James has just called his readers μοιχαλίδες (v. 4) — adulteresses — and defined friendship with the world as enmity with God. Verse 5 presses the charge with a second rhetorical question, and then does something it does nowhere else in the letter: it introduces a formal citation (ἡ γραφὴ λέγει) whose source cannot be located in any known text, Hebrew or Greek.

That is the first of three difficulties, and they compound:

1. No identifiable source. No verse in the LXX, the MT, or the surviving Jewish literature reads πρὸς φθόνον ἐπιποθεῖ τὸ πνεῦμα ὃ κατῴκισεν ἐν ἡμῖν.
2. The subject of ἐπιποθεῖ is unexpressed. Either God longs (and τὸ πνεῦμα is his accusative object) or the spirit longs (and τὸ πνεῦμα is its nominative subject). Both are grammatically clean.
3. πρὸς φθόνον uses a word that is never positive in Greek. If God is the subject, the verse predicates envy of God — which is why it has been read as jealousy instead; if the human spirit is the subject, the phrase describes the bent of the spirit God himself put in us, which raises its own problem.

The surrounding argument does supply a control. Verses 1-4 are about grasping — ἐπιθυμεῖτε (v. 2), ζηλοῦτε (v. 2), asking in order to spend on pleasures (v. 3) — and v. 4 casts the whole thing as marital infidelity. Verse 6 answers with μείζονα δὲ δίδωσιν χάριν, God’s greater grace, quoting Prov 3:34. A verse about God’s jealous claim on what belongs to him fits between adultery and grace exactly. A verse about the human spirit’s envious tendency also fits, as a further diagnosis of vv. 1-3. Neither reading breaks the paragraph, which is why the question has stayed open for eighteen centuries.

The rendering therefore has to be built so that the reader can see the seam rather than receive a resolved paraphrase (§4). The reading text adopts the majority construal (God as subject, jealously) and the apparatus states the alternative; the literal tier keeps the word order and the case ambiguity visible.

Built onbdag

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

ἢ δοκεῖτε ὅτι is a recognizable rhetorical unit: the disjunctive ἤ does not offer an alternative but presses the hearer who has resisted the previous question (cf. Rom 6:3 ἢ ἀγνοεῖτε; 1 Cor 6:9 ἢ οὐκ οἴδατε; Matt 26:53 ἢ δοκεῖς). It follows οὐκ οἴδατε in v. 4 and functions as its reinforcement: you do know this — or do you imagine Scripture is idle?

The expected answer is no. Because Greek manuscripts carry no punctuation, the extent of the question and the boundary of the quotation are both modern editorial decisions and must be reported as such rather than as manuscript evidence. NA28 closes the question at the end of the verse and opens the citation at Πρός. Two other arrangements have been printed:

• The question ends at λέγει, and Πρὸς φθόνον … begins a new sentence which is James’s own comment, not a citation at all.
• The citation formula governs nothing that follows, and the promised quotation arrives only in v. 6 (Prov 3:34) — the intervening clause being James’s parenthetical remark.

The last is the least strained solution to the no-source problem and is a serious minority position, but it requires reading λέγει in v. 6 as resuming a formula suspended across a clause and a half.

Built onbdf · na28

NT hapax. Adverb to κενός, already fixed in this corpus as ‘empty / hollow’, with the concrete sense (an empty vessel, an empty hand) prior to the moral one.

The force is without achieving anything — ineffectually, to no purpose. LXX usage supports this: the cognate εἰς κενόν is the standing idiom for fruitless effort (Isa 65:23; Lev 26:20), already fixed in this corpus as ‘in vain / into emptiness’.

What is denied is that Scripture speaks pointlessly, not that it speaks falsely. Rendering ‘in vain’ in modern English risks the falsity reading, since ‘in vain’ has drifted toward futility-of-outcome; and ‘for nothing’ is too colloquial. ‘To no purpose’ carries the ineffectuality without importing untruth.

The adverb is emphatically fronted before its verb. English cannot reproduce the fronting inside the ὅτι clause without strain, so the wooden tier keeps ‘emptily … says’ and the emphasis is recorded here.

Built onbdag · muraoka-lxx

The corpus policy on ἡ γραφή already stands: articular singular denotes a specific passage, and where it governs a verb of speaking (Gal 3:8; 4:30; Rom 9:17) the personification is the point. Capitalize, retain the article, never render ‘the passage’ or ‘the text’.

That policy holds here — but it sharpens the problem. James is not saying it is written somewhere; he is producing a specific speaking authority. And no source text answers.

The main proposals:

• A lost or non-canonical writing. Origen and several later Fathers name an Apocalypse of Elijah or comparable work; Jerome knew of the difficulty. Nothing survives that matches.
• A free summary rather than a quotation. James condenses the biblical theme of God’s jealousy for his people — Exod 20:5; 34:14; Deut 32:16-21; Zech 8:2 — into one sentence. The absence of any citation formula elsewhere in the letter except at 2:8, 23 and 4:6 makes a loose gathering-up plausible, and Jewish authors of the period do sometimes cite Scripture by gist.
• The formula points forward to v. 6 (Prov 3:34), with Πρὸς φθόνον … as James’s own words.

Built onbdag · na28

This is the hinge, and the corpus already has a decision on the noun that makes it hard.

φθόνος is fixed as ‘envy’, and the rationale is explicit: strictly the pain at another’s good and the wish to see it lessened; no positive register in Greek at any period. It is deliberately kept distinct from ζῆλος (‘zeal / jealousy’), which is object-dependent and can be positive, precisely because ζῆλος can be predicated of God (Exod 20:5 LXX ζηλωτής) and φθόνος cannot.

So if God is the subject of ἐπιποθεῖ, the verse says God longs πρὸς φθόνον — and the word for godly jealousy is the one word James did not use.

Two constructions of πρός + acc. are on offer:

1. Adverbial, of manner or degree — ‘jealously’, ‘with a jealous intensity’. πρός + acc. does form adverbial phrases of this kind (πρὸς ὀργήν, ‘angrily’; πρὸς βίαν, ‘forcibly’; cf. πρὸς ὀλίγον, ‘for a little’, in Jas 4:14 itself). On this reading the noun is being stretched past its normal register to name a fierce proprietary longing.
2. Directional or telic — ‘toward envy’, i.e. the spirit’s longing tends in the direction of envy. This construction is already fixed in this corpus at 1 Pet 4:12 (πρός + acc. telic with abstract nouns, ‘with a view to’) and matches the covetousness of Jas 4:1-3 exactly, where ἐπιθυμεῖτε and ζηλοῦτε are the diagnosis.

What is excluded in every tier: rendering φθόνος as ‘zeal’ (that is ζῆλος, and the substitution hides which word James wrote) and any paraphrase that deletes the noun (‘God longs intensely’).

Built onbdag · bdf

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

ἐπιποθέω is already fixed in this corpus as ‘long for / yearn after’, with the note that ἐπι- is intensive rather than adding a separate sense, that the verb is Paul’s standard affection-word (Phil 1:8; Rom 1:11; 2 Tim 1:4), and that it carries LXX devotional colouring (Ps 41:2 LXX, the deer panting for water streams).

That gloss holds here without modification. Two points bear on this verse specifically:

• The verb is affective, not volitional. It names an ache, not a decision. This is why it sits so oddly with φθόνος on the manner-reading — the longing is warm, the noun is cold.
• The subject is unexpressed, which is the whole crux. Greek is content to leave a third-singular subject implicit where context supplies it. Here context does not settle it.

Render with a yearning-verb. ‘Desire’ is too weak and too volitional (the standing rationale); ‘crave’ imports the ἐπιθυμία family, which James has just used pejoratively in v. 2; ‘is jealous over’ deletes the verb and adopts one reading of πρὸς φθόνον in the same stroke.

Built onbdag · muraoka-lxx

τὸ πνεῦμα is neuter, so nominative and accusative are formally identical. Everything turns on which it is.

Reading A — God is the subject; τὸ πνεῦμα is the object.

He (God) yearns jealously for the spirit he made to dwell in us.

The subject is supplied from ὁ θεός in v. 4 (twice) and is the natural subject of κατῴκισεν in the relative clause, which no one disputes. On this reading one subject governs both verbs and the sentence coheres: God put a spirit in us and wants it wholly for himself, so friendship with the world (v. 4) is infidelity. Verse 6a — but he gives greater grace — reads as the answering clause of the same speaker.

Reading B — τὸ πνεῦμα is the subject.

The spirit which he settled in us longs toward envy.

On this reading the sentence diagnoses rather than accuses: the human spirit as God gave it has an envious bent, which explains the πόλεμοι and μάχαι of v. 1 and the ἐπιθυμεῖτε … ζηλοῦτε of v. 2. Verse 6’s μείζονα δὲ δίδωσιν χάριν then answers a real problem with a real remedy — grace greater than the inclination — and the δέ becomes strongly adversative, which suits it well.

One further note on word order: on Reading A the object stands after its verb with the relative clause trailing (ἐπιποθεῖ τὸ πνεῦμα ὃ κατῴκισεν), which is unexceptionable; on Reading B the subject also follows its verb, which is common enough in Koine but slightly heavier. Word order is therefore a mild consideration in A’s favour, not an argument.

Built onbdf · bdag

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

The corpus already distinguishes two πνεῦμα policies: the divine Spirit (capitalized where the surrounding argument fixes the referent) and πνεῦμα with a human possessive, glossed ‘spirit’ and defined as the person considered from the side of what animates and orients them — explicitly not a detachable immaterial component and not the Holy Spirit (Phil 4:23).

At Jas 4:5 the referent is contested three ways:

• The human spirit as created — the life-breath God settled in every person (Gen 2:7; Eccl 12:7; Zech 12:1 LXX, where the same causative verb-family appears with πνεῦμα). This is the reading that makes κατῴκισεν ἐν ἡμῖν most natural and is required by Reading B above.
• The Holy Spirit given to believers — on the analogy of Rom 8:9-11 and 1 Cor 3:16, with ἐν ἡμῖν as the indwelling of the Spirit in the community. This requires Reading A’s object-construal and yields God yearns jealously for the Spirit he caused to dwell in us, which is awkward theologically (God longing for his own Spirit) and is why proponents of A often prefer the created-spirit referent even on that reading.
• The human spirit as renewed — the new life of Jas 1:18.

Policy: render lower-case ‘spirit’ in the reading text and record the pneumatological reading in apparatus. This is the same discipline already applied at 1 Pet 4:6 (σαρκί / πνεύματι), where the reading text is lower-case and the Spirit-reading is flagged rather than adopted. Capitalizing here would settle a contested question typographically.

Built onbdag

NT hapax in the form κατῴκισεν. The verb is the causative of κατοικέω (‘dwell’): to settle someone somewhere, to establish as a resident. Standard LXX vocabulary for God settling Israel in the land and for planting inhabitants in a place (Gen 47:6; Num 32:40 and the resettlement narratives of Kings).

Two features must survive in English:

1. The causative force. Not ‘which dwells in us’ but ‘which he caused to dwell / settled in us’. Render with a settling- or dwelling-verb plus a causative construction.
2. The permanence. κατοικ- vocabulary is settled residence, against πάροικος/παρεπίδημος (sojourner, resident alien) — a distinction 1 Peter builds on. What God did was install something to stay.

A significant part of the tradition reads the intransitive κατῴκησεν (‘which dwelt in us’), which removes God as the actor of the clause and makes the spirit simply resident. This is the easier reading and is textually secondary; see the textual section.

Excluded: ‘put in us’ (drops the dwelling), ‘breathed into us’ (imports ἐνέπνευσεν from Gen 2:7 LXX, a different verb), ‘gave us’ (deletes the residence entirely).

Built onbdag · muraoka-lxx

The critical text reads the causative aorist κατῴκισεν (‘he settled, caused to dwell’), with God as unexpressed subject. A substantial body of witnesses reads the intransitive κατῴκησεν (‘it dwelt’), with τὸ πνεῦμα as subject.

The two differ by one vowel and were homophones in Koine pronunciation, so the variation is easily generated in copying by ear or by eye. But it is not exegetically trivial:

• κατῴκισεν requires an external agent, supplied as God, and therefore keeps God active in the relative clause. It also makes Reading A (God as subject of ἐπιποθεῖ) considerably easier, since one subject then governs both verbs.
• κατῴκησεν removes God from the clause entirely and pushes toward Reading B, since the spirit is now the grammatical subject of the relative clause and readily becomes the subject of ἐπιποθεῖ as well.

On internal grounds the causative is the harder reading: it leaves a subject unexpressed and forces the reader to supply God, which is exactly the kind of difficulty scribes smooth. The intransitive is the easier and more obvious form, and its appeal is visible in the versional tradition, which frequently reads the spirit as simply dwelling.

The reading text follows the causative. The variant is recorded because the base text tradition here has a meaningful variant that affects sense (§4), and it must not be silently resolved.

Built onna28 · metzger-tc

The corpus has already fixed δοκέω in the formula εἴ τις δοκεῖ + infinitive as reflexive-intransitive — ‘thinks of himself that…’, not ‘appears to others’ — and as consistently marking a self-assessment about to be qualified (Phil 3:4; Gal 6:3; 1 Cor 3:18; 8:2; 14:37; Jas 1:26).

Here the construction differs (2nd plural with a ὅτι clause rather than an infinitive), but the pragmatic function is identical: the verb frames a supposition the author is about to demolish. Compare Matt 26:53 (ἢ δοκεῖς ὅτι) and Luke 13:2, 4 (δοκεῖτε ὅτι), both introducing views to be corrected.

Render ‘do you suppose’ or ‘do you think’. Excluded: ‘do you imagine’ (adds contempt the verb does not carry on its own), and any construal as ‘does it seem to you’ (which would make the addressees passive recipients of an impression rather than holders of a view).

Built onbdag

The pronoun is first plural, and the shift matters. The whole surrounding passage is second plural and adversarial — ὑμῖν, ὑμῶν, ἐπιθυμεῖτε, φονεύετε, μοιχαλίδες, δοκεῖτε — and at the citation James moves to ‘us’.

On Reading A this is unremarkable: the citation speaks of God’s claim on all his people, author included. On Reading B it is more pointed: James places himself inside the diagnosis, exactly as he does at Jas 3:1 (λημψόμεθα) and 3:2 (πταίομεν ἅπαντες), where the first plural is the load-bearing feature of the sentence.

The corpus rule for those verses applies here too: the first person must be preserved at every occurrence. Rendering ‘in you’ — as a small part of the tradition does, assimilating to the surrounding second plurals — converts a shared condition into an accusation of others and reproduces a scribal harmonization rather than the text.

On the scope of ἐν ἡμῖν: whether it is distributive (in each of us) or corporate (among us, in the community) is not marked in Greek, and the choice interacts with the referent of πνεῦμα — the corporate reading pairs naturally with the Holy Spirit construal, the distributive with the created-spirit construal. English ‘in us’ is open in the same way and is retained for that reason.

Built onbdf

The verse was already difficult in antiquity, and the patristic evidence divides along the same lines the modern discussion does — which is itself useful, since it shows the ambiguity is in the Greek and not an artifact of later theology.

• The source problem was felt early. Origen appears to have known the citation as coming from a writing outside the canon, and later writers repeat the puzzle. That the difficulty is attested from the third century tells against the modern suggestion that the citation formula is loose or careless — early readers took it as a genuine quotation and looked for the book.
• The envy-reading has good ancient support. Where the Greek commentators construe φθόνος in its ordinary sense, the verse becomes a statement about the human spirit’s proneness to envy, answered by the greater grace of v. 6. This is the natural reading for a native reader of Greek, since φθόνος of God would be jarring.
• The jealousy-reading is largely the Latin inheritance. The Vulgate’s rendering pulls toward God’s jealous longing, and the Western tradition after it has generally read the verse that way. Per §11 that resonance may be named as a point of theological-tradition interest but supplies no evidence for the rendering.

Built onbdag

Proposed — pending reviewJ. B. Mayor, The Epistle of St. James, 3rd ed. (London: Macmillan, 1913)

The verse sits in a stretch of James (4:1-10) that is prophetic-diatribe in register: rapid questions, a vocative of accusation (μοιχαλίδες), a citation, and then a sequence of ten aorist imperatives in vv. 7-10. Per §10 the connectives and the interrogative shape carry the argument and must be rendered, not smoothed.

Three consequences for the English:

1. ἤ is not dropped. The question is a second push, not a fresh topic.
2. The question stays a question. Converting to a declarative (‘Scripture does not speak to no purpose’) supplies the answer the rhetoric extracts from the hearer and dissolves the device.
3. The citation stays terse. No connective, no explanatory expansion, no source reference in the text. The abruptness of a quotation dropped in without introduction beyond ἡ γραφὴ λέγει is part of the force.

The letter’s overall register is elevated but not literary-ornate; the reading text should be plain and hard, not smoothed toward devotional cadence.

Built onbdf

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