,” second person, addressed to God in prayer) is dropped. The psalm is a petition; Paul makes it a principle.
• πᾶς ζῶν (“every living one”) becomes πᾶσα σάρξ. σάρξ is the LXX equivalent of בָּשָׂר and carries the creature’s weakness and mortality; it also keys into the σάρξ-thread of Galatians (3:3; 4:23, 29; 5:16-24; 6:12-13).
• ἐξ ἔργων νόμου is added, and fronted before the verb. This is the interpretive move: the psalm’s universal statement about the creature before God is specified as a statement about law-works.

Paraphrased the same way at Rom 3:20, which confirms the citation was a fixed element of Paul’s argument rather than an ad hoc allusion.

οὐ … πᾶσα is a Semitism. In Greek the sequence would strictly mean “not all flesh” (i.e. some but not all); the Hebrew לֹא … כָל behind it means “no … at all.” Render “no flesh,” and note the idiom rather than letting the Greek word order mislead.

No citation formula is used — no γέγραπται, no λέγει ἡ γραφή. The scriptural wording simply carries the sentence’s conclusion, which is itself evidence of how thoroughly the psalm’s language had become Paul’s own.

Built onmuraoka-lxx · bdf

The verse opens with an anarthrous generic ἄνθρωπος and closes with πᾶσα σάρξ, and the two frame the argument. The particular case (a person, any person, in the matter under dispute) widens to the whole human creature before God.

ἄνθρωπος: per §3 and standing corpus policy, masculine grammatical gender here is the Greek generic and the referent is inclusive. Reading tiers use “a person”; the literal tier keeps “a man” to track the form, with the generic force noted.

σάρξ: one English gloss across all registers, per standing policy. Here the sense is neutral-anthropological — the human being considered as creature, mortal and weak before God — not the polemical σάρξ of Gal 5:16-24. The register difference is carried in the note, never by changing the word. “No one” and “no human being” are excluded from the reading text: they delete a noun whose recurrence through Galatians is part of the letter’s argument.

Built onbdag · muraoka-lxx

The verse is textually stable in its substance; the variation is almost entirely a matter of divine-name expansion and preposition-levelling, and runs in a single direction — toward fuller and more uniform formulae.

• ἐὰν μή: read by the great majority. A few witnesses read εἰ μή, which is the commoner exceptive form and the easier reading; ἐὰν μή is to be retained.
• διὰ πίστεως Ἰησοῦ Χριστοῦ: the first formula. Some witnesses read Χριστοῦ Ἰησοῦ, and a few omit one name. The variation reflects the general scribal fluidity of the double name rather than any exegetical pressure; render the critical text as written.
• ἐκ πίστεως Χριστοῦ: the second and shorter formula. Part of the tradition expands to Ἰησοῦ Χριστοῦ, harmonizing it to the first. The shorter reading is both better attested and the harder one — scribes expand divine names, they do not contract them — and is retained.
• καὶ ἡμεῖς: some witnesses read ἡμεῖς δέ or omit the pronoun. The adjunctive καί is the harder reading and the one the argument requires.

None of the variation touches the verse’s three negations or the Psalm citation. What the apparatus does show is a scribal tradition smoothing a deliberately varied sentence toward uniformity — exactly the direction a translator must avoid.

Built onna28 · metzger-tc

The patristic handling is instructive mainly for what it does not assume.

Chrysostom (Hom. in Gal. 2) reads ἐξ ἔργων νόμου against the background of the Antioch dispute and presses the rhetorical point of καὶ ἡμεῖς: if those who had the law believed, the law plainly did not suffice. He takes the faith-genitive as objective without discussion, which is the reading the Greek Fathers generally assume.

Jerome (Comm. in Gal.) renders nisi per fidem Iesu Christi and, like the Latin tradition generally, leaves the genitive as ambiguous in Latin as it is in Greek — which is why the subjective/objective question did not become a controversy until modern English and German translations were forced to choose.

Augustine, especially in the anti-Pelagian works, reads opera legis as human works in general and iustificare as iustum facere. The Latin morphology of iustificare does real work in his argument that the Greek δικαιόω does not support; this is the clearest instance in the verse of a doctrinal position resting on a translation.

These readings are reception and are recorded as such (§11). None of them settles a lexical question.

Proposed — pending reviewJohn Chrysostom, Commentary on the Epistle to the Galatians, in Nicene and Post-Nicene Fathers, First Series, vol. 13, ed. P. Schaff, Buffalo: Christian Literature Co., 1889. · Jerome, Commentarii in Epistulam Pauli ad Galatas, Corpus Christianorum Series Latina 77A, Turnhout: Brepols, 2006.