--- title: The Grammar Doesn't Support 'A God' url: https://jw-apologetics.app/arguments/colwells-rule-john-1-1/ markdown: https://jw-apologetics.app/arguments/colwells-rule-john-1-1.md type: argument perspective: Protestant (historic Nicene) response to Jehovah's Witness teaching topics: [Deity of Christ, NWT Translation] challenges: [Jesus Is a Created Being, The New World Translation Is the Most Accurate Bible, The Trinity Is a False Pagan Doctrine] verses: [John 1:1, John 1:14, John 20:28, Colossians 2:9, John 1:18, John 1:6, Romans 1:21] strength: Moderate tone: Excellent attribution: North Avenue Church — Biggest Errors of JW Teaching sermon added: '2026-05-24' last_reviewed: '2026-05-24' index: https://jw-apologetics.app/llms.txt --- # The Grammar Doesn't Support 'A God' *John 1:1c — Colwell, Harner, and the construction the NWT mishandles* ## Doorstep Question > "In John 1:1, your NWT reads 'the Word was a god.' I want to ask one honest question about the Greek behind that 'a.' Where in the Greek does the indefinite article come from? Because Greek doesn't have an indefinite article — and the construction John actually used is one that Greek scholars who have catalogued every example in the New Testament say doesn't read 'a god' anywhere else either." ## JW Claim The Greek of John 1:1c does not have the definite article before theos ('god'). Where the article is missing in Greek, the word can be rendered indefinite — 'a god.' The NWT is being precise about the Greek where other translations smooth it over. The first 'God' in the verse has the article (ho theos = 'the God,' Jehovah); the second 'god' does not (theos alone = 'a god,' a divine being lesser than Jehovah). ## Strategy Don't lead with the name 'Colwell's Rule' at the door — the technical label will lose the room. Lead with what was observed. In 1933 a Greek scholar named E. C. Colwell catalogued every place in the New Testament where a noun like 'God' or 'Lord' appears as the predicate of a 'was/is' verb and comes *before* the verb — the exact word order John uses in John 1:1c (theos en ho logos — literally 'God was the Word'). What Colwell found is that in this specific construction, the noun normally lacks the article *even when it is definite*. In plain English: the Greek dropped 'the' as a matter of word order, not as a marker of indefiniteness. So the NWT's logic — 'no article, therefore indefinite, therefore "a god"' — is the move Colwell's data specifically rules out. Then add the honest caveat: in 1973 Philip Harner refined this for the John 1:1 construction, arguing it is *qualitative* — 'the Word had the very nature of God.' Note carefully: Harner refines Colwell but agrees on the negative point — *indefinite* ('a god') is not what the construction says. Whether you take it as definite (Colwell) or qualitative (Harner), the NWT's 'a god' is the one option Greek scholarship rules out. Close by showing the other evidence: the same anarthrous *theos* construction is used of the Father in John 1:18, in Romans 1:21, and elsewhere — no translator renders those 'a god.' And John 20:28 has Thomas address Jesus as 'my Lord and my God' (ho kurios mou kai ho theos mou) — *with* the article — and Jesus accepts the title. ## Key Insight The NWT's argument is 'no article means indefinite.' That single inference is what Greek scholars across the theological spectrum reject. Whether the predicate is definite or qualitative, it is not indefinite. The grammar doesn't support 'a god.' ## JW Responses & Counters ### Response 1 **JW says:** "Colwell's Rule has been criticised by scholars — even Bruce Metzger acknowledged limitations." **Counter:** Yes — Colwell's Rule has been refined. The most important refinement is Philip Harner's 1973 study in the Journal of Biblical Literature, which argued that the John 1:1 predicate is best read as *qualitative* — 'the Word had the very nature of God' — rather than strictly definite. Here's the thing: that refinement still excludes the NWT reading. Definite and qualitative both deny that John is saying 'a god.' The scholars criticising Colwell are not criticising him toward the NWT's direction; they're criticising him in the opposite direction. The NWT's 'indefinite' reading is the one option no major Greek scholar defends. ### Response 2 **JW says:** "Other anarthrous nouns in John 1 (verses 6, 12, 13, 18) are translated indefinite — why not verse 1?" **Counter:** Look more carefully at those verses. John 1:6 — 'there came a man sent from God' — *God* (theos) there is anarthrous and translated definite by every standard version, including the NWT itself. John 1:18 — 'no one has ever seen God' — anarthrous *theos*, translated definite. The pattern across John is that anarthrous *theos* regularly refers to the one God; it is not 'a god' in any of those occurrences. Translators do not treat anarthrous *theos* as indefinite anywhere else in John. Why does the NWT make the indefinite move uniquely at John 1:1c? Because the theology requires it, not because the grammar does. ### Response 3 **JW says:** "The Kingdom Interlinear's word-for-word column reads 'god was the Word' — lowercase. So 'a god' is the natural reading." **Counter:** The Kingdom Interlinear's word-for-word column is exactly what you expect from a literal gloss — lowercase 'god' because Greek has no capitalisation. That gloss does *not* support adding an indefinite article. The English line above the literal gloss in the Kingdom Interlinear itself reads 'the Word was a god' — but the literal Greek-to-English line under it does not have the 'a.' The 'a' was added in translation, not transferred from the Greek. ### Response 4 **JW says:** "Even Bruce Metzger said Colwell's rule doesn't strictly prove deity." **Counter:** Correct — and Metzger's own assessment of the NWT is the more important quote. In his article on the NWT, Metzger called the rendering 'the Word was a god' a 'frightful mistranslation,' 'erroneous,' and 'pernicious.' He's the Greek scholar the Watchtower itself selectively quotes — but his actual published judgement on John 1:1 is that the NWT is wrong. The internal grammar discussion among scholars is whether John means 'definite' or 'qualitative.' Nobody on the scholarly side argues for 'indefinite.' ### Response 5 **JW says:** "If John wanted to say 'the Word was God,' he could have written ho logos en ho theos." **Counter:** He couldn't, actually — and that's the heart of the matter. If John had written ho logos en ho theos ('the Word was the God'), with the article on theos, he would have made the Word and the Father interchangeable — a heresy called Sabellianism. By *omitting* the article on theos and putting it before the verb, John says two things at once: the Word is fully God in nature (the predicate is qualitative/definite), but the Word is not the Father (the article distinguishes them). The grammar threads the Trinitarian needle precisely. 'A god' destroys that precision and replaces it with polytheism. ## What to Avoid Don't start a Greek lecture. The doorstep version is one sentence: 'Greek scholars who have catalogued every example of this exact construction in the New Testament show that "a god" is the one reading the grammar doesn't support.' Save Colwell, Harner, and Metzger for the JW who pushes back. Also avoid arguing that the NWT translators were ignorant — they were not; the rendering is theologically driven, not grammatically driven. ## Homework Open your NWT to John 1:6, 1:18, and 1:1c. Notice that all three have anarthrous *theos*. Notice that the NWT translates the first two definite ('God' / 'God') and only the third indefinite ('a god'). Ask yourself: what rule is being applied that produces three different translations of the same grammatical pattern? ## Next Moves - [Two Verbs, Two Categories](https://jw-apologetics.app/arguments/en-vs-egeneto-john-1-tense.md) - ["My God and Your God" — John 20:17](https://jw-apologetics.app/arguments/john20-17-my-god.md)