James ends without a benediction, a greeting, or an Amen. The letter simply stops on a promise made to whoever goes after a wanderer.

The verse is the apodosis of the conditional begun in v. 19: if someone strays from the truth and someone turns him back — then let him know what he has done. Everything hangs on the third-person imperative γινωσκέτω, which addresses not the strayer but the retriever.

Three things drive the exegesis, and all three are genuinely open:

1. Who saves. σώσει has no expressed subject. It may be the human turner (the nearest available subject, and the one the whole sentence has been describing) or the Lord (on the ground that only God saves — and James has just said as much at 4:12, ὁ δυνάμενος σῶσαι καὶ ἀπολέσαι).
2. Whose life, whose sins. αὐτοῦ after ψυχήν is both textually unstable and referentially ambiguous, and πλῆθος ἁμαρτιῶν has no possessive at all. The sinner’s sins are the obvious referent; a long tradition — patristic and Reformation alike — has heard the turner’s own sins covered as well.
3. What ‘cover’ means. καλύπτω is the LXX verb of Ps 31:1 (‘blessed the one whose sins are covered’), and it stands here in a clause that reproduces Prov 10:12 almost as 1 Pet 4:8 does. Whether the covering is forgiveness by God, the arresting of sins that would otherwise have been committed, or both, James does not say.

The apparatus below keeps all three open. Every smooth English rendering that resolves them — ‘will save that person from death’, ‘will cover over a multitude of his sins’ — quietly settles a question James left standing, and the reader is entitled to see the seam.

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

Third-person present imperative of γινώσκω, standing at the head of the apodosis.

Two points. First, the person. Per standing corpus policy on third-person imperatives (γνωσθήτω, Phil 4:5; αἰσχυνέσθω / δοξαζέτω, 1 Pet 4:16), English requires a ‘let’ periphrasis, and the third person must not be converted to a second-person address. James has been speaking to ἀδελφοί μου (v. 19) and then narrows to a τις inside the conditional; the imperative follows that τις, not the vocative. ‘You should know’ flattens a deliberate narrowing.

Second, the lexeme. The corpus keeps γινώσκω (‘come to know’, relational-experiential) distinct from οἶδα (‘know’, settled state) — James himself uses οἶδα-family forms elsewhere (Ἴστε, 1:19). Here the sense is ingressive: let this dawn on him. The Byzantine variant γινώσκετε (2nd pl. imperative or indicative) levels the clause toward the vocative of v. 19 and is treated below.

The ὅτι is a plain complementizer supplying the content of the knowing — neither causal (‘because’) nor recitative. Contrast the causal ὅτι already fixed at Gal 4:6.

ἐπιστρέψας is the aorist active participle, substantival with the article, and it is transitive: ἁμαρτωλόν is its direct object.

The verb is the LXX standard for שׁוּב, the prophetic verb of covenantal turning (Deut 30:2; Isa 45:22; Hos 14:2 LXX), and it carries both intransitive (‘turn, turn back’) and transitive-causative (‘turn someone back’) uses. The causative is well attested in exactly this soteriological register: Luke 1:16-17 (πολλοὺς … ἐπιστρέψει ἐπὶ κύριον, of John the Baptist, echoing Mal 3:23-24 LXX), and in the LXX at Jer 23:22 and Sir 48:10.

James uses the verb twice in two verses — intransitively is not the case in v. 19 either: ἐπιστρέψῃ τις αὐτόν is also transitive. The corpus policy is that a turning-word must be kept in English at every occurrence so the road-image and the שׁוּב field stay visible. Hence ‘turns back’, not ‘brings back’, ‘restores’, or ‘reclaims’.

‘Converts’ (Vulg. converti fecerit) is accurate to the Latin but is confined to declared liturgical register: modern English ‘convert’ has narrowed to initial religious change of allegiance, whereas James’s ἁμαρτωλός is a member of the community (ἐν ὑμῖν, v. 19) who has strayed.

The aorist participle is complexive — the act taken as a whole — and its substantival use with the article names a class: whoever does this.

Anarthrous substantival adjective, and its anarthrous qualitative force matters: not the sinner but a member of that class.

The corpus already fixes ἁμαρτωλός as ‘sinner / sinful one’, a standing characterization rather than a single act, and distinguishes it from παραβάτης (‘transgressor’, one who has stepped across a marked line — James’s word at 2:9, 11) and παράπτωμα (‘trespass, false step’). The choice here is notable: the person described in v. 19 has wandered from the truth, and James now names him with the harder word.

ἐκ πλάνης ὁδοῦ αὐτοῦ — ‘out of the wandering of his way.’

πλάνη is the noun to πλανάω, already fixed in this corpus as ‘be led astray’ with the road-image retained as live (§8). The noun names the wandering itself, not the false opinion arrived at: LXX and NT usage (Prov 14:8; Matt 27:64; Rom 1:27; 2 Pet 2:18; 3:17; 1 John 4:6) keeps the straying-from-a-path sense recoverable throughout.

The pairing with ὁδοῦ is decisive. James has written a genitive of a road-noun after a wandering-noun; ‘the error of his way’ (the traditional English, from Vulg. ab errore viae suae) preserves the two nouns but modern English ‘error’ has become a purely cognitive word and the road has gone dead in it. ‘Wandering’ keeps the metaphor visible and keeps πλάνη audibly cognate with πλανηθῇ in v. 19 — a link the whole two-verse unit depends on.

Excluded from reading tiers: ‘from his wandering’ (deletes ὁδοῦ), ‘from his mistaken ways’ (converts a singular concrete road into a plural abstraction), ‘from the error of his ways’ (same, and the English idiom now carries a faintly comic register).

The genitive ὁδοῦ is best taken as the sphere in which the wandering occurs, or possessive — his way, and his straying on it. The Semitic ‘way’ idiom (דֶּרֶךְ) for a manner of life is fully at home here, and the corpus already retains ὁδοί as ‘ways’ at Jas 1:8.

Three unresolved questions converge on this clause.

(i) Who is the subject of σώσει? Greek supplies none. The nearest available subject is ὁ ἐπιστρέψας — grammatically the most natural, and the reading the sentence’s whole shape invites, since the verse exists to tell the turner what he has accomplished. But James has just insisted (4:12) that there is εἷς … ὁ δυνάμενος σῶσαι καὶ ἀπολέσαι, and 5:15 has ὁ κύριος as the one who raises the sick. A theological correction is therefore possible: the human agent turns, the Lord saves.

(ii) Whose ψυχή? αὐτοῦ most naturally continues the reference of the immediately preceding αὐτοῦ (the sinner’s way). But the Koine simple pronoun frequently does duty for the reflexive, so ‘his own life’ — the turner’s — is grammatically available, and this is where the second-century and Reformation traditions divide.

(iii) The text itself is unstable here. See the textual section.

Policy: the reading text renders ‘will save his life’, which in English is as ambiguous as the Greek — ‘his’ can take either antecedent — and does not adjudicate. The study tier renders ‘a life’ with the referent flagged. Excluded from the reading text: ‘will save that person from death’ (adopts the sinner-reading and deletes the pronoun) and ‘will save his own soul’ (adopts the turner-reading).

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

Corpus policy is settled: ψυχή renders Hebrew נֶפֶשׁ and denotes the whole living self, physical existence included; ‘soul’ imports a body/spirit dualism foreign to the context and is excluded as the base gloss.

James himself fixes the sense internally. At 1:21 he writes τὸν δυνάμενον σῶσαι τὰς ψυχὰς ὑμῶν, and at 5:20 σώσει ψυχὴν αὐτοῦ ἐκ θανάτου — the same verb, the same noun, and here an explicit ἐκ θανάτου. What is rescued is a life from death, on the model of Ps 32:19 LXX (ῥύσασθαι ἐκ θανάτου τὰς ψυχὰς αὐτῶν) and Prov 23:14.

‘Save a soul from death’ is deeply embedded English and may be named as reception (§11), but it supplies no evidence. It is retained in the liturgical tier and nowhere else.

The noun is anarthrous. This is worth noting rather than smoothing: James writes ‘a life’, not ‘the life’, which slightly favors leaving the referent open.

Anarthrous, unqualified, at the clause’s turning point — exactly as at Jas 1:15 (ἡ δὲ ἁμαρτία ἀποτελεσθεῖσα ἀποκύει θάνατον), which is the letter’s other statement about where sin ends.

Per standing policy the bare noun is retained with no supplied adjective. Whether the referent is physical death, eschatological ruin, or both undifferentiated is genuinely contested, and the Jewish sources James stands nearest — Sir 15:11-20, Wis 1:12-16, and the Ezekiel watchman oracles — do not separate them. ‘Eternal death’ and ‘spiritual death’ each supply an adjective the Greek withholds and settle the question in the reading text.

The internal link to 1:15 is the strongest evidence available and should govern the reader’s expectation: sin brought to term gives birth to death; turning a sinner back saves a life from it. The two verses are the letter’s opening and closing statements of the same trajectory, running in opposite directions.

The corpus already fixes καλύπτω as ‘cover’, with the metaphor treated as live (§8) and the resolving alternatives excluded.

The LXX field is decisive. ἐπικαλύπτω/καλύπτω renders כָּסָה of sins in Ps 31:1 LXX (μακάριοι ὧν ἀφέθησαν αἱ ἀνομίαι καὶ ὧν ἐπεκαλύφθησαν αἱ ἁμαρτίαι), where the covering stands in synonymous parallel with forgiveness, and it renders the same Hebrew verb at Prov 10:12 (πάντας δὲ τοὺς μὴ φιλονεικοῦντας καλύπτει φιλία) and Prov 17:9. Neh 3:37 LXX and Ps 84:3 LXX belong to the same field.

All three are live, and all three have ancient support. What is excluded from the reading text is any rendering that names one: ‘will bring about the forgiveness of’, ‘will atone for’, ‘will prevent’, ‘will hide’. ‘Cover’ holds them all, as the Greek does.

πλῆθος is a mass-noun of magnitude, not the adjective ‘many.’ Corpus policy retains the noun so the heap-image survives — decisive here, since the same phrase πλῆθος ἁμαρτιῶν stands at 1 Pet 4:8 where the Hebrew source has כָּל (‘all’), and the Greek’s noun-of-magnitude is a translational choice worth seeing.

Critically, no possessive stands with ἁμαρτιῶν. James supplies αὐτοῦ twice earlier in the sentence and withholds it here. That silence is the crux.

Rendering: keep the bare phrase, supply no possessive, and let the omission stand. ‘A multitude of his sins’ and ‘a multitude of sins of his own’ both write in what the Greek withheld.

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

The closing clause is close to Prov 10:12b, but not to the LXX of it.

MT Prov 10:12b: וְעַל כָּל־פְּשָׁעִים תְּכַסֶּה אַהֲבָה — ‘but love covers over all transgressions.’

LXX Prov 10:12b: πάντας δὲ τοὺς μὴ φιλονεικοῦντας καλύπτει φιλία — a paraphrase that keeps καλύπτει but loses the ‘transgressions’ object entirely.

1 Pet 4:8: ἀγάπη καλύπτει πλῆθος ἁμαρτιῶν. Jas 5:20: καλύψει πλῆθος ἁμαρτιῶν.

James and 1 Peter share καλύπτει/καλύψει πλῆθος ἁμαρτιῶν verbatim, and neither matches the LXX. Two consequences.

First, the wording is closer to the Hebrew than the LXX is, which suggests either an alternative Greek version or — more likely — a fixed early-Christian proverbial form circulating independently of the LXX text. That the two letters share it word for word, while neither cites it as scripture, points the same way.

Second, and important for §7: James is not obviously quoting Prov 10:12. He supplies no citation formula (contrast 2:8, 2:23, 4:6). The clause functions as a proverb absorbed into the argument. Whether James is echoing Proverbs, quoting a common Christian saying, or drawing on 1 Peter (or vice versa) is undecidable from the wording alone.

The practical rule: render the clause so that its verbal identity with 1 Pet 4:8 is visible in English, since that identity is the primary datum. Both should read ‘cover a multitude of sins.’

The structure of vv. 19-20 — one person strays, another intervenes, a life is or is not saved from death — reproduces the watchman oracles of Ezekiel 3:18-21 and 33:8-9 LXX closely enough to be worth stating.

Ezek 3:18-19 LXX: if the watchman does not warn the ἄνομος ἀποστρέψαι ἀπὸ τῶν ὁδῶν αὐτοῦ, τοῦ ζῆσαι αὐτόν, the man dies in his lawlessness and his blood is required of the watchman; if he warns and the man does not turn (μὴ ἀποστρέψῃ ἀπὸ τῆς ἀνομίας αὐτοῦ καὶ ἀπὸ τῆς ὁδοῦ αὐτοῦ), the watchman has delivered his own life (τὴν ψυχὴν αὐτοῦ ἐρρύσω).

Four of James’s terms are here in Greek: the ἀπο-/ἐπι-στρέφω turning-verb, ὁδός, ψυχή, and the life/death outcome. And crucially, Ezekiel has both possibilities in play — the strayer’s life and the watchman’s own — which is precisely the ambiguity of Jas 5:20.

This does not decide the question, but it makes the double reference intelligible rather than eccentric. A Jewish-Christian writer working in this register would have had both parties’ lives naturally in view.

Cf. also Dan 12:3 LXX (οἱ κατισχύοντες τοὺς λόγους μου … ὡς οἱ ἀστέρες) and Sir 48:10 on Elijah’s turning of hearts, which James has invoked two verses earlier.

Proposed — pending reviewHatch, Edwin, and Henry A. Redpath. A Concordance to the Septuagint and the Other Greek Versions of the Old Testament. 2nd ed. Grand Rapids: Baker, 1998.

Three points of instability, none doctrinally decisive but two exegetically live.

(1) γινωσκέτω / γινώσκετε. The third-singular imperative γινωσκέτω is read by ℵ A P Ψ and the great majority; the second-plural γινώσκετε (imperative or indicative) appears in B and some Latin support. The plural assimilates to the vocative ἀδελφοί μου in v. 19 and shifts the address from the individual retriever to the whole congregation — the easier reading, and the direction of scribal levelling elsewhere in James. The singular stands.

(2) ψυχὴν αὐτοῦ / ψυχὴν / ψυχὴν αὐτοῦ omitted or transposed. The witnesses fluctuate: ψυχὴν αὐτοῦ (ℵ, the Byzantine tradition, and most), ψυχὴν alone (B, and some versional support), and αὐτοῦ ψυχήν or ψυχὴν αὐτοῦ ἐκ θανάτου αὐτοῦ in scattered witnesses. NA28 prints ψυχὴν αὐτοῦ.

This is where the exegesis is touched. A bare ψυχήν leaves the referent wholly open and would slightly favor rendering ‘a life’; the pronoun narrows it, though — as argued above — not decisively, since Koine αὐτοῦ can do reflexive duty. The variation itself is best explained as scribes reacting to the ambiguity: some clarifying by adding, some by omitting.

(3) Final ἀμήν. Added by ℵ P and part of the Byzantine tradition, absent from B A and the earliest witnesses. This is the standard direction of accretion — per corpus policy on closing formulae, a word said aloud in worship migrates easily into a text read aloud in worship. James ends without it, and the abruptness is a genuine feature of the letter.

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

The Fathers divide on the same question modern commentators do, and the division is old enough to be evidence that the Greek was felt as open.

Origen takes the covering of sins to concern the one who does the turning: the retriever’s own sins are covered by the act. This reading was influential and is the source of the persistent Western tradition that the verse promises the rescuer something.

The Greek catena tradition and Œcumenius generally read the ψυχή as the sinner’s, treating the verse as a straightforward description of what pastoral retrieval accomplishes for the strayer.

Bede, commenting on the Catholic Epistles, reads the covering as forgiveness on the Ps 31:1 model and understands the whole clause of the sinner, while noting the reward accruing to the one who turns him.

What the patristic evidence establishes is negative but useful: no ancient reader found the verse unambiguous, and none felt licensed to resolve it by paraphrase. That is the discipline the rendering above tries to keep.

Proposed — pending reviewBede the Venerable. Commentary on the Seven Catholic Epistles. Trans. David Hurst. Cistercian Studies 82. Kalamazoo: Cistercian Publications, 1985.

ἐκ occurs twice in the verse — ἐκ πλάνης ὁδοῦ αὐτοῦ and ἐκ θανάτου — and the repetition is structural, not stylistic.

Per standing corpus policy, ἐκ marks source or separation and is kept distinct from ἀπό (origin), ὑπό (personal agency), and διά (means). Here both instances are separation: out of a condition occupied.

The two phrases are parallel and answer each other. The sinner is turned out of his wandering; a life is saved out of death. What the retrieval removes him from in the first clause is what he is removed from in the second — the wandering road and the death it ends in are the same trajectory (cf. 1:15). One English preposition must serve both occurrences, or the answering structure disappears; ‘from the error of his way … from death’ is acceptable in smooth registers, but the literal tier keeps ‘out of’ at both points.

Note also the anarthrous θανάτου against the anarthrous ψυχήν: nothing in the second half of the verse is definite except by inference.