After five verses of hammering questions (vv. 1-5), Paul stops interrogating and produces a text. The verse is a bare quotation of Gen 15:6 LXX, introduced by a single comparative conjunction and nothing else: no γέγραπται, no λέγει ἡ γραφή, no main clause of its own. The effect is abrupt — the Galatians’ own experience of the Spirit (vv. 2, 5) is set alongside Abraham’s believing without a word of explanation, and the explanation is deferred to v. 7 (γινώσκετε ἄρα).

Three things carry the weight.

1. The citation is verbatim (with one word-order change and the name supplied), which makes this the hinge where Paul moves from argument-from-experience to argument-from-Scripture. Abraham is almost certainly the opponents’ own ground, and Paul takes it from them.
2. ἐλογίσθη is a divine passive with an unexpressed subject and an unexpressed agent. English wants to supply both (‘God credited his faith to him’), and supplying either resolves a question the Greek leaves open.
3. εἰς δικαιοσύνην is a Semitism, not a purpose or a transformation. It marks the heading under which a ledger entry is posted — the accounting sense of λογίζομαι is primary, and the corpus has already fixed it so.

The syntactical difficulty — a καθώς clause with nothing to attach to — is not a defect to be repaired but the seam at which Paul’s argument turns.

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

καθώς is comparative (κατά + ὡς), normally answered by a correlative οὕτως or attached to a main clause. Here it has neither. Paul writes a comparison and supplies no second limb; the sentence as printed has no finite main verb outside the quotation.

Per the corpus policy on unmarked constructions and anacoluthon, the literal tier keeps the bare καθώς with no supplied verb and no supplied correlative, and the reading text marks the hanging comparison with a dash rather than closing the sentence. Excluded from the reading text: ‘Consider Abraham’ (supplies an imperative absent from the Greek) and ‘So also Abraham’ (supplies the correlative οὕτως Paul did not write).

At Rom 4:3 the same citation is introduced with τί γὰρ ἡ γραφὴ λέγει — a formal citation formula. The absence of any such formula here is the datum.

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

Ἀβραάμ is the indeclinable LXX transliteration of אַבְרָהָם, its case supplied by syntax (here nominative, in v. 8 dative τῷ Ἀβραάμ). Transliterate as written; supply no gloss of the name’s etymology in the reading text.

The name’s presence here is itself the argument’s pivot. Abraham occurs nine times in Galatians (3:6, 7, 8, 9, 14, 16, 18, 29; 4:22), clustered entirely in this stretch, and the density strongly suggests Paul is not introducing a figure of his own choosing but contesting one already in play. Second-Temple Jewish tradition had a well-developed Abraham: the first to keep Torah before it was given (Sir 44:19-21, where the covenant is tied to his keeping of the law and to circumcision; Jub. 23:10; cf. 1 Macc 2:52, where it is Abraham’s faithfulness in testing that is reckoned righteousness — אמונה read as proven loyalty). If the agitators in Galatia argued from Abraham that circumcision and Torah-keeping belong to Abrahamic sonship, Paul’s move is to go behind Gen 17 to Gen 15 — to the chapter where Abraham has nothing but a promise and believes it.

Note that Gen 15:6 precedes Gen 17 (circumcision) in the narrative, and Paul presses exactly this chronology in Rom 4:9-12. He does not press it here; he simply quotes.

πιστεύω with a dative of person is the third of the three constructions the corpus distinguishes: πιστεύω εἰς + acc (‘believe into’, personal committal, Gal 2:16), πιστεύω ὅτι (propositional assent, Jas 2:19), and πιστεύω + dat (credit someone’s word — Gen 15:6 LXX; Jas 2:23; Acts 27:25; 1 John 5:10).

The dative is the person whose word is taken as reliable, not a recipient of a gift and not a beneficiary. Abraham credits what God has said. This matters because the rendering must not drift toward either of the other constructions: ‘believed in God’ imports εἰς and shifts the object from the word to the person; ‘had faith in God’ converts the verb into a noun-phrase and makes faith a possession rather than an act; ‘trusted God’ is excluded as base gloss because πείθομαι/πέποιθα own the trust-domain in this corpus.

The aorist is complexive: the believing taken as one whole act, not a process and not a state. Paul neither narrates a moment of conversion nor describes a settled disposition; he quotes a verb.

Context note: in Hebrew, הֶאֱמִן בַּיהוָה (hiphil of אמן + בְּ) covers both ‘give credence to’ and ‘rely on’, and the LXX’s dative is a serviceable but narrower rendering. The breadth belongs in the note, not in the English — per §1 and §9 the Greek as written is the base text.

λογίζομαι is primarily an accounting word — to calculate, to post an entry to a ledger — with ‘hold a considered view’ derived from it (the corpus fixes it as ‘reckon / count’, kept distinct from ἡγέομαι ‘consider’ and οἴομαι ‘suppose’). Here the aorist passive is the LXX’s rendering of Hebrew חָשַׁב, which belongs to the same domain: the entry is made, and the thing entered stands credited.

Three features must survive in English, and each is a place where translations routinely repair what the Greek withholds.

The passive stays passive. The agent is unexpressed. ‘God credited it to him’ supplies the agent; the construction withholds it, though in context no other agent is conceivable. This is the standing divine-passive policy.

The subject is unexpressed. Greek has no stated subject for ἐλογίσθη — not πίστις, not τὸ πιστεῦσαι, nothing. English must supply ‘it’, and the referent of that ‘it’ is the contested question below.

The ledger image stays live (§8). ‘Credited’ is permitted in reading tiers; ‘counted’ is the liturgical option (and the word by which this verse entered English); ‘regarded as’ and ‘accepted as’ are excluded as abstracting the accounting away, and ‘imputed’ is confined to declared doctrinal register, since in modern English it has become a technical term of a particular system rather than a word about books.

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

εἰς + accusative here is the Semitizing εἰς of predication, rendering Hebrew לְ: it marks the heading or category under which something is entered, not purpose, not transformation, not direction. The corpus has already fixed this for the Gen 15:6 citation family and for Ps 105:31 LXX (Phinehas), and one English rendering must serve Gen 15:6 LXX, Rom 4:3, 5, 9, 22, Gal 3:6, and Jas 2:23 so the shared substrate stays visible.

‘as righteousness’ in reading tiers; ‘unto righteousness’ in the literal tier to display the preposition. Excluded: ‘so that he became righteous’ (transformative — makes εἰς telic and converts a ledger entry into an ontological change), ‘in order to make him righteous’ (same, with an added agent), ‘to his credit as righteousness’ (over-explains the metaphor the preposition only implies).

δικαιοσύνη retains the single corpus gloss ‘righteousness’ across the forensic and ethical poles. Here the pole is plainly forensic-relational — a standing credited by another — but the English must not vary: δικαιοσύνη at Gal 3:6, 21 and δικαιόω at 2:16-17, 3:8, 11, 24 are one family and one argument, and ‘justice’ narrows to the social-forensic while ‘uprightness’ narrows to the ethical.

Per §9: the diachronic finding governs. In the first century δικαιοσύνη with λογίζομαι and εἰς is ledger-and-standing language, not the developed sixteenth-century technical vocabulary; the later doctrinal reading is reception and is handled in the Reception section, never blended into the base rendering.

The καί joining the two clauses is paratactic, reproducing the Hebrew waw of Gen 15:6 (וַיַּחְשְׁבֶהָ). Render plain ‘and’ per the corpus policy against manufacturing logical relations the connective withholds (ascensive δέ, Phil 2:8; additive καί after a negation, Jas 2:24; consecutive καί after imperatives, Phil 4:7).

This is not pedantry. The relation between believing and being reckoned righteous is precisely what the chapter is about, and rendering ‘and so it was credited’ or ‘with the result that’ supplies a causal or consecutive link that Paul has left to the juxtaposition. The Hebrew leaves it to the waw; the LXX leaves it to καί; the English leaves it to ‘and’.

Note also that the καί must always be rendered (§10) — dropping it (‘Abraham believed God, it was credited to him’) removes the one word binding the two halves of the citation.

Paul quotes Gen 15:6 LXX essentially verbatim, with two differences from the Göttingen text:

The LXX’s Ἀβράμ is the correct pre-Gen 17:5 form of the name; Paul (and Rom 4:3, and Jas 2:23) uses Ἀβραάμ throughout, the familiar post-renaming form. This is normalization rather than a point being made.

The word-order change (name fronted before the verb) is forced by the substitution of καθώς for the LXX’s narrative καί: with the comparative conjunction heading the clause, the subject takes the next slot.

Hebrew (MT): וְהֶאֱמִן בַּיהוָה וַיַּחְשְׁבֶהָ לּוֹ צְדָקָה. Two points bear on the Greek. (i) The Hebrew wayyaḥšəbehā has an explicit feminine singular object suffix — ‘he reckoned it’ — whose antecedent is most naturally the believing; the LXX’s agentless passive ἐλογίσθη drops both the explicit divine subject and the object suffix, which is how the Greek comes to have no stated subject at all. (ii) The Hebrew names YHWH (בַּיהוָה); the LXX has τῷ θεῷ, and Paul follows the LXX. Per §1 the Greek is rendered as written — θεός is ‘God’, not converted to the divine name — and the Hebrew substrate is recorded here rather than restored to the text.

No significant variation in the Greek witnesses for Gal 3:6. A few witnesses prefix καί or substitute γάρ for καθώς, reflecting the felt difficulty of the unattached comparative clause — which is evidence that the hanging καθώς is the harder and original reading.

Proposed — pending reviewJohn William Wevers, ed., Genesis, Septuaginta: Vetus Testamentum Graecum Auctoritate Academiae Scientiarum Gottingensis editum I (Göttingen: Vandenhoeck & Ruprecht, 1974)

The verse is remarkably reticent, and the reticence is worth stating plainly before the doctrinal traditions are brought in.

Paul asserts that Abraham believed God and that something was reckoned to him as righteousness. He does not say here (i) who did the reckoning, (ii) what precisely was reckoned, (iii) whether the believing was itself meritorious or merely the occasion, (iv) that circumcision was absent, or (v) that works were excluded. Items (iv) and (v) he argues elsewhere (Rom 4:9-12 for the chronology; Gal 3:10-12 for the exclusion); item (i) is supplied by the divine passive and no one has seriously contested it. Items (ii) and (iii) are the fault lines.

What the verse does do, structurally, is convert an argument from the Galatians’ experience (vv. 1-5) into an argument from Scripture (vv. 6-14), and it does so by placing Abraham — a figure the opponents presumably invoked for circumcision — at the head of the faith side of the ledger. The inference is drawn in v. 7 (οἱ ἐκ πίστεως, οὗτοι υἱοί εἰσιν Ἀβραάμ) and the Gentile scope in v. 8 (ἐνευλογηθήσονται ἐν σοὶ πάντα τὰ ἔθνη, Gen 12:3 / 18:18). Reading v. 6 apart from vv. 7-9 misses the point: Paul quotes Gen 15:6 not primarily to define how an individual is justified but to establish who Abraham’s children are.

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

Within the New Testament. Gen 15:6 is cited at Rom 4:3 (with the full citation formula τί γὰρ ἡ γραφὴ λέγει), restated at Rom 4:5, 9, 22, and cited at Jas 2:23 — where it is joined to Abraham’s offering of Isaac and introduced with ἐπληρώθη ἡ γραφή, a verb the corpus keeps on the ‘fill up’ root rather than resolving to ‘fulfil’. That Paul and James cite one verse to different ends is a primary datum of early Christian exegesis and belongs in the reader’s view; per the corpus policy on ἔργα and δικαιόω, both authors are rendered with the same English words so the reader can weigh the relation for himself rather than receiving it pre-resolved.

Second Temple Judaism. 1 Macc 2:52 reads the verse as Abraham’s proven faithfulness under testing (Ἀβραὰμ οὐχὶ ἐν πειρασμῷ εὑρέθη πιστός, καὶ ἐλογίσθη αὐτῷ εἰς δικαιοσύνην) — the same Greek clause, applied to the Aqedah rather than to the promise of Gen 15. Sirach 44:19-21 ties the Abrahamic covenant to his keeping of the law and to the covenant in his flesh. Whether Paul is contesting precisely this line of reading is a reasonable inference and not a demonstrable one.

Rabbinic literature. Later rabbinic treatment of Gen 15:6 is extensive and is not uniformly a works-reading; the Mekhilta tradition associating Israel’s faith at the sea with Abraham’s is one strand. Any single rabbinic citation here would be later than Paul by centuries and is reported, if at all, as parallel rather than as background.

The early Church. Both the faith-as-credited and the gift-of-righteousness readings are present early. Chrysostom (Hom. in Gal. 3) presses the chronological argument — Abraham was justified before the law and before circumcision — and treats faith itself as what is reckoned. Augustine, especially in the anti-Pelagian works, insists that the faith which is reckoned is itself God’s gift, so that nothing in the transaction originates with the man; the argument is theological rather than lexical and should be recorded as such.

Later tradition. Gal 3:6 and Rom 4:3 became load-bearing texts in the sixteenth-century debates over iustificatio, with the Latin imputatum est and reputatum est mediating the Greek passive, and the English ‘counted’ / ‘imputed’ tradition descending from there. Per §11 this resonance may be named but supplies no evidence for the rendering: the base text follows the accounting sense of λογίζομαι and the Semitizing force of εἰς, and the doctrinal history is recorded here.

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, 1889)