Verse 7 is the hinge of Paul’s Jerusalem report. Verse 6 ended with a flat negative — the men of reputation added nothing to him. Verse 7 swings to the positive with the strongest contrastive pairing Greek affords (ἀλλὰ τοὐναντίον, ‘but on the contrary’): far from supplementing Paul, they recognized what God had already done.

Three features carry the weight.

First, πεπίστευμαι is a perfect passive with a retained object: ‘I stand entrusted with the gospel.’ The agent is unnamed, which in Paul regularly means God. The entrusting is not something Jerusalem conferred; it is something Jerusalem saw.

Second, the two genitives τῆς ἀκροβυστίας and τῆς περιτομῆς are abstract nouns standing for bodies of people. The question is whether they mark two messages or one message with two spheres — and the Greek, being a bare genitive, does not say. Verse 8 glosses the same arrangement with εἰς + accusative (‘toward the nations’), which tilts strongly toward sphere.

Third, Paul writes Πέτρος here and in v. 8, and Κηφᾶς at 1:18 and 2:9, 11, 14. The switch within a single paragraph is a datum, and is one of the main reasons many read vv. 7-8 as Paul quoting the language of the Jerusalem agreement itself.

Grammatically the verse has no main verb. The participle ἰδόντες is left hanging and is finally picked up by ἔδωκαν in v. 9 — one of the more conspicuous broken constructions in Paul.

τοὐναντίον is crasis for τὸ ἐναντίον, the articular neuter adjective used adverbially — ‘the opposite thing,’ hence ‘on the contrary.’ It occurs three times in the NT, all with ἀλλά or δέ and all marking a sharp reversal of what was just denied (2 Cor 2:7; 1 Pet 3:9; Gal 2:7).

The combination ἀλλὰ τοὐναντίον is pleonastic by strict logic: ἀλλά already contrasts. The doubling is Koine reinforcement of the kind this corpus preserves elsewhere (πάλιν ἄνωθεν, Gal 4:9; ἔτι μᾶλλον καὶ μᾶλλον, Phil 1:9), and both members must be rendered. Compressing to a bare ‘but’ or ‘instead’ discards the emphasis.

Note that this is not the ascensive ἀλλὰ καί already fixed at Phil 1:18, which heightens rather than opposes. Here ἀλλά is doing genuine adversative work against the οὐδέν προσανέθεντο of v. 6, and ‘but’ is correct.

Built onbdag · bdf

A perfect passive of πιστεύω in the first person singular, with τὸ εὐαγγέλιον standing as a retained accusative — the object the verb would have governed in the active (πιστεύω τινί τι, ‘entrust something to someone’). When the dative person becomes the subject of the passive, the thing-accusative stays. The same construction recurs at Rom 3:2 (ἐπιστεύθησαν τὰ λόγια τοῦ θεοῦ), 1 Cor 9:17, 1 Thess 2:4, and 1 Tim 1:11.

Two things must survive into English.

The tense. The perfect names a completed act of entrusting whose result stands in force as Paul writes. ‘I was entrusted’ (a simple past) severs the standing result; ‘the gospel was committed unto me’ preserves the force but relocates the subject. English present perfect with Paul as subject — ‘I have been entrusted’ — keeps both.

The agent. Unexpressed. In Pauline usage the entrusting party in this formula is God (explicit at 1 Thess 2:4, δεδοκιμάσμεθα ὑπὸ τοῦ θεοῦ πιστευθῆναι τὸ εὐαγγέλιον; cf. 1 Tim 1:11, κατὰ τὸ εὐαγγέλιον … ὃ ἐπιστεύθην ἐγώ). Supplying ‘God entrusted me with’ is exegetically defensible but writes in what Paul withheld, and it is precisely the withholding that makes the sentence work: the Jerusalem leaders are the subject of ἰδόντες, not of the entrusting. They saw a fact already constituted.

This is also why ‘they saw that I had been given the task of preaching’ is a weaker rendering than it looks: it converts a deposit held in trust into an assignment handed out, and loses the commercial-fiduciary colouring of the πίστις root that Paul exploits elsewhere (παραθήκη, 1 Tim 6:20; 2 Tim 1:12, 14).

Built onbdag · bdf

Two parallel genitives, τῆς ἀκροβυστίας and τῆς περιτομῆς, each depending on εὐαγγέλιον (the second with the noun elided). Both nouns are abstract-for-concrete: ‘the uncircumcision’ = the uncircumcised as a body, ‘the circumcision’ = the circumcised. The idiom is Paul’s own (Rom 3:30; 4:9; Gal 2:12; Eph 2:11; Col 3:11), and the corpus policy on περιτομή already requires the abstract form be retained so the κατατομή / περιτομή wordplay of Phil 3:2-3 stays visible.

The genitive’s force is where the dispute lies.

English can hold a bare genitive, so the literal tier does: ‘the gospel of the uncircumcision.’ The reading text adopts ‘for the uncircumcision,’ which follows v. 8’s own gloss, with the alternative recorded here. What must be avoided in every tier is ‘the gospel to the Gentiles’ — it both resolves the genitive and substitutes ἔθνη (Paul’s word in v. 8, not here) for ἀκροβυστία, erasing a deliberate change of vocabulary between adjacent verses.

Built onbdf · bdag

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

Galatians uses the Aramaic Κηφᾶς at 1:18 and 2:9, 11, 14, and the Greek Πέτρος only here and in v. 8. In a letter where the author otherwise holds consistently to the Aramaic form, two Greek-form occurrences clustered in two consecutive verses are hard to attribute to chance.

The most common explanation is that vv. 7-8 incorporate language from the Jerusalem agreement itself — a formulation drafted in Greek, in which Peter was named by his Greek name and the two spheres were described with the abstract nouns ἀκροβυστία and περιτομή rather than with Paul’s own ἔθνη. On this view Paul is quoting the terms he was granted, which is rhetorically apt: his point is that the pillars’ own formula vindicates him.

Translation policy (already fixed in this corpus): transliterate each form as written, never substitute one for the other, and record the alternation in the parsing table. The man is identified in a note; he is not renamed in the rendering.

Built onna28 · metzger-tc

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

ἰδόντες is a nominative masculine plural aorist participle with no finite verb to attach to in v. 7. The sentence begun here is interrupted by the parenthetical explanation of v. 8 (ὁ γὰρ ἐνεργήσας …), resumed by a second participle in v. 9 (καὶ γνόντες), and only then given a main verb: δεξιὰς ἔδωκαν.

Three consequences for rendering.

1. No English sentence may be closed at the end of v. 7. The verse division is modern and falls inside a construction. Reading tiers should end with a dash.
2. No finite verb may be supplied unbracketed. ‘They saw that…’ is an acceptable reading-tier resolution of the participle only if the suspension is still signalled by punctuation; what is excluded is a self-standing sentence (‘On the contrary, they saw that I had been entrusted…’) which quietly repairs the break.
3. The subject must not be named. Greek supplies none; it is recoverable from οἱ δοκοῦντες in v. 6 and specified only in v. 9 (Ἰάκωβος καὶ Κηφᾶς καὶ Ἰωάννης). Inserting ‘James, Cephas and John’ here imports the identification Paul deferred.

The break itself is evidence of something. Paul is writing at speed and under pressure, and the sentence keeps acquiring qualifications it cannot carry — exactly as at Gal 1:20 and 2:4-6, each of which also breaks. Smoothing it reproduces the scribal and versional instinct rather than the text.

Built onbdf

Proposed — pending reviewJ. B. Lightfoot, Saint Paul's Epistle to the Galatians, London: Macmillan, 1865

The verse is stable in the Greek tradition. NA28 prints the text above without a variation unit of consequence.

Two minor points worth recording.

ἀλλὰ τοὐναντίον. A few witnesses write the uncontracted ἀλλὰ τὸ ἐναντίον. Crasis and its resolution are orthographic and carry no semantic weight; the contracted form is printed.

Πέτρος. No significant support for Κηφᾶς here, which is the reverse of the situation at 2:11 and 2:14, where a substantial body of Byzantine witnesses reads Πέτρος against the Κηφᾶς of the earliest manuscripts. The direction of scribal activity is consistently toward the more familiar Greek name, which strengthens the case for Κηφᾶς at 2:11, 14 and leaves Πέτρος here and in v. 8 as the harder, original distribution.

The Latin tradition reads sed e contra cum vidissent quod creditum est mihi evangelium praeputii sicut et Petro circumcisionis, resolving the participle into a temporal clause with cum and the subjunctive — a translational necessity, not a textual witness, and no warrant for repairing the Greek anacoluthon. Note also that the Latin’s creditum est mihi makes the gospel the subject and Paul the indirect object, reversing the Greek construction; this is the route by which ‘the gospel … was committed unto me’ entered the English tradition.

Built onna28 · metzger-tc

ἀκροβυστία is a biblical-Greek formation, the LXX’s standard equivalent of עָרְלָה (Gen 17:11, 14, 23-25), used in preference to the polite classical ἀκροποσθία. Its primary referent is the concrete foreskin; Paul extends it to an abstract status-term paired with περιτομή (Rom 2:25-27; 3:30; 4:9-12; Gal 2:7; 5:6; 6:15; Col 3:11).

The corpus policy, already fixed at Gal 5:6, is to render ‘uncircumcision’ in reading tiers so the abstract parallel with περιτομή survives, and to record the concrete sense and the coarse register in the notes rather than laundering either (§5: describe, don’t launder). ‘The uncircumcised’ shifts from status to persons and breaks the parallel with 6:15’s καινὴ κτίσις; ‘the Gentiles’ substitutes a different lexeme entirely — and here that substitution would additionally erase the fact that Paul writes ἀκροβυστία in v. 7 and ἔθνη in v. 8.

That shift between adjacent verses is itself worth noting. If vv. 7-8 do draw on the Jerusalem formulation, the move from ἀκροβυστία/περιτομή (the categories of Torah observance) to ἔθνη/περιτομή (ethnic-geographic categories) may mark the seam where Paul’s own idiom resumes.

Built onbdag · muraoka-lxx

The Fathers read the verse chiefly as a proof that Paul’s apostleship is of the same order as Peter’s, and they are alert to the recognition language.

Chrysostom presses ἰδόντες hard: the pillars did not teach Paul, they saw (εἶδον) — and what they saw was already complete. He treats the διαίρεσις of spheres as a division of labour sanctioned by God’s own prior allocation, not a concession negotiated between parties, and he is careful that the two genitives name audiences rather than two messages, since on his reading there is one gospel differently addressed.

Jerome, commenting on the Latin, notes the force of e contra: Paul expected to be examined and was instead acknowledged. He also registers the shift of name but without drawing the source-critical inference that modern scholarship draws from it.

The later Western reading of the verse as a charter for jurisdictional division between apostolic sees is reception, not exegesis, and supplies no evidence for the rendering (§11).

Proposed — pending reviewJohn Chrysostom, Commentary on the Epistle to the Galatians (Homilies on Galatians), in Nicene and Post-Nicene Fathers, First Series, vol. 13

The verse does two things at once that are easy to pull apart and should not be.

It asserts a single gospel with a divided mission field. Paul has already pronounced anathema on anyone proclaiming a gospel at variance with the one received (1:8-9); he cannot three chapters later be conceding two gospels. What is divided is the sphere, and v. 8 says so by switching to εἰς + accusative.

And it locates the ground of Paul’s standing outside Jerusalem’s gift. πεπίστευμαι is passive and perfect: the trust was constituted before the conference and stands independent of its outcome. The pillars’ role in the sentence is to see, and the main verb when it finally arrives in v. 9 is a handshake — recognition, not authorization. Whatever Galatians 2 is, it is not a report of Paul receiving a licence.

The cost of over-smoothing the syntax is theological, not merely stylistic: a tidy sentence with a named subject and a supplied main verb reads as the minutes of a meeting that conferred something. The broken Greek reads as a man insisting that nothing was conferred.