John 1:1c: Grammar, Manuscripts, Camps

John 1:1c has been debated for the entire history of Christian theology. Every major theological tradition has had to say something about it. The verse is twelve Greek words long and every one of them carries weight.

καὶ θεὸς ἦν ὁ λόγος

kai theos ēn ho logos

The traditional English rendering "and the Word was God" obscures several things about the Greek that a serious reader should know. The word order is not "the Word was God" but "God was the Word." The noun theos ("God") is anarthrous (without the definite article). The subject ho logos ("the Word") has the definite article. The verb ēn ("was") is imperfect tense. Each of these features carries interpretive weight, and the cumulative decision about how to translate the clause depends on what one does with all of them together.

This page walks through the grammar in detail, the manuscript evidence, and the engagement with mainstream evangelical scholarship. It is a long page because the question is a serious one and shortcuts do not work.

The grammar of anarthrous predicate nouns

The core grammatical question is what to do with theos in John 1:1c. In Greek, a noun used as a predicate (saying what the subject is) can appear either with the definite article or without it. The presence or absence of the article is not random. It carries meaning, and the meaning matters.

When a predicate noun has the article, it usually identifies the subject as the particular instance of its class. "The man is the teacher" (ho anthrōpos estin ho didaskalos) identifies the man as the specific teacher under discussion, not just as a teacher in general.

When a predicate noun lacks the article, it can function in several ways. It can be indefinite ("a teacher"), definite ("the teacher," where context makes the article unnecessary), or qualitative ("characterized by teacher-ness," emphasizing the nature or character of the subject). Which of these three functions is operating in any given sentence depends on the sentence's structure, word order, and context.

In 1973, the Greek scholar Philip B. Harner published a landmark article in the Journal of Biblical Literature, volume 92 (pages 75 through 87), titled "Qualitative Anarthrous Predicate Nouns: Mark 15:39 and John 1:1." Harner examined every relevant construction in the Fourth Gospel and in Mark. His findings were significant and have shaped subsequent grammatical analysis.

Harner's key observation is that in the Fourth Gospel, when a predicate noun lacks the article and precedes the verb (the construction in John 1:1c, where theos precedes ēn), the noun is almost always qualitative in function, not indefinite or definite. Harner counted fifty-three such constructions in the Fourth Gospel. Of these, forty convey qualitative force; thirteen convey definite force in context; zero convey indefinite force ("a god" rather than "God" or "what God is"). The statistical pattern is clear.

The implication for John 1:1c is substantial. The construction theos ēn ho logos places the anarthrous theos before the verb, exactly the pattern Harner identified as qualitative. The most grammatically defensible reading is not "the Word was God" (which suggests identification) or "the Word was a god" (which suggests indefiniteness), but rather "what God was, the Word was," or "the Word was fully what God is," or "the Word shared the nature of God."

Colwell's rule and its misapplication

A common Trinitarian response to Harner's analysis appeals to a rule formulated by E. C. Colwell in 1933. Colwell observed that definite predicate nouns that precede the verb tend to drop the article. On this reading, theos in John 1:1c is definite despite lacking the article, because it precedes the verb. And if theos is definite, the translation "the Word was [the] God" can be defended.

This application of Colwell's rule is a methodological error. Colwell's rule describes what happens to definite predicate nouns; it does not establish that any anarthrous predicate preceding a verb is definite. Colwell himself did not draw that inference. The inverse of Colwell's rule (an anarthrous preverbal predicate is definite) is a logical fallacy, the inverse error. Harner's analysis specifically addresses this mistake and demonstrates with statistical evidence that qualitative force predominates in these constructions in John.

Daniel Wallace, a prominent evangelical Greek grammarian, has addressed this in his Greek Grammar Beyond the Basics (1996). Wallace acknowledges the inverse-error problem with Colwell's rule and agrees that the qualitative reading of John 1:1c is the most grammatically defensible. Wallace's position is that theos in John 1:1c is qualitative, indicating that the Word shared the nature of God, without being identical to the Father who is ho theos. He holds this together with a Trinitarian theological commitment: the Word shares the divine nature (qualitative), and yet is a distinct Person from the Father (not identified with ho theos). This is an internally coherent Trinitarian reading of the verse.

Wallace's grammatical conclusion is exactly the conclusion Biblical Unitarians draw. Theos in John 1:1c is qualitative. The Word shared the nature of God. The difference is what theological framework this qualitative sharing implies. Wallace fills the qualitative sharing with Nicene Trinitarian content. Biblical Unitarianism fills the qualitative sharing with Memra content: the Word shared the nature of God because the Word was God's own self-expression, not a second divine Person.

What the qualitative reading actually allows

The qualitative reading does not force any particular theological conclusion. It rules out two extremes.

It rules out the Jehovah's Witness "a god" reading. The New World Translation renders John 1:1c as "the Word was a god," treating theos as an indefinite predicate and making the Word a lesser created divine being. Harner's analysis shows this is not the predominant function of the construction. The indefinite reading is grammatically possible but statistically unusual and contextually strained.

It rules out the simple identification reading. If the translator intends to identify the Word with ho theos (the Father), the Greek has a straightforward way to do that: use the article on theos and produce theos matching ho theos of verse 1b. The fact that John does not use the article on theos in 1:1c signals that John is not making a simple identity claim. Something more nuanced is being said.

What the qualitative reading allows is a range of theological interpretations, all of which share the feature that the Word shared the nature of God without being identified with the Father as a particular Person.

Nicene Trinitarian reading. The Word is a distinct divine Person who shares the one divine substance (ousia) of the Father. The Son and the Father share the same divine nature. The Son is distinct from the Father as a Person but identical in substance.

Biblical Unitarian Memra reading. The Word is God's own self-expressive agency, not a distinct divine Person at all. The Word shares the nature of God because the Word is God speaking. God's own speech has the nature of God because it is God's speech. When this Word takes flesh in Jesus, God's own self-expression has become embodied in a human being.

Biblical Unitarian Proclaimer reading. An original synthesis proposed by some Biblical Unitarian scholars (notably Dustin Smith in his Fourth Gospel work), suggesting that theos in 1:1c refers specifically to God's role as the one who proclaims or announces his purposes. The Word is "what God was" in his proclaimer-capacity. This reading is still being developed and is not yet established scholarship in the wider academic world.

All three of these readings are grammatically defensible. The decision among them depends on which theological framework one brings to the text. Biblical Unitarianism has textual reasons (the Memra background, the first-century Jewish setting, the consistent Father-Son distinction throughout the Fourth Gospel, John 17:3 as the anchor) to prefer the Memra reading.

The Coptic evidence

An interesting piece of evidence from the manuscript tradition is the early Coptic translation of John 1:1c. Coptic is the language of Egyptian Christianity, and the Sahidic Coptic version of the New Testament dates to the second or third century. The Sahidic Coptic renders John 1:1c as:

ⲁⲩⲱ ⲛⲉⲩⲛⲟⲩⲧⲉ ⲡⲉ ⲡϣⲁϫⲉ

Transliterated: auō neunoute pe pshaje

Translated: "and the Word was a god" or, more precisely in Coptic grammar, "and the Word was god-ish" or "divine-in-nature."

The Sahidic Coptic uses an indefinite article before noute ("god"), producing a reading that corresponds to "a god" in English. This has sometimes been cited by Jehovah's Witnesses as manuscript support for their translation.

The honest reading of this evidence is more nuanced. Coptic does not have exactly the same grammatical apparatus as Greek; its indefinite article can carry qualitative as well as indefinite force. The best modern Coptic scholarship (James Brown, Dirk Jongkind, and others) suggests the Sahidic reading falls into the qualitative-with-indefinite-flavor range. It is closer to "the Word was divine" than to "the Word was a god" in the sense of a lesser created deity.

The Coptic evidence does support Harner's conclusion that theos in John 1:1c is qualitative. The early Egyptian church, reading John in its own Coptic translation, understood the verse to be saying something qualitative about the Word's nature, not identifying the Word as the one God.

The manuscript tradition and textual criticism

There is no significant textual variation in John 1:1c in the Greek manuscript tradition. The verse reads essentially the same in the major witnesses. Papyrus P66 (dating to c. 175-225 CE), Papyrus P75 (c. 175-225 CE), Codex Vaticanus (4th century), Codex Sinaiticus (4th century), and Codex Alexandrinus (5th century) all read kai theos ēn ho logos. The variant tradition is not where the question lies. The question is entirely at the level of translation and interpretation of a textually stable verse.

This is a significant point. Some theologically contested verses have manuscript variants that affect the reading (for example, John 1:18 and 1 John 5:7-8, both of which are treated in other deep dives on this site). John 1:1c does not. The Greek text is what it is. The debate is about what the Greek text means.

Murray Harris and the question of theos for Jesus

Murray Harris's Jesus as God: The New Testament Use of Theos in Reference to Jesus (1992) is the most thorough evangelical treatment of the question of whether Jesus is called theos in the New Testament. Harris examines every passage where theos might refer to Jesus and evaluates each on its grammatical and contextual merits.

Harris concludes that there are seven passages in which theos certainly or probably refers to Jesus: John 1:1, John 1:18, John 20:28, Romans 9:5, Titus 2:13, Hebrews 1:8, and 2 Peter 1:1. He is cautious about others (John 20:17, 1 John 5:20) and concludes that the New Testament does, in a limited set of passages, use theos of Jesus.

Harris's work is honest and careful. What it establishes is that theos is occasionally applied to Jesus in the New Testament. What it does not establish, and what Harris himself is careful not to overstate, is what the application means. In the qualitative reading of John 1:1c, theos is applied to the Word not as an identification of the Word with the Father but as a qualitative description of the Word's sharing in the nature of God. The same qualitative framework can be applied to the other passages Harris identifies, in most cases without requiring the Nicene ontological reading.

Biblical Unitarians do not need to deny that the New Testament sometimes calls Jesus theos. Biblical Unitarians need only insist that this application is the Shaliach's application, not the identification of Jesus with the one true God. The Messiah is the authorized representative of the one true God, and as such he can be addressed with divine titles (theos, kurios) without being the one true God himself. This is the Old Testament's own pattern. Moses is called elohim in his representative capacity (Exodus 7:1). The Davidic king is called elohim in his royal capacity (Psalm 45:6). The judges of Israel are called elohim when they represent God's justice (Psalm 82:6). The application of divine titles to authorized human agents is a Hebrew Bible pattern that the New Testament continues when it applies these titles to the Messiah.

The Harner-Wallace convergence

A striking feature of contemporary New Testament Greek scholarship is the broad agreement, across theological lines, that theos in John 1:1c is qualitative. Harner established this in 1973 for mainstream scholarship. Wallace confirmed it for conservative evangelical scholarship in 1996. Most mainstream Greek grammars now teach the qualitative reading as the default.

What the scholars disagree about is the theological payload of the qualitative reading. A Trinitarian reads "the Word shared the nature of God" and hears "the Son is a distinct divine Person of the same substance as the Father." A Biblical Unitarian reads "the Word shared the nature of God" and hears "the Word is God's own self-expressive agency, sharing God's nature because it is God's speech."

The grammar does not settle the theology. The grammar rules out several theological positions (the NWT reading, the simple-identity reading). Within the range the grammar allows, theological commitments decide what the qualitative sharing means.

This is worth saying clearly. Biblical Unitarianism does not depend on a contested Greek reading that only Biblical Unitarian scholars accept. The Greek reading is the mainstream reading. What Biblical Unitarianism does is fill the qualitative space with Memra content rather than Nicene content. The filling is a theological decision, and the case for the Memra filling is the case made in the Prologue study, in the Shema study, and throughout the positive case zone.

Interacting with Daniel Wallace's Trinitarianism

Because Wallace is so prominent in conservative evangelical grammar, and because his analysis of John 1:1c agrees at the grammatical level with Biblical Unitarians, his theological move from grammar to Nicene Trinitarianism deserves specific attention.

Wallace argues that the qualitative reading ("the Word shared the nature of God") requires a Trinitarian framework because the qualitative sharing of divine nature is what Nicene Trinitarianism affirms. The Word is of the same divine essence as the Father (homoousios), and this is what "qualitative sharing" describes.

The Biblical Unitarian response is that "qualitative sharing of divine nature" does not require a Nicene Trinitarian framework. The Memra framework accounts for qualitative sharing without a second divine Person. God's own speech shares God's nature because it is God speaking. The qualitative sharing is the reflective sharing of a self-expressive agency, not the ontological sharing of two distinct Persons.

Wallace's move from qualitative grammar to Nicene theology is not required by the grammar. It is a theological reading applied to the grammar. A Biblical Unitarian can accept the grammar entirely and reach a different theological conclusion, because the Memra framework is available as an alternative filling of the qualitative space.

This means that engaging Wallace at the grammatical level alone cannot resolve the debate. The debate has to be engaged at the level of the frameworks that fill the qualitative space. Which framework is more consistent with the rest of the Johannine corpus, with the Shema, with the New Testament's sending-language, with the apostolic confession in Acts? That is the argument Biblical Unitarianism makes, and it is made throughout the positive case zone of this site.

What this page does not settle

This page does not settle the question of Christology. It does establish several things.

First, the Greek of John 1:1c is textually stable. There is no manuscript variant that affects the translation significantly.

Second, the best grammatical reading of the Greek is qualitative. Harner's analysis is broadly accepted. Wallace confirms it. The mainstream Greek grammars teach it.

Third, the qualitative reading does not force a Nicene Trinitarian framework. It is compatible with Trinitarianism, but it is also compatible with the Biblical Unitarian Memra framework. Theological commitments, not grammar, decide between these.

Fourth, the Biblical Unitarian case does not depend on fringe grammar. It depends on the mainstream grammar filled with Memra content. A reader can accept Harner and Wallace and still reach the Biblical Unitarian conclusion.

The decision between Trinitarian and Biblical Unitarian readings of John 1:1c is made on the basis of the frameworks one brings to the verse. The Prologue study makes the case for the Memra framework. This deep dive provides the grammatical foundation that the Prologue study builds on.