One sentence, built entirely out of σύν- compounds and the verb-root ὑποκρίνομαι. Peter has withdrawn from the mixed table (v. 12); v. 13 reports the contagion. Three movements, each marked by a καί:

1. συνυπεκρίθησαν αὐτῷ — the rest of the Jews at Antioch did the same thing with Peter. The verb is an NT hapax and a triple compound, built so that the σύν- carries the solidarity and the ὑποκρ- carries the verdict.
2. καὶ οἱ λοιποὶ Ἰουδαῖοι — a bounded remainder: all the Jewish believers at Antioch other than Peter (and Paul).
3. ὥστε καὶ Βαρναβᾶς συναπήχθη — a result clause with a finite indicative, i.e. an achieved outcome, and the ascensive καί lands the sentence on the one name Paul least wanted to write: his own companion of v. 1 and his partner in the right hand of fellowship (v. 9).

Two lexical decisions carry most of the weight. ὑπόκρισις is the stage-actor’s word, and Paul has chosen it rather than ψεῦδος (lie) or ἀπάτη (deception): the charge is not that Peter believed something false but that his conduct did not correspond to his convictions (cf. v. 14, where Paul says Peter lives ἐθνικῶς). And συναπήχθη is a passive of being swept along — Barnabas is not charged with initiating anything, and the passive leaves the agent of the carrying-off formally unexpressed.

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

NT hapax (Gal 2:13), absent from the LXX; attested in Hellenistic prose (Polybius uses it of joining in a feigned performance). A triple compound: σύν + ὑπό + κρίνομαι.

The base verb ὑποκρίνομαι means to answer from beneath the mask — hence to play a part on stage, and so to act a part in life. The ὑπό- element is lexicalized and spent by the first century; it is not an independent morpheme carrying ‘under’ in any felt sense. The σύν-, by contrast, is fully live: it is answered by the dative αὐτῷ and names the others’ joining of Peter’s performance.

Two rules follow for the rendering:

• The σύν- must be audible, as with every other σύν- compound in this corpus (συνεργός, σύμψυχος, συγκοινωνός, συμμιμητής, συνεκλεκτός, συμπρεσβύτερος, and συναπάγω two clauses later). ‘They were hypocritical too’ deletes it and converts a charge of corporate solidarity in a performance into a parallel moral failing.
• The drama-root must be audible, because Paul resumes it four words later with the cognate noun ὑποκρίσει. The clause is built from one root twice, and if the verb and noun receive unrelated English words (‘acted hypocritically … their insincerity’) the figure disappears.

Excluded from the reading text: ‘were hypocrites’ (converts a verb of performance into a character-designation), ‘pretended along with him’ as sole rendering in the literal tier (loses the κρίν-/acting root that the noun needs), ‘dissembled’ (opaque in modern English; retained in declared liturgical register, following the Vulgate’s simulationi).

The aorist is complexive: the episode taken as one whole, not a series of occasions.

From ὑποκρίνομαι; properly the actor’s performance, then any conduct that presents what is not the case. The NT usage is uniformly pejorative (Matt 23:28; Mark 12:15; Luke 12:1; 1 Tim 4:2; 1 Pet 2:1), and the corpus has already fixed the privative ἀνυπόκριτος as ‘without hypocrisy / unhypocritical’, keeping the mask-image recoverable.

What the word does not mean here is conscious duplicity of belief. Paul’s charge in v. 14 is precisely that Peter does live ἐθνικῶς — his convictions are not in question. What is wrong is that the withdrawal from the table enacted a position Peter did not hold, and so misrepresented the gospel by performance. The LXX background is thin (2 Macc 6:25 of Eleazar refusing to feign; Sir 1:29; 32:15 of the double-tongued), but it points the same way: the fault is in the presentation, not in the mind.

Rendering: ‘play-acting’ in the literal and study tiers, where the stage must be visible and where the cognate link to συνυπεκρίθησαν must hold; ‘pretense’ in the reading text, which carries the same root-link to a verb of pretending; ‘dissimulation’ in declared liturgical register.

σύν + ἀπό + ἄγω: to lead away together with. Three NT occurrences, all passive and all of being carried off by something: here, 2 Pet 3:17 (συναπαχθέντες τῇ τῶν ἀθέσμων πλάνῃ), and Rom 12:16, where the construction is different and the sense disputed.

Every morpheme is doing work. The ἀπό- marks removal from the position previously held; the σύν- marks the company one goes with; and ἄγω supplies the image of being led, which is why the verb takes a dative of the force that does the leading. The physical picture is of someone caught up in a departing crowd.

Rendering: ‘was carried off along with’ in the literal tier, where all three morphemes must register; ‘was swept along with’ in reading tiers. Excluded: ‘was led astray’ (that is πλανάω, already fixed with the road-image, and 2 Pet 3:17 shows Greek can supply πλάνη separately when it wants that sense), ‘was carried away’ alone (deletes the σύν-), ‘joined them’ (converts a passive into an act of will — precisely the distinction the verse maintains between Barnabas and the others).

The passive must stay passive. Paul’s syntax distinguishes the rest of the Jews, who are the subject of an active-in-sense verb (συνυπεκρίθησαν, aorist passive in form but with active sense — they did the acting), from Barnabas, who is the subject of a true passive. Whether Paul intends the distinction as mitigation of Barnabas or as an aggravation (the very man who should have stood was moved) is not settled by the grammar, and renderings that make Barnabas the agent (‘even Barnabas went along with them’) close the question.

Two decisions here, both consequential.

λοιπός (already fixed in the corpus as ‘the rest / the remaining ones’, with the substantival-plus-genitive construction at Phil 4:3) marks a bounded remainder relative to a set already named. The set is the Jewish believers at Antioch; the one subtracted is Cephas, named in v. 11. ‘The other Jews’ blurs the remainder-logic; ‘all the other Jews’ over-extends it.

Ἰουδαῖοι is the §5 flashpoint. The referent here is beyond serious doubt: Jewish members of the Antiochene congregation, Paul’s own co-religionists and fellow believers — the group he includes himself in at v. 15 (ἡμεῖς φύσει Ἰουδαῖοι). The term is ethnic-civic, not a designation of unbelief, and it carries no polemical freight of its own.

Keep Ἰουδαῖος lexically distinct from Ἰσραήλ / Ἰσραηλίτης (already fixed): Paul’s choice between the two is deliberate, and Ἰουδαῖος is the civic-ethnic designation he uses of himself and Peter in vv. 14-15.

The verse contains three instances of καί and all three must be rendered (§10).

1. Clause-initial καί — continuative, advancing the narrative from v. 12. Render ‘and’.
2. καὶ οἱ λοιποὶ Ἰουδαῖοι — adjunctive, standing immediately before the subject. Per the standing entry on adjunctive καί before a clause subject, this adds a further party who does the same thing, and must be rendered ‘also’. Reducing it to a bare connective destroys the claim that a second group performed Peter’s act.
3. ὥστε καὶ Βαρναβᾶς — ascensive, ‘even’. This is the rhetorical point of the whole verse: the escalation to the one name that costs Paul most to write. Dropping it leaves a bare report of an additional person and erases the climax.

The distribution is deliberate: Paul adds a group (καί 2), then adds beyond the group one man (καί 3). English must keep ‘also’ and ‘even’ distinct, or the two-stage escalation collapses into one.

ὥστε with a finite indicative, not with an infinitive. The corpus has already fixed ὥστε + infinitive as ‘so that / with the result that’, noting that Koine has largely lost the classical conceived/actual distinction (BDF §391).

With the indicative, however, the construction is unambiguous even in Koine: it reports an actual, achieved result as an independent assertion of fact. Paul is not describing what the hypocrisy was apt to produce; he is stating what it did produce. Two consequences:

• The connective must be rendered. A bare ‘and even Barnabas was swept along’ reduces a stated consequence to a coordinate item.
• No modal hedging is permitted: ‘so that even Barnabas was in danger of being carried away’ softens a plain indicative, which §6 forbids.

‘So that’ is retained in reading tiers; ‘with the result that’ is available where the result-force needs to be explicit but is clumsy in narrative register.

The bare dative completing συναπήχθη is genuinely open between two classifications, and the choice bears on how much Barnabas is charged with.

English ‘by their pretense’ leans instrumental; ‘in their pretense’ leans circumstantial. The reading text adopts ‘by’, on the strength of the 2 Pet 3:17 parallel, with the alternative flagged. The literal tier keeps the bare dative visible with a bracketed supplement for the σύν- (‘along [with them] by their play-acting’) so that neither classification is foreclosed.

Agency proper (ὑπό + gen) is not in play: Paul had that construction available and did not use it.

The possessive genitive stands before the article-noun unit (αὐτῶν τῇ ὑποκρίσει) rather than in the unmarked position after the noun (τῇ ὑποκρίσει αὐτῶν). The corpus has already fixed this as a marked, emphatic construction (τῆς ὑμῶν δεήσεως, Phil 1:19; τῆς ἐμῆς παρουσίας, Phil 1:26; τῶν αὐτοῦ κτισμάτων, Jas 1:18).

The force here is to specify whose performance it was — theirs, not his own. Barnabas contributed nothing to the pretense; he was moved by somebody else’s. English has no morphological equivalent, so the emphasis is carried by placement and stress (‘by their pretense’) and recorded here; per the standing entry, ‘their own’ overtranslates and is excluded.

Note also that αὐτῶν is plural while the preceding dative αὐτῷ is singular. The pretense of the clause’s end is attributed to the group of v. 13a, not to Peter alone — Peter is the one they acted with, but by the time Barnabas is swept off it is a corporate performance. Rendering both pronouns with a single English referent erases the widening.

Galatians 2 runs a sustained pattern of σύν- compounds, and v. 13 has two of them in eleven words:

• συμπαραλαβών (v. 1) — Paul took Titus along with him
• ὁ σὺν ἐμοί (v. 3) — Titus, the one with me
• συνυπεκρίθησαν (v. 13) — they played the part along with him
• συναπήχθη (v. 13) — Barnabas was carried off along with them
• συνεσταύρωμαι (v. 20) — I have been crucified together with Christ

The pattern is not decorative. Chapter 2 is about who is with whom: Paul with Titus and Barnabas going up to Jerusalem, the pillars giving the right hand of κοινωνία (v. 9), then the table-fellowship at Antioch fracturing into a different kind of togetherness, and finally the one togetherness that matters (v. 20). The two compounds in v. 13 are the low point of the sequence — solidarity in the wrong direction — and English must keep the ‘with’ audible at every point or the arc from v. 1 to v. 20 becomes invisible.

Per the standing corpus policy on σύν- compounds, the prefix survives in the literal tier without exception.

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

The verse is textually stable; NA28 prints it without variation in the critical apparatus of any consequence.

Two minor points worth recording:

• Some witnesses read the simplex ὑπεκρίθησαν for the compound συνυπεκρίθησαν. The compound is both the harder reading and the better attested, and the direction of the variation is the expected one — a rare triple compound simplified toward a familiar verb. The variant must not be followed, since the σύν- is what links the clause to αὐτῷ.
• The spelling of Βαρναβᾶς fluctuates in the manuscript tradition (Βαρνάβας, Βαρναβᾶς) as accentual and orthographic variation of a Semitic name; this carries no semantic weight and is not rendered differently.

Nothing in the tradition softens the charge, which is itself notable: this is a verse that reports an apostle and a revered missionary in the wrong, and scribes did not touch it.

The verse sits at the centre of one of the sharpest exegetical quarrels of the early church.

Origen, Chrysostom, Theodoret, and Jerome (following a Greek tradition Jerome largely took over) read the whole Antioch episode as a staged disagreement: Peter withdrew deliberately so that Paul could publicly rebuke him and the Gentile believers could be set free without Peter having to oppose his own constituency. On this reading ὑπόκρισις is a kind of pious stagecraft — Chrysostom calls the rebuke οἰκονομία, an accommodation — and neither apostle is really at fault.

Augustine rejected this flatly, in a correspondence with Jerome (Epp. 28, 40, 82) that remained sharp for years. His argument was that if Paul’s reported words in vv. 14-21 are themselves a performance, then the letter’s own account of truthfulness collapses, and nothing in Scripture can be trusted to say what it means. Peter was genuinely wrong; Paul genuinely opposed him; the text says so.

Later tradition has sometimes also pressed the episode into service for or against Petrine authority. That is reception, not exegesis, and supplies no evidence about what the Greek says (§11).

Proposed — pending reviewJohn Chrysostom, Commentary on the Epistle to the Galatians, in Patrologia Graeca 61 · Augustine, Epistulae 28, 40, 82 (correspondence with Jerome), in Corpus Scriptorum Ecclesiasticorum Latinorum 34 · Jerome, Commentarii in Epistulam ad Galatas, in Patrologia Latina 26

The name is the climax of the sentence and its force depends entirely on what the reader already knows.

Within Galatians itself: Barnabas went up to Jerusalem with Paul (2:1) and received the right hand of κοινωνία alongside him from James, Cephas, and John (2:9). He is, in the argument Paul is constructing, a co-witness to the Gentile mission’s legitimacy. The ascensive καί — ‘even Barnabas’ — presupposes exactly this: the one man whose standing with Paul was established two verses earlier.

The wider record (Acts 4:36; 9:27; 11:22-26; 13-14) fills this in further: a Hellenistic Jew from Cyprus, the man who vouched for Paul in Jerusalem when nobody else would, who fetched him from Tarsus to Antioch, and who was his senior partner on the first missionary journey. Acts 15:39 records a later parting between them over Mark. Whether that separation and this incident are related is not something Galatians says, and nothing may be read into the verse from it.

Particular note: Paul does not excuse Barnabas, and he does not name him in order to settle a score. He names him because the spread of the pretense to Barnabas is the measure of how serious the thing was. A performance that could carry off Barnabas could carry off anyone.

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)