James closes the faith-and-works argument (2:14-26) with a comparison, not a further proof. The sentence is built as an exact two-limb balance — ὥσπερ … οὕτως καί — and the two limbs are made to match word for word: subject, χωρίς + genitive, νεκρός, ἐστιν. Whatever the first limb asserts about a corpse, the second asserts about a workless faith.

Three decisions carry the verse.

First, πνεῦμα is anarthrous and stands in a clause about a body’s death, which puts the ordinary Greek and Hebrew sense ‘breath’ squarely in view (Gen 2:7; Ps 103:29 LXX; Eccl 12:7). ‘Spirit’ is not wrong — the word is the same — but modern English ‘spirit’ hears an immaterial component departing, which pushes the analogy toward an anthropological claim James is not making.

Second, the tertium comparationis is contested. On the commoner reading, works are to faith as breath is to the body: the animating principle, without which the thing is a corpse. On the reverse reading, faith is the body and works the breath, so the comparison runs the other way and works are the evidence of life rather than its source. The Greek word order gives σῶμα/πίστις in the same slot and πνεύματος/ἔργων in the same slot, which favors the first alignment; but the analogy is a simile, not an equation, and James’s own argument (v. 22: faith worked with his works and was brought to completion by works) resists pressing it into a causal ordering.

Third, νεκρά is a verdict on a kind of faith, not a definition of faith as such. James has already qualified the subject twice (v. 14 ‘if someone says he has faith’; v. 17 ‘faith by itself’), and v. 26 recapitulates rather than extends. The line is a diagnosis: the thing on the table is not living faith in a weakened state but a corpse.

The verse ends the section with no connective following, and 3:1 begins a new topic. The abruptness is the rhetoric.

Proposed — pending reviewDavids, Peter H. The Epistle of James: A Commentary on the Greek Text. New International Greek Testament Commentary. Grand Rapids: Eerdmans, 1982.

The full correlative pair. ὥσπερ is the strengthened comparative (ὡς + περ), regularly answered by οὕτως in formal comparison (Matt 12:40; 13:40; 24:27; Rom 5:12, 19, 21; Jas 2:26). Both limbs must be marked in English, and the adjunctive καί must be rendered separately: it attaches to ἡ πίστις (‘faith also’) rather than merely joining clauses, adding the second case to the first rather than restating it. Dropping it — as smoother English versions do — removes the additive move that makes the comparison an argument rather than an illustration.

The structural parallelism is exact and deliberate:

James has used the same οὕτως καί hinge at 2:17 with the identical predicate (οὕτως καὶ ἡ πίστις … νεκρά ἐστιν), so v. 26 is a closing inclusio on v. 17 with the corpse-image supplied. The repetition is structure, not redundancy, and must be reproduced identically in English.

Per the standing policy, γάρ is rendered and never dropped: epistolary and paraenetic argument structure is load-bearing (§10).

What it grounds is a real question. Read narrowly, it grounds v. 25 (Rahab’s case) — but Rahab needs no explanation from a corpse-analogy. Read broadly, it grounds the whole of 2:14-25, functioning as a summarizing ‘for’ that supplies the underlying principle at the close. The second reading is preferable on discourse grounds: the verse repeats v. 17 verbatim in its predicate, and summary-γάρ at the end of a unit is well attested. English ‘for’ is as open as the Greek and is retained rather than resolved to ‘because’ or ‘after all.’

σῶμα is the concrete physical body — the thing that can be flogged, imprisoned, or buried — and per the standing glossary entry it is never rendered ‘person,’ ‘self,’ or ‘life.’ Here the referent is specifically a corpse: a body from which breath has gone.

That fixes the register of the whole clause. James is not offering an anthropology; he is pointing at a dead body. The image is deliberately unpleasant, and §6 forbids softening it.

Anarthrous πνεύματος, governed by χωρίς. Two renderings are live.

Breath. πνεῦμα is the ordinary Greek word for breath and wind, and this is the settled LXX equivalent of רוּחַ / נְשָׁמָה in exactly this kind of clause: Gen 2:7 (God breathes and the man becomes a living being), Ps 103:29 LXX (ἀντανελεῖς τὸ πνεῦμα αὐτῶν, καὶ ἐκλείψουσιν, καὶ εἰς τὸν χοῦν αὐτῶν ἐπιστρέψουσιν), Eccl 12:7, Ezek 37:8-10 (bodies assembled but with no πνεῦμα in them, until breath comes and they live). Ezekiel 37 is the closest conceptual parallel available: assembled, complete, articulated — and dead, for want of breath. Given James’s sapiential and LXX-saturated idiom, this is the reading with the strongest substrate.

Spirit. The anthropological sense is equally ordinary in Koine (Luke 8:55; 23:46; Acts 7:59), and the corpus already fixes πνεῦμα with a human possessive as ‘spirit’ (1 Pet 4:6; Phil 4:23). Nothing in the grammar excludes it.

Excluded in every tier: ‘the Spirit’ (capitalized). The divine Spirit is not in view, and no ancient reading supports it here.

Proposed — pending reviewJohnson, Luke Timothy. The Letter of James. Anchor Bible 37A. New York: Doubleday, 1995.

χωρίς + genitive marks absolute separation, not degree (already fixed at Phil 2:14; Heb 9:22; 1 Tim 2:8). The corpse is not short of breath; it has none. The faith is not deficient in works; it has none.

The repetition of the preposition across both limbs is the hinge of the comparison and must be rendered with the same English word both times. Varying to ‘apart from … without’ for style breaks the parallel the whole verse is built on. This is the fourth occurrence of χωρίς in the paragraph (2:18, 20, 26 twice), and James is hammering one word.

ἔργων is anarthrous here, against the articular τῶν ἔργων at 2:18 (twice), 2:20, 2:21, 2:22, 2:24, 2:25. The shift is worth recording: articular ἔργα throughout the argument point back to the concrete deeds under discussion (feeding the destitute, Abraham’s offering, Rahab’s reception of the messengers), while the anarthrous form at the close generalizes — works of whatever kind, works as such.

What the anarthrous form does not do is turn James’s ἔργα into Paul’s ἔργα νόμου. §9 forbids importing a later doctrinal frame: James’s works throughout ch. 2 are acts of mercy toward the poor and obedience to the royal law (2:8, 15-16), not boundary-marking Torah observance. Paul marks his own category explicitly with νόμου, and James never does.

Per the standing ἔργον policy the plural is retained and rendered ‘works’ at every occurrence in the chapter, so the concordance across 2:14-26 stays intact.

Articular ἡ πίστις. The article is either generic (‘faith’ as a category) or anaphoric (‘the faith’ just described — the faith that says without doing, v. 14; the faith of the demons, v. 19). English idiom drops it in either case, but the anaphoric reading matters exegetically: James is not defining faith, he is delivering a verdict on a specimen.

The subject is qualified in every parallel statement of the paragraph: v. 14 πίστιν λέγῃ τις ἔχειν (‘says he has faith’), v. 17 καθʼ ἑαυτήν (‘by itself’), v. 20 χωρὶς τῶν ἔργων, v. 26 χωρὶς ἔργων. The qualifier is never absent, and any English rendering that leaves it out (‘faith is dead’) reports something James did not say.

Per the standing entry, ‘faith’ is retained rather than ‘belief’; the content of the faith comes from context (v. 19 shows what a bare propositional πίστις amounts to).

νεκρόν … νεκρά — the same predicate adjective, agreeing in each limb with its subject, and fronted before the copula in both. The fronting is emphatic and should be preserved where English word order permits.

The corpus already fixes substantival νεκρός as physically dead, with the metaphorical sense requiring an explicit marker (1 Pet 4:5-6). Here the first limb is literal and the second metaphorical, and the metaphor is licensed by the simile itself, not by the lexeme.

Critically, the word does not mean ‘weak,’ ‘ineffective,’ or ‘incomplete.’ James had ἀργή available and used it at v. 20 (‘idle, out of work’ — itself a pun on ἔργα); at v. 26 he chose the stronger word. A corpse is not a body performing badly. Renderings such as ‘faith without works is useless’ or ‘amounts to nothing’ substitute a functional assessment for a mortal one and shade a hard text toward comfort (§6).

See also the textual note: the variant ἀργή at v. 20 shows scribes moving between these two adjectives, which is itself evidence that they were felt as distinct.

Proposed — pending reviewMayor, Joseph B. The Epistle of St. James: The Greek Text with Introduction, Notes and Comments. 3rd ed. London: Macmillan, 1913.

The verse is a simile with two terms on each side, and how they align determines what is being claimed.

The rendering does not adjudicate: the English reproduces the Greek’s parallelism exactly and leaves the alignment where James left it.

On the Paul/James question more broadly: the two authors use ἔργα for different referents, and δικαιόω is not at issue in this verse at all. Nothing in v. 26 requires the harmonizing or the contrasting reading, and the apparatus does not supply one.

Proposed — pending reviewDibelius, Martin. James: A Commentary on the Epistle of James. Rev. Heinrich Greeven. Hermeneia. Philadelphia: Fortress, 1976.

NA28 prints χωρὶς ἔργων (anarthrous) with no bracketing and no variant flagged in the main apparatus for the second limb. Some witnesses read τῶν ἔργων, assimilating to the articular form used throughout 2:18-25; the anarthrous reading is the harder and better attested, and the assimilation runs in the predictable direction.

At 2:20 — one variant worth recording here because it bears directly on the force of νεκρά at v. 26 — the witnesses divide between ἀργή (‘idle, out of work’, read by the earliest witnesses and printed by NA28) and νεκρά (‘dead’, the Byzantine reading, assimilated to vv. 17 and 26). The scribal movement is from the rarer ἀργή toward the νεκρά of the surrounding verses. That the two adjectives were interchangeable enough to swap, yet distinct enough for the swap to be detectable, confirms that James’s choice of νεκρά at v. 26 is a choice.

The reading πνεύματος is unvaried; no witness substitutes πνοῆς (‘breath’), which would have settled the lexical question. The Vulgate’s spiritu and the Peshitta both keep the ambiguous word, so the ancient versions did not have to choose either.

The Greek and Latin commentators uniformly read the verse as the summation of the argument rather than as a fresh point, and they read the analogy in the majority alignment — works stand to faith as breath to the body.

Augustine’s handling is the most consequential for later reception. In De fide et operibus he took James to be correcting a misreading of Paul already circulating in the first century, and he pressed the corpse-image to argue that a faith which produces nothing is not a lesser faith but not the saving thing at all. That reading — that James and Paul use πίστις with different referents rather than contradicting each other — governed the Western tradition and is still the standard harmonizing move.

The Greek tradition tends to treat the verse less polemically, as ordinary paraenesis: the assembly is being told that professed belief without mercy to the destitute (2:15-16) is a corpse.

Per §11, none of this is evidence for the rendering. It is recorded as reception, and it is worth recording chiefly because the Reformation-era debate over this verse is downstream of Augustine’s framing rather than of the Greek.

Proposed — pending reviewAugustine. De fide et operibus (On Faith and Works). CSEL 41. Trans. Gregory J. Lombardo, Ancient Christian Writers 48. New York: Newman Press, 1988.