In his new grammar book, Coptic in 20 Lessons (Peeters Leuven, 2007) Bentley Layton has a valuable comment on Coptic John 1:1c, which says auw ne.u.noute pe p.Saje. He shows it can be diagrammed in this manner:
auw = and
ne = past tense marker, "was"
u.noute = a god
pe = is
p.Saje = the Word
Thus, literally, the Coptic sentence says "and a god was the Word."
True, Layton includes the traditional English "and the Word was God." But it can be noted that the traditional translation is inconsistent with his own grammatical exposition on page 7. "A god" does not equal "God." And elsewhere, in other examples in his grammar, Layton translates the Coptic construction of indefinite article + common noun into English as "a" + noun.
So clearly, the literal translation of Coptic John 1:1c's indefinite construction is "and the Word was a god." This is especially so since John's context of the Word is that of an entity, not just an abstract idea. (Cf. Layton, page 34)
The Coptic indefinite syntactical pattern at John 1:1c does not support the traditional definite reading, "the Word was God."
It is now confirmed by yet another reputable Coptic grammatical resource that the literal translation of Coptic John 1:1c is "and the Word was a god." This is what the Coptic text actually says.
Anything beyond that is commentary and paraphrase.
Your argument begins from a genuine morphological fact and then repeatedly turns that fact into a semantic conclusion that the Coptic grammars themselves do not permit. The predicate in Sahidic John 1:1c is indeed morphologically indefinite: ⲟⲩ- is the singular indefinite article, and an interlinear analysis may therefore gloss ⲟⲩⲛⲟⲩⲧⲉ as “a-god.” No competent Trinitarian treatment needs to deny that. Nor is the denotative construal “the Word was a god” grammatically impossible. What does not follow is your much stronger claim that this is what the Coptic sentence uniquely “actually says,” whereas “the Word was God” or “the Word was divine” must be secondary theological paraphrase. That conclusion depends upon collapsing four analytically distinct levels—morphology, predicate semantics, translation, and ontology—into a single mechanical English gloss.
ReplyDeleteYour appeal to Bentley Layton actually provides the clearest demonstration of the problem. On the page of ‘Coptic in 20 Lessons’ that you cite, Layton presents the bound-group analysis ⲁⲩⲱ ⲛⲉ-ⲩ-ⲛⲟⲩⲧⲉ ⲡⲉ ⲡ-ϣⲁϫⲉ with the interlinear representation “And past tense marker-a-god is the-Word.” But on the very same page, immediately beneath that analytical gloss, his sentence-level translation is “And the Word was God.” There is no contradiction unless you assume in advance that an interlinear morpheme gloss must be identical with an idiomatic semantic translation. Layton plainly does not make that assumption.
Indeed, your diagram oversimplifies Layton's grammatical analysis when you write pe = “is.” In a Sahidic nominal sentence, ⲡⲉ is not simply an independently tense-bearing equivalent of English is. In the relevant predicate-first construction it is the nexus/pronominal element, while ⲛⲉ- provides preterite conversion. The clause is structurally a preterite-converted nominal predication, not an English copular sentence whose individual Coptic morphs can simply be replaced one by one with English words. This is precisely why a pedagogical interlinear is evidence for morphological structure, not by itself a complete theory of predicate meaning. Plumley's independent treatment of the predicate-first nominal sentence makes the same point structurally: his ⲟⲩⲛⲟϭ ⲡⲉ, morphologically something like “a-great-one is-he,” is translated idiomatically “He is great.” Sterling's morphological paradigms likewise identify ⲛⲉ- as the preterite converter in nominal/copulative constructions. Thus even before theology enters the discussion, your equation “formal ⲟⲩ- = obligatory English a” is methodologically too crude.
More importantly, page 34 of the very Layton grammar you invoke expressly contradicts your proposed semantic rule. Layton heads the relevant discussion “ADJECTIVAL PREDICATES IN THE NOMINAL SENTENCE” and states that such predicates are normally formed with the indefinite article ⲟⲩ- / ϩⲉⲛ-, which is not translated into English when the construction has adjectival meaning. He then makes the point directly with the exact noun under dispute. He says that nouns too can be used as “adjectival” predicates and gives ⲛⲟⲩⲧⲉ, “god,” as his example. His paradigm includes ⲟⲩⲛⲟⲩⲧⲉ ⲡⲉ = “he is divine.” He immediately adds that the same form is ambiguous because it may instead refer to an entity, in which case ⲟⲩⲛⲟⲩⲧⲉ ⲡⲉ = “he is a god.” He gives exactly analogous contrasts such as “human” versus “a person” and “luminous” versus “a light.”
Layton's more technical ‘A Coptic Grammar’ is even less open to the interpretation you place upon him. In §292 he explicitly distinguishes denotation from description. A denotative predicate names its subject as an instance of a class; a descriptive predicate characterizes its subject by the quality associated with the lexical class. He then states that a gendered common-noun predicate is ambiguous between those two functions and again uses the exact example ⲟⲩⲛⲟⲩⲧⲉ ⲡⲉ, which he renders both “He is a god” and “He is divine.” He further explains that descriptive predicates can be indefinite article phrases and that the two semantic functions potentially belong to gendered common nouns generally, though their distribution varies with lexical semantics.
DeleteThat evidence directly disposes of your claim that because the Logos is “an entity, not just an abstract idea,” the predicate must be denotative. You have confused the ontological status of the subject with the semantic function of the predicate. Descriptive predicates are normally predicated of entities. A man is an entity when you say “the man is human”; a person is an entity when you say “he is righteous”; God is an entity when you say “God is merciful.” Nothing about an entity-subject requires every nominal predicate applied to it to function as a class-membership expression. Layton's exact ⲟⲩⲛⲟⲩⲧⲉ example was chosen precisely to illustrate that a gendered common noun may characterize an entity rather than merely enumerate it as one specimen among others.
Nor is this some idiosyncrasy of Layton. Plisch's ‘Einführung in die koptische Sprache’ describes what he calls the “two-faced character” of the Sahidic indefinite article and, in his discussion of qualifying nominal sentences, gives the exact expression ⲟⲩⲛⲟⲩⲧⲉ ⲡⲉ, paraphrasing it as ein Göttlicher ist er, approximately “he is a divine one.” Plisch elsewhere gives the particularly transparent non-Christological control ⲟⲩⲇⲓⲕⲁⲓⲟⲥ ⲡⲉ, which he translates idiomatically Er ist gerecht, “He is righteous,” while noting literally ein Gerechter, “a righteous one.” The indefinite morphology remains completely real; what changes is the semantic function and therefore the natural target-language realization.
An Arabic-language Sahidic grammar by ʿAbd al-Ḥalīm Nūr al-Dīn provides a lexically exact corroboration from another pedagogical tradition. In his treatment of nominal predication he calls the relevant function khabar waṣfī, a “descriptive predicate,” and explicitly states that certain nouns, not merely adjectives or abstract mass terms, may function in this way while taking the indefinite article. Among his examples he specifically names NOUTE and gives ⲟⲩⲛⲟⲩⲧⲉ ⲡⲉ. His own Arabic gloss remains nominal, so one should not falsely claim that he independently translates it “he is divine”; the important fact is his grammatical classification of NOUTE itself as admissible in descriptive predication. That is precisely the point your count-noun argument says cannot happen.
The wider grammatical tradition points in the same direction. Brankaer observes that a noun used as nominal predicate often bears the indefinite article and may thereby express a quality. Allen likewise treats nouns of quality with indefinite determination as adjectival predicates. Richter's published Sahidic nominal-sentence tables include ⲟⲩ-ⲇⲓⲕⲁⲓⲟⲥ ⲡⲉ in property-oriented predication. Ernshtedt's specialist work on Coptic determination explicitly distinguishes object/entity-oriented and qualitative values within indefinite predicate constructions, while Polotsky's structural studies distinguish identifying from qualifying functions in the Coptic nominal sentence. None of those sources proves that John 1:1c MUST be qualitative; collectively they establish something fatal to your argument, namely that formal indefiniteness does not constitute a semantic rule of denotative class-membership. The interpretive question remains to be decided at the level of lexical semantics and discourse.
The same point was already understood long before the modern controversy. Ludwig Stern's 1880 ‘Koptische Grammatik’ explicitly remarks that the Coptic indefinite article can have generic force and can correspond in German not only to an indefinite article but even to a definite article or to articlelessness. koptischegrammat00ster In his treatment of nominal predicates he says that predicative adjectives regularly bear ⲟⲩ- in the singular and then translates such constructions naturally as “my God is great,” “the Lord is righteous,” “the flesh is weak,” and “I am good.” koptischegrammat00ster Thus the distinction between Coptic morphological indefiniteness and obligatory reproduction of English “a/an” is not a recent Trinitarian expedient devised to escape John 1:1.
DeleteFor balance, Schwartze's much earlier grammar should also be acknowledged. In an 1850 discussion of the contraction of preterite ⲛⲉ with indefinite ⲟⲩ, he cites John 1:1 and parenthetically glosses it eigentlich es war ein Gott, roughly “properly/literally, it was a god.” That is genuine evidence that a morphologically literal “a god” gloss has a respectable history. But Schwartze is discussing contraction, not constructing a semantic theory under which a gendered common-noun predicate can only be denotative. His observation therefore supports what a careful Trinitarian should already concede: “a god” is a grammatically available denotative construal. It does not establish the thesis you need, namely that descriptive predication is excluded.
Your use of Lambdin suffers from the same shift from a pedagogical default to an absolute semantic rule. Lambdin does indeed say that Coptic article usage “corresponds closely” to English. But the complete passage immediately calls the relationship a general correspondence and proceeds to discuss exceptions: indefinite Coptic substance nouns where English has no article and abstract nouns that may carry an article where English again uses none. The full paragraph therefore cannot be turned into the proposition “Coptic ⲟⲩ- must always appear as English a/an unless the noun is a substance or abstraction.” Lambdin says no such thing. His wording announces a useful introductory approximation and then teaches students that translation is not mechanically isomorphic. His discussion also notes that even the plural indefinite article may sometimes simply be omitted in English “as the context requires.”
You are entirely correct that Lambdin himself gives ⲛⲧⲟϥ ⲟⲩⲛⲟⲩⲧⲉ ⲡⲉ as “He is a god.” But that example establishes availability, not exclusivity. Lambdin is illustrating an ordinary indefinite nominal predicate; he is not stating that every predicate containing a gendered count noun plus ⲟⲩ- must be denotative. His example therefore does exactly what one should allow it to do: it proves that “a god” is grammatical. It cannot cancel Layton's more explicit semantic treatment of precisely the same expression, where the same ⲟⲩⲛⲟⲩⲧⲉ ⲡⲉ is assigned both “he is a god” and “he is divine.” Lambdin's elementary exercise and Layton's semantic analysis are not contradictory; the former instantiates one licensed interpretation, while the latter explains why it is not the only one.
That is why your dismissal of the substance and abstract-noun examples as a “smokescreen” misses the real argument. Those examples are not needed to prove the qualitative possibility of ⲛⲟⲩⲧⲉ. They merely show from Lambdin's own page that the slogan “corresponds closely” cannot be made into a rigid translation equation. The direct answer to your count-noun objection is much stronger: Layton uses the count noun ⲛⲟⲩⲧⲉ itself to illustrate descriptive predication; Plisch uses the same noun in a qualifying predicate; Nūr al-Dīn specifically classifies NOUTE among nouns capable of descriptive predicate function. The appeal to water and truth can therefore be removed altogether, and your conclusion still fails.
Your treatment of “literal translation” also needs greater linguistic precision. A literal morphological gloss and a literal semantic translation are not the same operation. If “literal” means “show every overt morph in the target language,” then “a-god” is an entirely legitimate classroom gloss for ⲟⲩ-ⲛⲟⲩⲧⲉ. But if “literal translation” means “represent as directly as possible what the source construction predicates in context,” then one must first determine whether the predicate is denotative or descriptive. “A god” is not semantically neutral; it selects the denotative interpretation. “Divine” selects the descriptive interpretation. Since Sahidic grammar licenses both for precisely this form, you cannot declare one “what the text says” and the other “commentary” without simply assuming the point under dispute.
DeleteLayton's own practice makes this impossible to miss. If the interlinear a-god exhausted the meaning, his immediately following “And the Word was God” would be elementary incompetence. But on page 34 he explicitly tells you why there is no inconsistency: nouns such as ⲛⲟⲩⲧⲉ may serve as adjectival/descriptive predicates while retaining ⲟⲩ-. Your argument therefore requires you to treat Layton's morpheme gloss as authoritative while treating Layton's explicit semantic explanation of the same construction as though it did not exist. That is not a sustainable reading of his grammar.
Your characterization of “the Word was God” as the “traditional definite reading” likewise confuses two different questions. A Trinitarian qualitative reading of English “the Word was God” is not equivalent to reading the Coptic as ⲡⲛⲟⲩⲧⲉ, “the God,” nor does it identify the Logos as the same person as the Father of 1:1b. Orthodox Trinitarian exegesis is precisely the opposite: 1:1b establishes personal or referential distinction—“the Word was with God”—while 1:1c predicates what the Logos is. English can use “God” predicatively without inserting the article “the.” Thus the choice is not “Coptic indefinite versus Trinitarian definite.” The real dispute is between a denotative indefinite construal (“the Word was a god”) and a descriptive deity predication naturally rendered in English as “the Word was divine” or, in traditional idiom, “the Word was God.” To call the latter “definite” merely because English capitalizes God imports English naming conventions into the Coptic syntax.
This distinction also explains why a qualitative reading is fully coherent from a Trinitarian standpoint without pretending that ⲟⲩ- itself means homoousios. Coptic grammar does not lexicalize the Nicene Creed. “He is divine” is a linguistic predication of deity or divine quality; whether that predication belongs within a doctrine of consubstantiality is a wider contextual and theological inference. But precisely the same limitation applies to your reading in the opposite direction. ⲟⲩ- does not lexicalize “created,” “ontologically inferior,” “secondary deity,” or “not Almighty.” Even if you prefer the denotative “a god,” you have not thereby obtained the Watchtower ontology. You have only classified the Logos under the semantic class denoted by ⲛⲟⲩⲧⲉ. The further claims of creaturehood and inferior divine nature have to come from somewhere else.
Your demand for “actual Sahidic Coptic Scripture” therefore does not rescue the argument. The biblical data actually demonstrate why article morphology cannot function as a theological ranking system. Sahidic passages can predicate ⲟⲩⲛⲟⲩⲧⲉ of the unique biblical God in explicitly characterizing expressions: Deuteronomy 4:24 and 6:15 describe the Lord with the indefinite divine predicate modified by “jealous”; Deuteronomy 4:31 predicates it with mercy; Deuteronomy 32:4 has ⲟⲩⲛⲟⲩⲧⲉ ⲙⲡⲓⲥⲧⲟⲥ, a faithful God; Joshua 24:19 characterizes him as a holy God; Nahum 1:2 again uses the indefinite divine predicate in describing the Lord; Baruch 3:2 directly addresses the biblical God with an indefinite divine predicate characterized by mercy. The same morphology can of course be genuinely denotative in passages such as Judges 6:31 concerning Baal or in idolatrous and polemical contexts. That contrast is exactly the point: the determiner does not itself supply the ontology; lexical modification, subject, and discourse do.
DeleteNor do those Old Testament examples need to be exaggerated into proof that every unmodified ⲟⲩⲛⲟⲩⲧⲉ is qualitative. They are sufficient to answer the claim you actually make: that indefinite article + NOUTE inherently conveys one nonspecific god among gods and therefore cannot participate in full-deity predication. It plainly can occur predicatively of the unique God of biblical monotheism. The numerical New Testament expressions in 1 Corinthians 8:6 and Ephesians 4:6 likewise use ⲟⲩⲛⲟⲩⲧⲉ in “one God” constructions referring to the Father. Wright and Ricchuiti rightly treat those as numerical cases rather than direct qualitative parallels; their value here is negative but decisive: the surface form ⲟⲩⲛⲟⲩⲧⲉ does not intrinsically mean a subordinate or additional deity.
Within John itself, your appeal to the many places where Horner uses English a proves less than you suppose. There are indeed numerous Sahidic ⲟⲩ-N predicates naturally rendered “a prophet,” “a devil,” “a murderer,” “a Samaritan,” “a sinner,” “a thief,” “a hired hand,” “a Jew,” “a king,” or “a robber.” A sound Trinitarian argument should not suppress this. It demonstrates that denotative classification is highly productive. But those are predominantly role, status, ethnonym, or behavioral class nouns. You still require the unstated premise that all gendered common nouns in the same broad predicate frame must possess identical predicate semantics. Layton's §93 expressly denies so crude a rule by making lexical semantics and context relevant to the denotation/description decision, and §292 proves the matter with ⲛⲟⲩⲧⲉ itself. The fact that “Barabbas was a robber” is naturally classificatory does not determine whether “the Logos was noute” is classificatory or descriptive any more than the English sentence “John is a doctor” determines the semantics of “John is human.”
The same Gospel furnishes a particularly instructive structural control in John 4:24: Sahidic has the morphologically indefinite predicate ⲟⲩⲡⲛⲉⲩⲙⲁ before the definite subject ⲡⲛⲟⲩⲧⲉ. Yet the proposition concerns what God is—conventionally, “God is spirit”—rather than ranking God as one spirit among many. Again, the point is not that pneuma and noute have identical lexical behavior. They plainly do not. The point is methodological: an overt ⲟⲩ- in a Sahidic nominal predicate does not by itself dictate an English class-member interpretation. The semantic work of the predicate must actually be analyzed.
Your treatment of Horner creates the same problem. Horner renders John 1:1c “… and [a] God was the Word.” His own editorial convention says that square brackets indicate words represented in Coptic but not required by English. Whatever one thinks of his final English choice, his typography therefore explicitly distinguishes the overt Coptic indefinite morphology from obligatory English reproduction of “a.” In other passages he often does translate an ⲟⲩ-N predicate with a, because in those passages he judges the predicate to be denotative. That variation is not evidence that he violated Coptic grammar at John 1:1; it is precisely what one expects if translation operates at the level of predicate semantics rather than article substitution.
DeleteThe quantitative evidence also cuts against your implicit assumption that Coptic ⲟⲩ- merely makes explicit what Greek anarthrous θεός already “literally” says. Wright and Ricchuiti examined a reconstructed set of twenty-five occurrences of Greek anarthrous nominative singular θεός relevant to Sahidic and report that twenty-one are rendered with the definite Coptic article and only four with the indefinite. Whatever judgment you make about their ultimate qualitative interpretation of John 1:1c, this distribution is sufficient to refute a mechanical rule Greek anarthrous θεός → Sahidic ⲟⲩⲛⲟⲩⲧⲉ. The Sahidic translator did not simply add ⲟⲩ- whenever Greek omitted the article. The Coptic form at John 1:1c therefore requires target-language semantic and syntactic explanation rather than being treated as a transparent graphical representation of Greek articlelessness.
The same caution is required when you say that the Greek itself can be “literally” translated “a god was the Word.” Greek has no indefinite article corresponding to English a. Supplying a before θεός is already a semantic decision about an anarthrous predicate; it is not a morpheme-for-morpheme rendering. Harner's classic study of pre-verbal anarthrous predicate nouns argued that the relevant Greek construction in John 1:1c prominently predicates quality rather than merely indefiniteness. That Greek argument does not by itself settle the Coptic, but it establishes that a qualitative source-language value was available to a Coptic translator. A translation into Sahidic must then be interpreted according to Sahidic grammar—which, as Layton explicitly demonstrates, has precisely the resources to express descriptive deity with ⲟⲩⲛⲟⲩⲧⲉ.
Your appeal to Egypt's long exposure to Koine Greek does not change this. Familiarity with Greek can make a version historically valuable; it cannot erase the grammar of the receptor language. A translator who understands Greek perfectly still has to express the source text through Sahidic nominal syntax, Sahidic determination, and Sahidic lexical semantics. Modern versional scholarship accordingly treats Coptic versions as translations whose evidential value depends upon translation technique, target-language structure, manuscript history, and the recoverable versional reading—not as encoded interlinears of Greek. Nor can the antiquity of the Sahidic tradition itself decide the Christological issue. A pre-Nicene date neither makes the translators Nicene theologians nor turns them into proto-Watchtower subordinationists. Dating is not semantics.
Your proposed burden of proof is therefore backwards in one important respect. If someone asserted that John 1:1c must be qualitative merely because it contains ⲟⲩ-, that person would indeed bear a burden of proof. But you are asserting the converse universal: that because ⲟⲩ- occurs before the count noun ⲛⲟⲩⲧⲉ, only denotative “a god” represents what the Coptic text actually says, while descriptive “divine” is a theological paraphrase. One counterexample in the grammatical system is sufficient to disprove such an exclusivity claim, and Layton provides the exact counterexample with the exact noun: ⲟⲩⲛⲟⲩⲧⲉ ⲡⲉ may mean “he is divine.” Plisch and Nūr al-Dīn corroborate the same lexical possibility in different descriptive frameworks. You therefore bear the burden of establishing that the descriptive function explicitly licensed for ⲟⲩⲛⲟⲩⲧⲉ is somehow blocked specifically in John 1:1c. Merely pointing to other passages where different nouns are naturally denotative does not meet that burden.
DeleteEven your proposed zero-article alternative does not solve the problem. Sahidic does possess zero determination, and no careful argument should deny it. But zero determination is construction-sensitive. Polotsky observes that the absolute zero-article noun appears excluded from the binary nominal sentence he analyzes, and Plisch states more categorically within his nominal-sentence system that an undetermined nominal predicate is not articleless apart from a bare infinitive. Thus the counterfactual “if the translator meant quality, he would simply have omitted the article” requires a grammatical rule that the specialist literature does not give you. More importantly, it is positively contradicted by Layton: descriptive nominal predicates themselves are normally formulated with ⲟⲩ- / ϩⲉⲛ-.
The strongest form of your case is consequently much narrower than the one you present. You can legitimately say that the Sahidic predicate is morphologically indefinite and that “the Word was a god” is a grammatically possible denotative rendering. You can point to Lambdin's “He is a god,” to Schwartze's old literal gloss, to many Johannine ⲟⲩ-N class predicates, and to the real productivity of indefinite classification in Sahidic. Those are genuine facts. What you cannot legitimately add is that this makes “a god” the unique meaning of the clause, that “divine” is merely theological commentary, that a count noun such as ⲛⲟⲩⲧⲉ cannot be descriptively predicated, or that “the Word was God” necessarily imports the definite ⲡⲛⲟⲩⲧⲉ of 1:1b. Layton's grammar directly contradicts all four moves.
The linguistic conclusion can therefore be stated without pretending that Coptic morphology alone proves Nicaea. John 1:1b distinguishes the Logos from the divine referent with whom he is; John 1:1c then predicates ⲛⲟⲩⲧⲉ of that already established Logos. Sahidic grammar allows that predicate to be descriptive rather than a second definite referring expression. In that sense, “the Word was God” can function as an idiomatic English rendering of deity predication while preserving the personal distinction established in the preceding clause. The construction does not say “the Word was the Father,” and Trinitarian theology does not claim that it does. Conversely, the Coptic indefinite article says nothing whatever about the Logos being created, ontologically inferior, or a secondary deity. Those are additional theological propositions that you have imported into the form every bit as surely as you accuse Trinitarians of importing theirs.
So the appropriate conclusion is not “the Sahidic proves the Trinity,” but neither is it “the Sahidic literally says a god and anything else is paraphrase.” The linguistic evidence is more disciplined than either slogan. The morphology is indefinite; the predicate semantics can be denotative or descriptive; “a god” is possible; “divine” is equally grammatical; and the decision between them must be made from lexical semantics, discourse, translation technique, and context. Layton says this with the exact noun ⲛⲟⲩⲧⲉ. Plisch corroborates it with the same noun. Nūr al-Dīn classifies the same noun within descriptive predication. Plumley, Stern, Brankaer, Allen, Richter, Polotsky, Ernshtedt, and the broader nominal-sentence tradition independently undermine the equation between overt Coptic indefiniteness and obligatory English “a/an.” Your argument therefore succeeds in proving a morphological fact and one legitimate semantic possibility. It fails precisely where it claims exclusivity—and it is that unsupported exclusivity, not the Trinitarian reply, that turns a grammatical observation into theological eisegesis.
Delete