One short sentence closes the longest sustained piece of scriptural argument in Galatians. Everything from v. 21 (“you who want to be under law, do you not hear the law?”) has been building toward it, and Paul states the conclusion in the flattest possible terms: οὐκ ἐσμὲν παιδίσκης τέκνα ἀλλὰ τῆς ἐλευθέρας — a nine-word verdict with no verb in the second half.

Three things are worth watching.

First, the inference marker. διό is inferential and it reaches back over the whole Hagar-Sarah reading, not just over the Gen 21:10 quotation immediately before it. Paul is not deducing “therefore we are free” from “cast out the slave woman”; he is stating the standing to which the whole allegory has been driving, of which the expulsion command is one element.

Second, the person. ἐσμέν, first person plural, not the second person of v. 28 (ἐστέ) or v. 30’s imperative. Paul steps inside the circle he has drawn. Whatever else the Galatian crisis involved, he does not lecture from outside it.

Third, the asymmetry of the two genitives: anarthrous παιδίσκης, articular τῆς ἐλευθέρας. The slave woman is a category; the free woman is the one already identified — Sarah, and behind her the Jerusalem above of v. 26, called “our mother.” The Greek makes the second term definite and the first indefinite, and the English should not level them.

And there is one lexical decision with real force behind it: παιδίσκη and ἐλευθέρα are legal status terms out of the Greek slave economy, and Paul has been using them that way since v. 22. Softening either one to “servant” or “freeborn lady” dissolves the argument.

διό is the standard Koine inferential particle (δι’ ὅ, “on account of which”), stronger and more conclusive than οὖν and rarer than ἄρα. Paul uses it about twenty-seven times, characteristically at the close of a movement rather than in the middle of one (Rom 1:24; 2 Cor 4:13; Eph 2:11; Phil 2:9).

The scope question is the live one. Two readings:

1. διό follows from the Gen 21:10 citation in v. 30 — since Scripture says cast out the slave woman’s son, we must be sons of the free woman.
2. διό follows from the whole allegorical argument of vv. 21-30, of which v. 30 is the scriptural clincher.

The second is preferable. The logic of (1) is odd: the expulsion command by itself would not establish who we are, only what happens to those on the wrong side of the ledger. Reading (2) makes v. 31 the summary of vv. 21-30 and the hinge into 5:1’s τῇ ἐλευθερίᾳ ἡμᾶς Χριστὸς ἠλευθέρωσεν — the ἐλευθερ- root carrying straight over the chapter break.

Either way the particle must be rendered. Epistolary connectives are argument structure (§10); dropping it would leave a bare assertion where Paul wrote a conclusion.

Two variants, both instructive and neither doctrinally weighty.

1. διό vs. ἄρα οὖν vs. ἡμεῖς δέ. NA28 reads διό with 𝔓⁴⁶ ℵ* A C 6 33 1739 and the Vulgate. Byzantine witnesses (and D ψ) read ἄρα οὖν, which is Paul’s other favorite inferential combination (Rom 5:18; 7:3; 8:12; 9:16; 14:12) and is the easier, more familiar formula — hence likely the secondary smoothing. A third group (ℵ² B D* F G 062) reads ἡμεῖς δέ, “but we,” which would make the verse a contrastive statement rather than an inference. B’s support gives this some weight, and it produces a neat parallel with v. 28’s ὑμεῖς δέ.

2. ἐσμέν vs. ἐστέ. A handful of witnesses read the second person “you are.” This is almost certainly assimilation to the ὑμεῖς δέ of v. 28 and the second-person address of vv. 21, 30.

Nothing in the theology turns on this. Whether Paul wrote “therefore we are not” or “but we are not,” the claim is the same. Recorded here because §4 requires that the seam be shown rather than hidden.

Built onna28 · metzger-tc

παιδίσκη is a diminutive of παῖς and in Hellenistic usage denotes a female slave — not a young girl and not a hired servant. The documentary papyri use it in sale contracts and household inventories; it is a status word.

Its weight here is entirely LXX-derived. In Gen 16:1-6 and 21:10-13 LXX, Hagar is repeatedly ἡ παιδίσκη (Hebrew שִׁפְחָה / אָמָה). Paul is not choosing a generic word for a slave woman; he is quoting the vocabulary of the Genesis narrative, which is why the term recurs unchanged at Gal 4:22, 23, 30 (twice), and 31. The LXX’s own choice matters: Gen 21:10 has ἔκβαλε τὴν παιδίσκην ταύτην καὶ τὸν υἱὸν αὐτῆς, and Paul reproduces it in v. 30 nearly verbatim.

One further note: παιδίσκης here is anarthrous, where every prior occurrence in the section (vv. 22, 23, 30) carried the article. The article in vv. 22-30 was anaphoric to the Genesis narrative — the slave woman, Hagar. Anarthrous here, the noun is qualitative: we are not children of a slave woman, i.e. we do not stand in that category of descent at all. English is obliged to supply an indefinite article, but the qualitative force is the reason it is the right one.

Built onbdag · muraoka-lxx

ἡ ἐλευθέρα is the substantival feminine of ἐλεύθερος, “free,” and it is the exact legal counterpart of παιδίσκη: the free-status woman over against the slave-status woman. Paul introduced the pair at v. 22 (ἕνα ἐκ τῆς παιδίσκης καὶ ἕνα ἐκ τῆς ἐλευθέρας) and it has done all the structural work of the passage since.

The word is not in the Genesis narrative. Sarah is never called ἐλευθέρα in Gen 16 or 21 — the LXX simply names her. Paul has supplied the term, generating the antithesis by contrasting Hagar’s actual designation with a status word of his own choosing. That is a deliberate move: it converts a family narrative into a statement about two legal conditions, which is what makes it transferable to the two διαθῆκαι of v. 24 and to the Galatians’ situation.

The article is retained here (τῆς ἐλευθέρας) where παιδίσκης is anarthrous. The free woman is definite and already identified: Sarah in the narrative, and by way of v. 26 the ἄνω Ἰερουσαλήμ who is ἐλευθέρα … μήτηρ ἡμῶν. The link is lexical, not merely thematic — ἐλευθέρα in v. 26 and τῆς ἐλευθέρας here are the same word.

The root then carries straight into 5:1: τῇ ἐλευθερίᾳ … Χριστὸς ἠλευθέρωσεν. Chapter divisions obscure what the Greek makes obvious — v. 31 and 5:1 are one thought, and the ἐλευθερ- stem occurs four times across the two verses.

Built onbdag

τέκνα — “children,” not “sons.” Paul has υἱός available and uses it freely in this chapter (vv. 4, 6, 7, 22, 30), where inheritance and legal standing are in view. Here and at v. 28 (ἐπαγγελίας τέκνα) and v. 25 (μετὰ τῶν τέκνων αὐτῆς) he uses τέκνον, the relational term of birth and derivation.

The distinction is not accidental: τέκνον is built on τίκτω, “bear, give birth,” and the whole allegory has been running on birth vocabulary — γεννῶσα (v. 24), γεγέννηται (v. 23), γεννηθείς (v. 29), ὠδίνω (v. 19), τίκτουσα and ὠδίνουσα in the Isa 54:1 citation (v. 27). τέκνα belongs to that field. υἱός belongs to the inheritance field of vv. 1-7.

So the rendering keeps them apart: υἱός “son,” τέκνον “child,” per the standing corpus policy. Collapsing τέκνα to “sons” here would import juridical sonship into a clause whose point is maternal derivation — which mother we come from.

Vocative masculine plural, Paul’s third address of the chapter (vv. 12, 28, 31). Rendered “brothers,” per §3 and the standing corpus policy: masculine plural is the Greek default for a mixed-sex group and that convention is data about the language, not a claim about the composition of the audience.

That the Galatian assemblies included women is not in doubt (cf. 3:28, οὐκ ἔνι ἄρσεν καὶ θῆλυ), and the referent here is certainly mixed. The note carries that; the main text does not silently rewrite the grammar. It is worth observing that Paul addresses a mixed assembly as ἀδελφοί in the middle of an argument whose two poles are two women — the vocative’s grammatical gender and the allegory’s referential gender are simply running on different tracks.

The clause is οὐκ ἐσμὲν παιδίσκης τέκνα ἀλλὰ τῆς ἐλευθέρας — and the second half has no noun. τέκνα is elided after ἀλλά and must be supplied from the first half: “but [children] of the free woman.”

Greek does this readily; English can do it too, and here it should. Repeating “children” (“but children of the free woman”) is not wrong, but the terseness of the Greek is rhetorically live — the sentence lands hard because it stops early. In the smooth rendering the ellipsis is retained (“but the free woman’s”), which reads naturally in English and reproduces the compression. The literal tier brackets the supplied noun.

One word-order feature: παιδίσκης τέκνα places the genitive before its noun, whereas the standard order would be τέκνα παιδίσκης. Fronting the genitive topicalizes it — whose children is the question at issue, and the answer-word comes first. English cannot reproduce this without awkwardness (“a slave woman’s children we are not”), but the wooden tier keeps the Greek sequence visible.

Built onbdf

Galatians 4:21-31 is Paul’s reading of Gen 16 and 21 under the label ἀλληγορούμενα (v. 24). The structure:

• v. 21 — challenge to those who want to be ὑπὸ νόμον
• vv. 22-23 — the narrative datum: two sons, two mothers, two modes of birth (κατὰ σάρκα / δι’ ἐπαγγελίας)
• vv. 24-27 — the correspondences: two covenants, Sinai/Hagar/present Jerusalem in slavery over against the Jerusalem above, free, our mother, with Isa 54:1 as proof
• vv. 28-29 — application to the Galatians, including the persecution motif
• v. 30 — Gen 21:10 as the scriptural verdict
• v. 31 — the conclusion, in the first person

What v. 31 does not do is state a new premise. It states the standing of the addressees, and it does so in terms that make 5:1 the next move rather than a new topic. The rhetorical shape is worth noting: after four verses of dense typology and two scriptural citations, Paul closes with a single unadorned sentence containing no metaphor at all. The plainness is the point.

Chrysostom’s homily on the passage stresses the reversal: the Galatians, in seeking to come under law, were choosing descent from the slave woman, and Paul’s conclusion tells them what they in fact already are. He is careful to note that Paul says “we,” not “you” — the apostle does not exempt himself from the household he is describing.

Jerome’s commentary on Galatians (the fullest patristic treatment) takes τῆς ἐλευθέρας as pointing forward to libertas in 5:1 and treats the chapter break as an artificial interruption — a judgment the Greek supports.

The Latin tradition rendered the verse itaque, fratres, non sumus ancillae filii sed liberae. Note ancilla for παιδίσκη: a servant-girl, which is already a softening relative to the Greek slave-status term, and filii for τέκνα, which collapses the child/son distinction Paul maintains. Both are reception, not evidence (§11), and neither should shape the rendering — but the drift is instructive, since English translations have tended to inherit exactly those two softenings.