Three moves in one verse, and the third is where the trouble is.

First, οὖν ties the command back to what precedes — but to how much of it? If it reaches only to v. 15c (“if he has committed sins, it will be forgiven him”), then confession here is what the sick person and the elders do in the situation just described. If it reaches back over the whole of vv. 13-15, it generalizes: the whole congregation, not just the sickbed.

Second, the verbs change. James has used προσεύχομαι three times in vv. 13-14 and now switches to εὔχομαι twice (v. 16), before returning to προσευχή/προσεύχομαι in vv. 17-18. Whether that is stylistic variation or a real shift of register is worth flagging, not settling.

Third, the closing maxim. πολὺ ἰσχύει δέησις δικαίου ἐνεργουμένη is four content-words and at least three live questions: whether ἰσχύω is being used in its strength-sense (per the standing glossary decision) or has drifted toward the commercial ‘avail’; whether δικαίου names a special class of holy man or simply anyone who is right with God; and above all what ἐνεργουμένη is doing — middle (‘as it goes to work’), passive with divine agent (‘when it is energized [by God]’), or attributive to δέησις (‘an effectual prayer’). Every smooth English rendering picks one. This apparatus does not.

One more thing worth noticing before the details: James does not name who does the healing. ἰαθῆτε is a passive with no agent, and the corpus policy is to leave it that way.

And the sentence’s plainest feature is the least discussed: ἀλλήλοις … ἀλλήλων. Twice in nine words. Confession and prayer both run in both directions. There is no clergy and no penitent here — only a congregation.

Inferential οὖν, and per the standing corpus policy it is always rendered and never dropped (§10). ‘So then’ carries both the inference and the discourse shift.

What is genuinely open is scope.

The grammatical datum that favours the broad reading is precisely the person-shift, which the corpus treats as discourse architecture elsewhere in James (2:3-4; 2:16). The datum that favours the narrow reading is the anaphoric article on τὰς ἁμαρτίας (see below), which points back to a noun in v. 15.

Note also the textual wrinkle: a considerable body of witnesses reads ἐξομολογεῖσθε δέ rather than οὖν. δέ would loosen the connection to a mild transition rather than an inference. NA28 prints οὖν, and the harder inferential particle is more likely to have been softened to δέ than the reverse.

Built onna28 · metzger-tc · bdf

The corpus glossary already fixes ἐξομολογέω as confess / acclaim, settled on Phil 2:11, where the LXX covenantal-doxological register dominates. That entry needs a note here rather than an amendment.

The compound’s range in biblical Greek runs in two directions from the same root idea of open, out-loud acknowledgment:

• Acclaim, praise — the dominant LXX use, rendering Hebrew hôdâ (Ps 6:6; 9:2; 105:1 LXX; Sir 51:1), and the sense at Matt 11:25, Luke 10:21, Rom 14:11, and Phil 2:11.
• Confess (sins) — Lev 5:5; 16:21; Num 5:7; Neh 1:6; 9:2-3 LXX; Dan 9:20; and in the NT Matt 3:6, Mark 1:5, Acts 19:18, Jas 5:16, with the cognate simplex at 1 John 1:9.

With τὰς ἁμαρτίας as object the second sense is fixed beyond dispute. What must be preserved is the out-loud force of the ἐξ-: this is not interior contrition (μεταμέλομαι) nor a change of mind (μετανοέω), both available to James. The Leviticus texts are the controlling background, and they concern spoken acknowledgment in a public cultic setting.

The present imperative is durative — a standing practice, not a single occasion.

Excluded from reading tiers: ‘admit’ (drops the cultic-public register), ‘acknowledge’ (too cerebral), ‘own up to’ (register collapse). Also excluded is any rendering implying a directional or hierarchical act — the reciprocal dative forbids it (below).

Built onbdag · muraoka-lxx

The critical text has the bare article and no possessive: τὰς ἁμαρτίας, ‘the sins’.

Two things follow.

(1) The article is anaphoric. ἁμαρτίας stands in v. 15 (κἂν ἁμαρτίας ᾖ πεποιηκώς). The article here most likely points back to it: the sins — those under discussion. This is the strongest single argument for the narrow scope of οὖν.

(2) A possessive must be supplied, and the supplement must be recorded. English cannot say ‘confess the sins to one another’, so reading tiers render ‘your sins’. But a substantial body of witnesses supplies ὑμῶν in the Greek, and that supplement is precisely the scribal instinct English translators are reproducing. NA28 prints the bare form; the addition is the easier reading. The literal tier keeps ‘the sins’ so the reader can see what is and is not in the text.

On the lexeme itself, the standing corpus policy holds: ἁμαρτία is rendered sin and kept distinct from παράπτωμα (‘trespass / false step’), παράβασις (‘transgression’), and ἀδικία (‘unrighteousness’), all of which James had available and none of which he wrote here.

Built onna28 · metzger-tc · bdf

The reciprocal pronoun stands twice in nine words, once as dative with ἐξομολογεῖσθε and once as genitive with ὑπέρ. This is the verse’s structural spine and the corpus anti-variation discipline applies at full force: identical English wording (‘one another’) at both occurrences, in matching slots, even where a translator’s ear would prefer variation.

The reason is not stylistic tidiness. It is that the reciprocity is the whole ecclesiology of the sentence. Whatever the elders of v. 14 are doing, v. 16 is not a continuation of it in the same key: the confession runs both ways, and so does the prayer. James has no word here for a confessor, a penitent, or an officiant.

James uses ἀλλήλων four times (4:11 twice; 5:9; 5:16 twice), always of members acting on members.

One small note on English: Koine also uses the plural reflexive ἑαυτοῖς for the reciprocal (1 Pet 4:8, 10), and the corpus renders both ‘one another’ with the underlying form recorded in the parsing table. Here the genuine reciprocal stands.

Built onbdf · bdag

James switches verbs. Three times in vv. 13-14 he writes προσεύχομαι (προσευχέσθω, προσευξάσθωσαν) and the cognate noun εὐχή appears at v. 15. Now at v. 16 he writes εὔχεσθε twice, and at vv. 17-18 he returns to προσευχή/προσηύξατο.

εὔχομαι is the older and broader verb: to pray, and also simply to wish (Acts 26:29; 27:29; Rom 9:3; 2 Cor 13:7, 9; 3 John 2). προσεύχομαι is the near-technical biblical-Greek compound for prayer addressed to God, overwhelmingly dominant in the LXX and NT.

The corpus policy on προσεύχομαι already keeps it distinct from δέησις/δέομαι. εὔχομαι is added as a third member of the field, rendered pray and its distinctness from προσεύχομαι recorded in the parsing table rather than manufactured in English. There is no English pair that reproduces the Greek one, and inventing one (‘pray … make petition’) would collide with δέησις, which arrives four words later as a distinct noun.

ὑπὲρ ἀλλήλων is the standard benefactive genitive already fixed in the corpus (for the sake of / on behalf of), to be kept typographically distinct in notes from ὑπέρ + acc. (‘above, beyond’).

Built onbdag · muraoka-lxx

Proposed — pending reviewHenry George Liddell, Robert Scott, and Henry Stuart Jones, A Greek-English Lexicon, 9th ed. with revised supplement (Oxford: Clarendon Press, 1996)

ὅπως + aorist subjunctive is telic. James uses ἵνα elsewhere (5:9, 12) and ὅπως only here; the two are largely interchangeable in Koine for purpose clauses (BDF), so no distinct force should be pressed from the choice.

The aorist is complexive — the healing taken as a single whole event, not a process.

What matters is the syntactic relation of the purpose clause to the two imperatives. Three readings:

1. Attached to both imperatives: confess and pray, so that you may be healed.
2. Attached to εὔχεσθε only: the healing is the aim of the prayer, and the confession is preparatory.
3. Attached to the whole as one act, with confession and prayer functioning as a single practice.

Word order does not decide it: the clause follows the second imperative, which slightly favours (2), but Greek regularly lets a final clause govern a preceding coordinated pair. The rendering keeps the clause where Greek puts it and does not resolve the attachment.

On the passive: ἰαθῆτε has no expressed agent. Per standing corpus policy the divine passive stays passive and the agent stays unexpressed — ‘so that God may heal you’ supplies what James withheld, and he had ὁ κύριος available two words earlier (v. 15) and did not use it here.

Built onbdf · na28

ἰάομαι is the standard verb of healing in the LXX and NT: physical cure (Matt 8:13; 15:28; Luke 9:11; Acts 9:34) and, in a well-established metaphorical extension, restoration from sin or covenantal breach (Isa 6:10 LXX, cited at Matt 13:15 and Acts 28:27; Isa 53:5 LXX ἰάθημεν, taken up at 1 Pet 2:24; Jer 3:22; Hos 6:1; Ps 40:5 LXX).

Both poles are live here, and the choice is exegetically consequential.

English ‘heal’ carries both senses and is retained precisely for that reason. Excluded from reading tiers: ‘be cured’ (adopts the physical reading and narrows to medicine), ‘be restored’ (adopts the second), ‘be made whole’ (retained in declared liturgical register, where the older English has fixed it, but in modern registers it reads as a therapeutic abstraction).

Keep lexically distinct from σῴζω (‘save’, v. 15, 20) and ἐγείρω (‘raise’, v. 15), both of which stand in the immediate context and both of which are themselves double-sensed. James has stacked three ambiguous verbs in three verses. That is a feature, not an accident, and no one of them should be used to settle another.

Built onbdag · muraoka-lxx

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

The maxim opens with a fronted πολύ — topicalizing, and the emphasis must be preserved by keeping the degree-word early in the English clause.

On the construction: ἰσχύω is stative and intransitive (‘be strong, have strength adequate to’), and the corpus glossary already fixes it as be strong (for) / have strength (for) with ‘can do’ excluded — that entry was settled on Phil 4:13, and refined at Gal 5:6 to add ‘count for’ for the construction with an accusative of respect. πολύ here is exactly that: an accusative of respect or internal accusative specifying the degree of the strength, not a direct object.

So woodenly: much has strength a petition of a righteous man.

Gnomic present: a general truth, rendered with a simple English present, never a progressive.

Built onbdag · bdf

δέησις — the corpus glossary fixes this as petition (liturgical alt. supplication), from δέομαι ‘lack, need’: a specific request arising from want, narrower than προσευχή, and lexically distinct from it because Paul pairs the two at Phil 4:6 and 1 Tim 2:1.

That distinction is worth pressing here, because James has now used three prayer-words in four verses: προσεύχομαι (13, 14), εὔχομαι and εὐχή (15, 16), and δέησις (16). A translation that renders all of them ‘prayer’ hides the range. The noun is anarthrous and qualitative — a petition, not the prayer.

δικαίου is a substantival genitive adjective. Per standing corpus policy δίκαιος shares an English root with δικαιοσύνη (‘righteousness’) and δικαιόω, and that root-link is live in James, who has argued at length about being δικαιωθείς (2:21-25) and about the fruit of δικαιοσύνη (3:18), and who has just spoken of τὸν δίκαιον murdered by the rich (5:6).

The two readings are not as far apart as they look, since in LXX usage δίκαιος is a covenant-standing word rather than a merit-grade. But the rendering must not adjudicate: retain the bare ‘of a righteous person’ and supply no qualifier — ‘of a truly righteous man’, ‘of a godly person’, and ‘of anyone who is righteous’ each settle the question in the reading text.

On gender: the masculine substantival adjective is the Greek default for a referentially inclusive statement (§3). Reading tiers use ‘a righteous person’; the literal tier keeps ‘a righteous [man]’ with the supplied noun bracketed and the generic force recorded.

Built onbdag · muraoka-lxx

Proposed — pending reviewDale C. Allison Jr., A Critical and Exegetical Commentary on the Epistle of James, International Critical Commentary (London: Bloomsbury T&T Clark, 2013)

This is the hard word of the verse, and there is no rendering that does not decide something. Three questions, partly independent.

(1) Voice. The present participle ἐνεργουμένη is morphologically identical for middle and passive.

The corpus already faced this exact ambiguity at Gal 5:6 and adopted a rendering hospitable to both (‘working through’). The same discipline applies: ‘at work’ in reading tiers, which can be heard either as the petition working or as the petition having been set working, with the alternative flagged.

(2) Attachment. Does the participle modify δέησις attributively, or is it circumstantial with ἰσχύει?

• Attributive: an operative petition of a righteous person has great strength — i.e. a class of prayer is specified.
• Circumstantial: a righteous person’s petition, when it is at work, has great strength — i.e. a condition or manner is added.

The participle stands last in the clause, after the noun and its genitive, which slightly favours a predicative-circumstantial reading; but attributive position with an anarthrous noun is unremarkable in Koine. The rendering keeps the participle at the end, set off by commas, so the attachment remains as open in English as in Greek.

(3) The KJV problem. ‘The effectual fervent prayer of a righteous man availeth much’ is one of the most familiar sentences in English devotional usage. It is worth naming as reception (§11) precisely so it can be excluded as evidence. Two things are wrong with it as a rendering: it doubles the participle into two adjectives (‘effectual fervent’), and ‘fervent’ has no lexical warrant in ἐνεργέω at all — it imports the intensity-vocabulary of ἐκτενής (already fixed in the corpus as at full stretch, with the temperature-metaphor ‘fervent’ excluded on the same grounds). The familiarity of the phrase is not an argument for it.

Excluded in every tier: ‘fervent’, ‘earnest’, ‘persistent’ (all substitute an intensity- or duration-word for a work-word), and ‘powerful and effective’ (converts a participle into a second predicate adjective and makes the clause tautologous with ἰσχύει).

Built onbdag · bdf · silva-becnt

Proposed — pending reviewRalph P. Martin, James, Word Biblical Commentary 48 (Waco, TX: Word Books, 1988)

Three variants, none doctrinally weighty, all worth seeing.

οὖν / δέ. NA28 reads οὖν with strong early support; a large body of witnesses, including the Byzantine tradition, reads δέ. The variant matters only for the strength of the connection to vv. 14-15 (inference vs. mild transition). οὖν is the harder reading and the more likely to have been softened.

τὰς ἁμαρτίας / τὰς ἁμαρτίας ὑμῶν. The addition of the possessive is well attested and is exactly the clarification a scribe (or a translator) supplies. NA28 prints the bare articular noun. The addition is secondary; the anaphoric article is what English has to render with a possessive anyway, which is why the supplement should be bracketed in the literal tier rather than silently absorbed.

εὔχεσθε / προσεύχεσθε. Part of the tradition assimilates the second imperative to the προσεύχομαι of vv. 13-14 and 17-18. This is levelling toward the commoner and more distinctly religious compound, and it removes the very lexical switch discussed above. Retain εὔχεσθε.

Built onna28 · metzger-tc

James does not leave the maxim unexplained. Verse 17 opens Ἠλίας ἄνθρωπος ἦν ὁμοιοπαθὴς ἡμῖν — Elijah was a man of like feelings with us — and then reports that he prayed with prayer (προσευχῇ προσηύξατο, a Semitizing cognate-dative intensifying the verb, on the pattern the corpus records at Luke 22:15 and John 3:29) and the heavens shut for three and a half years.

This is exegetically decisive in one respect and suggestive in another.

Decisive: the ὁμοιοπαθής clause is best read as forestalling the inference that δικαίου names an unreachable class. James chooses the most spectacular intercessor in Israel’s memory and immediately insists on his ordinariness. Whatever δίκαιος means, it is not a barrier the congregation cannot cross.

Suggestive: the Elijah illustration is about a prayer that did something in the world — it stopped and started rain. That is at least a hint in favour of the middle reading of ἐνεργουμένη (prayer going to work) over a reading that locates the efficacy purely in the disposition of the one praying. It is a hint, not a proof, and the participle remains open.

Note also the structural bracket: vv. 13-18 begin and end with prayer, and the paragraph’s last word before the closing appeal (vv. 19-20) is about turning a wanderer back — which returns the letter to the community-fracture theme and lends some weight to the broader, restorative reading of ἰαθῆτε.

Built onbdf

Two threads in the patristic reception, and they need to be kept apart from the lexical evidence (§11).

On ἀλλήλοις, the Greek commentators generally read the reciprocity at face value: confession is mutual within the body. Origen (Hom. in Lev.) and later writers do connect the verse with the practice of disclosing sin to a spiritual physician, and by the time of the medieval Western commentators (Bede among them) the verse is standard in discussions of penitential discipline. That development is real history and it is worth knowing that this verse carried the weight of it. It is also, straightforwardly, later than the text: the Greek supplies no officiant, no absolving party, and no directionality.

On δικαίου, the Fathers characteristically take v. 17’s ὁμοιοπαθής as James’s own gloss and use it against the reading that the maxim applies only to spiritual elites. This is a good instinct and it is grounded in the text rather than imposed on it.

The methodological point: the fact that a devotional or ecclesial tradition has grown around a sentence tells us how the sentence has been used. It tells us nothing about what the words mean. That cuts both ways — against those who read later penitential practice into ἀλλήλοις, and equally against those who react so far against it that they empty the reciprocal of any concrete practice at all.

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