Two clauses, no connective between them, and the whole weight of the paragraph on the second.

The first clause states the sanction that backs the command of v. 12: ἡ γὰρ κρίσις ἀνέλεος τῷ μὴ ποιήσαντι ἔλεος — the judgment is merciless to the one who has not done mercy. It is a proverb in shape, an exact inversion of the beatitude (Matt 5:7), and it is the plainest thing in the verse.

The second clause is one of the shortest sentences in the letter and one of the most argued over: κατακαυχᾶται ἔλεος κρίσεως — four words, no article, no connective, no stated subject beyond the bare noun. Three questions are live and each of them changes the sense. Whose mercy is it — God’s, or the merciful person’s? Is κατακαυχάομαι hostile (‘boast down over, triumph over an opponent’) or celebratory (‘exult’)? And what is the relation between the clauses: does the second reverse the first, or does it explain how anyone escapes it?

The apparatus below argues that the asyndeton is deliberate, that the clause reads as a gnomic maxim, and that James has left the subject of the mercy unspecified on purpose. What must not happen is that a smooth English rendering settles what four Greek words leave open.

γάρ grounds v. 12: speak and act as those about to be judged by a law of freedom, for the judgment is merciless to the unmerciful. Per §10 the particle is rendered, never dropped — the sanction-clause is what makes v. 12 a warning rather than an observation.

The internal seam is unmarked. Between ἔλεος and κατακαυχᾶται the critical text has no καί, no δέ, no ἀλλά. The corpus policy on asyndeton (fixed at Phil 3:2, Phil 4:8, Gal 4:5, Gal 5:4) applies: no connective is supplied in English, and the abruptness is the data. Editors punctuate with a raised point (NA28) or a semicolon; both are editorial, since the earliest majuscules are unpunctuated here.

That matters, because the unmarked seam is precisely what leaves the relation between the clauses open. A supplied ‘but’ would make the second clause a reversal of the first. A supplied ‘for’ would make it an explanation. James supplied neither.

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

κρίσις names the judicial process — the trial, the act of judging — as against κρίμα, the verdict rendered, and κριτήριον, the court itself. James uses κριτήρια at 2:6 of the courts into which the rich drag the poor, and the near-context is thoroughly forensic.

The first occurrence is articular (ἡ κρίσις) and anaphoric: it points back to μέλλοντες κρίνεσθαι in v. 12. This is the judgment already named, not judgment in the abstract. The second occurrence is anarthrous and genitive (κρίσεως), which is qualitative — judgment as such, as a principle over against mercy. The English cannot easily mark the shift without clumsiness; the literal tier keeps the article at the first and withholds it at the second, and the difference is recorded here.

Note also the internal chiasm the two nouns produce across the verse: κρίσις — ἔλεος — ἔλεος — κρίσις. The mercy-words stand at the centre, the judgment-words at the edges. Whatever else the second clause does, it is formally the mirror of the first.

ἀνέλεος is a NT hapax and effectively absent from earlier Greek in this form; the commoner classical adjectives are ἀνελεής and ἀνηλεής. It is an α-privative formation on ἔλεος and is two-termination (one form serving masculine and feminine), hence nominative feminine here agreeing with κρίσις.

The privative shape must survive in English. ‘Merciless’ is idiomatic and correct; ‘without mercy’ is preferred in the literal tier because it keeps the ἔλεος morpheme audible against the three other occurrences of the root in this verse. What is excluded is any positive substitute — ‘harsh’, ‘severe’, ‘unsparing’ — each of which names a quality of the judgment rather than the absence of the specific thing the unmerciful withheld. The whole force of the clause is lex talionis in vocabulary: no ἔλεος given, no ἔλεος received.

ἔλεος is the LXX standard for חֶסֶד — enacted covenant loyalty, not sentiment and not clemency toward the guilty as such. The corpus already fixes it as ‘mercy’ (Gal 6:16) and keeps it lexically distinct from χάρις (‘grace’), οἰκτιρμός (‘compassion’), and χρηστότης (‘kindness’).

The idiom ποιέω ἔλεος — literally ‘do mercy’ — is the fixed LXX rendering of עָשָׂה חֶסֶד (Gen 24:12; Judg 1:24; 1 Kgdms 20:8; and in the NT at Luke 1:72; 10:37 ὁ ποιήσας τὸ ἔλεος μετʼ αὐτοῦ). It denotes mercy done — an act of relief performed for someone, not a feeling entertained. That is exactly the point in a letter whose immediately preceding argument is about seating the poor man under a footstool (2:2-4) and whose next paragraph is about sending the naked and hungry away warmed with words (2:15-16).

So the object of ποιήσαντι must stay a direct object in English. ‘To the one who has shown no mercy’ is acceptable in smooth registers; ‘to the one who has not been merciful’ is excluded, because it converts a deed into a disposition and erases the Semitic idiom on which James’s whole ethic of enacted faith runs.

Note the neuter gender: ἔλεος in Hellenistic Greek is neuter (τὸ ἔλεος), against the classical masculine (ὁ ἔλεος). This is morphology, not meaning, but it fixes ἔλεος as the nominative subject of κατακαυχᾶται in the second clause and as accusative object in the first.

Article + negated aorist participle, substantivized: ‘to the one who has not done mercy.’

Two features carry weight.

The negative. μή is the regular participial negative in Koine (BDF §430) and, per the standing corpus entry, carries no independent emphasis. But its presence here does mark the participle as generic and characterizing rather than as a fact asserted about a named person: this is a class description, ‘whoever it may be that has done no mercy,’ not an accusation levelled at anyone in particular. Had James written οὐ, the description would be asserted of an actual person.

The aorist. Complexive, summing up a course of conduct as a single whole — not punctiliar (‘on one occasion failed to show mercy’) and not future-referring. English present perfect (‘has not done’) tracks this best: it views the whole record from the standpoint of the judgment. A simple past (‘did not show mercy’) locates the failure at a point; a present (‘does not show mercy’) converts a settled record into ongoing behaviour.

The dative is one of reference or disadvantage — the judgment falls toward this person. Not agency.

The verse turns on this verb, and it is genuinely contested.

Morphologically it is κατά + καυχάομαι. The corpus already fixes καυχάομαι as ‘boast’ — ordinarily pejorative in Greek, but redirected by Paul by changing its object on the Jer 9:22-23 LXX model. The κατά- prefix here is not perfective but adversative, and it governs the genitive: to boast down over someone, i.e. to exult over a defeated or inferior party.

The other NT occurrences fix the hostile register. Rom 11:18 twice: μὴ κατακαυχῶ τῶν κλάδων — do not boast over the branches. Jas 3:14: μὴ κατακαυχᾶσθε καὶ ψεύδεσθε κατὰ τῆς ἀληθείας — do not exult over and lie against the truth. Both are contests with a loser.

So the image is not that mercy is nicer than judgment, nor that mercy is preferred. The image is that mercy and judgment have met as adversaries and mercy has come off on top and says so. The vaunt is part of the word: κατακαυχάομαι names not merely winning but crowing over the win.

Rendering policy: ‘triumphs over’ in reading tiers, because English ‘triumph over’ preserves both the contest and the resulting exultation; ‘exults over’ where the boast-element must be foregrounded; ‘boasts down over’ woodenly in the literal tier so the κατά- prefix and the καυχ- root stay visible against the corpus’s other καυχάομαι renderings. Excluded: ‘prevails over’ and ‘wins out over’ (drop the vaunt), ‘rejoices against’ (archaic and now opaque), and ‘mercy is better than judgment’ (dissolves a verb into a comparative and discards the contest entirely).

The present tense is gnomic — what is characteristically the case — and renders as a simple English present, never a progressive.

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

The clause has no possessive, no article, and no stated agent of the judgment. ἔλεος stands bare. This is the central exegetical question of the verse, and it is old.

The reading text therefore keeps the bare noun with no possessive supplied. ‘God’s mercy triumphs over judgment’ and ‘your mercy will triumph at the judgment’ each adjudicate. So, more subtly, does a capital M.

What may be said without adjudicating: the verse does not make mercy a purchase price. The first clause is a negative sanction (no mercy done, no mercy received), not a positive tariff, and the second clause is a maxim about the outcome of a contest, not a statement about a transaction. Per §6, the rendering must not be shaded toward either a works-righteousness reading or a reading that empties the first clause of its threat.

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.

The clause is thoroughly Jewish in shape.

The verb-plus-object idiom is Hebrew: עָשָׂה חֶסֶד — Gen 24:12, 14 (Abraham’s servant asks God to ‘do ḥesed’ with his master); Judg 1:24; 1 Kgdms 20:8; 2 Kgdms 9:1; Zech 7:9 (ἔλεος καὶ οἰκτιρμὸν ποιεῖτε ἕκαστος πρὸς τὸν ἀδελφὸν αὐτοῦ). The LXX renders it with ποιέω + ἔλεος consistently enough that the collocation is a translation-Greek signature.

The measure-for-measure principle behind the first clause is likewise standard: the one who withholds gets withheld from. Prov 21:13 LXX — ὃς φράσσει τὰ ὦτα τοῦ μὴ ἐπακοῦσαι ἀσθενοῦς, καὶ αὐτὸς ἐπικαλέσεται καὶ οὐκ ἔσται ὁ εἰσακούων. Sir 28:1-4 develops it at length. The Matthean form (Matt 5:7; 6:14-15; 18:23-35) is the closest NT parallel, and James’s clause reads as the negative of the beatitude — whether by direct dependence on the Jesus tradition or as shared sapiential currency is not determinable from the wording alone.

The pairing ἔλεος / κρίσις is itself scriptural. Ps 100:1 LXX sets them side by side as the twin themes of praise; Mic 6:8 LXX has ποιεῖν κρίμα καὶ ἀγαπᾶν ἔλεον; Zech 7:9 pairs κρίμα δίκαιον with ἔλεος. James is not inventing an antithesis but working a familiar pair.

Note what he does not do: he does not cite. There is no γέγραπται, no λέγει ἡ γραφή. The verse is proverbial in register, not quotational, and the English should read as a maxim rather than as a citation.

Proposed — pending reviewMartin Dibelius, James: A Commentary on the Epistle of James, rev. Heinrich Greeven, Hermeneia, Philadelphia: Fortress, 1976.

Three points, none of them large, all worth recording.

ἀνέλεος. 𝔓74 ℵ B C read ἀνέλεος; A and the Byzantine tradition read ἀνίλεως or ἀνηλεής. NA28 prints ἀνέλεος as best attested among the early witnesses and as the form least likely to be produced by scribal normalization toward commoner adjectives. No semantic consequence.

κατακαυχᾶται. A number of witnesses read κατακαυχάσθω (third-person present imperative middle: ‘let mercy boast over judgment’), and some Latin evidence reflects a similar reading. The imperative converts a maxim about how things stand into an exhortation, which is a considerable change in force. The indicative is the harder reading in context — the whole verse is grounding (γάρ) rather than commanding — and is better attested; the imperative is best explained as assimilation to the imperatival environment of v. 12. Adopt the indicative and record the variant.

The connective. A minority of witnesses supply a connective (δέ or καί) between the clauses. As with the several other asyndeton points in this corpus, the supplied conjunction is the smoothing to be avoided: it shows scribes feeling exactly the abruptness the critical text preserves, and it settles the clause-relation question by fiat. Follow NA28’s asyndeton.

The Fathers divide along the same line the modern commentaries do, which is itself evidence that the ambiguity is in the Greek rather than in later theology.

The Greek catena tradition on James (represented in the fragments attributed to Didymus and in the Catena of Andreas) generally takes the mercy as God’s and reads the clause as a statement about the divine character at the assize: the Judge’s own mercy outweighs his sentence. This reading fits the anarthrous abstract noun and the LXX pairing of ἔλεος and κρίσις as divine attributes.

The Latin tradition, following the Vulgate’s superexaltat autem misericordia iudicium, tended to hear the merciful person’s works in view, in step with the strong Western emphasis on almsgiving as availing at judgment (already visible in the Shepherd of Hermas and 2 Clement). Note that superexaltat is a good rendering of the κατά- compound’s force — Jerome heard the exultation, not merely the victory — but it makes the clause transitive with mercy as agent, which the Greek genitive does slightly differently.

What both traditions agree on, and what should be preserved: this is a statement about an outcome, not a formula for procuring one. Neither the Greek nor the Latin exegesis of the early centuries reads the verse as making acts of mercy the ground of acquittal in the way that later polemics on both sides assumed. Per §11, that patristic weight is a matter of reception and is named here rather than used as evidence for the rendering.

Proposed — pending reviewGerald Bray (ed.), James, 1-2 Peter, 1-3 John, Jude, Ancient Christian Commentary on Scripture: New Testament XI, Downers Grove: InterVarsity, 2000.

Both clauses are maxims. The first is proverbial in shape (nominal, verbless, generalizing participle); the second is gnomic (present indicative, no article, four words).

The genre-consistency rule (§10) means the English must read as wisdom-saying, not as epistolary argument. That argues for terseness: short clauses, no explanatory expansion, no supplied conjunction, no supplied possessive. James chose four words for the second clause when he could have written a sentence; the English should not be longer than it has to be.

It also argues against smoothing the asyndeton. The abrupt shift from a threat to a triumph-cry, with nothing between them, is the rhetorical event of the verse. Every conjunction an English translator supplies is a small act of explanation, and this verse is one that gains its force by declining to explain.

Proposed — pending reviewRichard Bauckham, James: Wisdom of James, Disciple of Jesus the Sage, New Testament Readings, London: Routledge, 1999.