Should Jesus be worshipped?

by Melody 51 Replies latest watchtower bible

  • Wonderment
    Wonderment

    aqwsed: "the anarthrous θες almost always carries a qualitative force, not an indefinite or polytheistic sense."

    ---

    You state this as a broad fact, but Grammarian D. Wallace observed:

    “It is nevertheless difficult to distinguish indefinite from qualitative nouns at times (just as at other times it is difficult to distinguish qualitative from definite nouns).”

    You seem inclined to make the same mistake that some Trinitarian scholars tend to make, namely, that anarthrous nouns in the predicate can only be qualitative, not indefinite - a theological evasion at that.

    I have a proposition for you. Share with us the number of predicate pre-verbal anarthrous nouns in the NT & LXX in the NOMINATIVE - indicating a qualitative force, and those with an indefinite sense.

  • Earnest
    Earnest
    aqwsed12345 : The use of the anarthrous θες in both the NT and patristic Greek does not automatically mean “a god” in the sense of a lesser, separate deity.

    Agreed. It depends on context.

    So the passage in Clement's Exhortations could be translated to read : "the Word of [the] God became man, that you may learn from man how man may become divine" rather than "man become God" or "become a god".

    The thrust of my short message that it is helpful to distinguish between θεός and ὁ θεός when the word "God" is used in NT and Patristics remains true.

  • aqwsed12345
    aqwsed12345
    @Wonderment

    Your response mischaracterizes both my original argument and the state of scholarly discussion concerning the semantics of anarthrous predicate nominatives in Koine Greek. While it is true that Daniel Wallace and others have recognized that the distinction between qualitative and indefinite force can, at times, be subtle or even ambiguous in certain contexts, it is a category mistake to extrapolate from the mere possibility of ambiguity to the assertion that the two functions are equally likely, or that the context of John 1:1c (or the patristic texts under discussion) is genuinely ambiguous between these readings.

    Wallace's own extensive research into anarthrous predicate nominatives explicitly concludes that in the majority of pre-verbal predicate nominative constructions, particularly those referring to persons or entities already thematically salient in the discourse, the qualitative force is primary. This is not a mere theological assumption but a result of detailed grammatical and syntactic analysis. For example, Wallace writes: "An anarthrous predicate nominative preceding the verb is primarily qualitative in meaning, indicating the nature or essence of the subject" (Greek Grammar Beyond the Basics, p. 269). While Wallace acknowledges occasional difficulty in distinguishing the two, he does not claim that the indefinite is always possible, nor that it is linguistically neutral or equally valid in every context.

    Turning to your demand for a statistical breakdown, numerous studies have been undertaken (including those by Philip Harner, Colwell, and Dixon) cataloging pre-verbal anarthrous predicate nominatives in both the NT and the Septuagint. These studies overwhelmingly show that the qualitative interpretation predominates, especially in theological and soteriological contexts. For instance, in Harner's well-known analysis, the majority of such predicates (including θεός in John 1:1c) are best understood as qualitative, highlighting the nature or essence of the subject, not its membership in a broader class or its indefiniteness.

    This is not a "theological evasion" but the consensus of Greek linguistics as applied to the Johannine Prologue. The context of John 1:1 explicitly sets up a distinction in personhood between the Logos and the Father ("the Word was with [the] God"), and then asserts the nature of the Logos ("the Word was God" in essence). The lack of the article in θεός is grammatically required by the predicate position and the rules of Greek syntax (Colwell's Rule), and in this context, an indefinite reading ("a god") is both contextually and theologically improbable. To read it as indefinite in this context is to ignore not only the grammar but the entire theological and literary context of the Prologue, which asserts the Logos' unique role as Creator and source of life and light (John 1:3-4), titles and attributes elsewhere reserved for YHWH alone in Jewish monotheism.

    Moreover, your insistence that Trinitarian scholars "tend to evade" the indefinite sense ignores the broader spectrum of scholarly opinion, which is not confined to confessional bias. Even non-Trinitarian grammarians and those outside theological circles acknowledge that the grammar and context support the qualitative sense in John 1:1c. The very fact that the definite reading is rejected by most serious scholars (precisely because the article is missing) further demonstrates the careful distinctions made by competent grammarians.

    In summary, your appeal to Wallace is misplaced: his work, and that of other scholars in the field, supports the view that the anarthrous θεός in John 1:1c is best understood qualitatively, not indefinitely. The text is not asserting that the Logos is "a god" among many, but that the Logos shares the very nature or essence of God, while remaining distinct in person from the Father. The burden of proof rests on those who wish to claim that John is teaching either polytheism or a subordinate ontological status for the Logos. The linguistic, literary, and theological evidence all support the qualitative interpretation as both contextually and grammatically correct.


    @Earnest

    Your response concedes a critical point but then subtly attempts to relativize the implications of the anarthrous θες in both the NT and patristic literature by appealing to a generic distinction between “God” and “divine.” However, this move fundamentally misunderstands both the function of θες in these theological contexts and the settled intentions of the Church Fathers. The essence of the issue is not merely semantic, as if the difference between “God,” “a god,” and “divine” were a matter of translation preference; rather, it is ontological and theological—rooted in the very heart of Christian confession.

    First, while it is certainly true that “context determines meaning,” it does not follow that θες, when used anarthrously, is best rendered “divine” in an abstract, adjectival sense, or that it is ever left so vague as to support either subordinationism or a merely qualitative reading void of true divinity. When the Church Fathers—and especially Clement—speak of man “becoming θες,” they are employing the language of theosis, which, as I noted, is participatory and relational, never essential or rivalrous. The context, therefore, does not point to an “abstract divinity,” nor to a second deity, but to a real, though always derivative, participation in the divine life. This is precisely why the Church Fathers can speak so freely of deification without lapsing into polytheism or undermining the Creator-creature distinction: the divine nature remains absolutely simple, unique, and unshared as to its essence (ousia), even as it is communicated by grace and adoption to rational creatures.

    This dynamic is central to both Nicene and pre-Nicene theology, and it stands in stark contrast to Arian readings that imagine a spectrum or hierarchy of divinity, wherein some are “gods” by lesser, contingent, or subordinate essence. The very reason the Church Fathers distinguish between θες and anarthrous θες is not to imply separate grades of deity, but rather to guard against both polytheism and modalism. In the case of John 1:1c, as demonstrated by Colwell, Harner, and a long line of grammatical and patristic authorities, the predicate θες—though anarthrous—bears qualitative force, asserting that the Logos possesses the very nature of God, not that he is a lesser or different kind of divine being. To suggest that the Word is merely “divine” in the sense of possessing godlike qualities, rather than being true God from true God, is to reintroduce precisely the ontological subordination that the Church condemned at Nicaea.

    Furthermore, your suggestion that the “distinction between θες and θες” is simply a matter of clarification in usage overlooks the specific theological grammar of both Scripture and the Fathers. When θες is applied to the Father with the article, it typically marks him as the arche, the personal source within the Trinity; when applied to the Son or Spirit without the article, it affirms their full participation in the divine essence without confusion of persons. Nowhere do the Fathers, Clement included, suggest that Christ is divine merely by courtesy or predication, nor that Christians are “divine” in any naturalistic or pantheistic sense. The entire point is that theosis is by grace, the Son’s divinity is by nature, and the difference is always maintained.

    In summary, your appeal to a generic “divinity” as the solution to the problem of θες and θες in patristic and biblical texts elides the crucial distinctions affirmed by the Christian tradition. The Church Fathers’ precision in their use of θες serves not to support the possibility of subordinate or lesser deities, but rather to articulate the mystery of the Trinity and the uniqueness of Christian monotheism: one God in three Persons, and one economy of salvation by which man is made “god by participation, not by nature.” Any exegesis or translation that attempts to dilute this distinction—by rendering the Logos as “a god” or by reducing Christian theosis to an abstract “divinity”—inevitably fails to account for both the grammar and the substance of Christian confession. The distinction between θες and θες is not a Trojan horse for Arianism, but a necessary instrument for safeguarding the paradox of unity and distinction at the heart of Christian faith.

  • slimboyfat
    slimboyfat

    In which of his books does Benjamin Sommer discuss Justin Martyr?

  • Wonderment
    Wonderment

    aqwsed: In summary, your appeal to Wallace is misplaced: his work, and that of other scholars in the field, supports the view that the anarthrous θεός in John 1:1c is best understood qualitatively, not indefinitely.

    ---
    No, my appeal to Wallace is not misplaced. I learned decades ago that Grammarians say one thing, and translators often do another. One thing is theory or theology, which Wallace as author admits being guilty of in his Preface, and another, the real world practice. He admits it is legitimate to disagree with him. Trinitarians do obviously evade the indefinite sense in pre-Verbal anarthrous nouns in Nominative constructions, as in John 1.1 for theological reasons. They appeal to other Trinitarians, of course, like Colwell, Dixon and Harner to make you believe that John 1.1 cannot be rendered as "a god." Their PhD's are supposed to seal the matter.

    The NWT offered a list of such nouns similar to the one found in John 1.1 in the discussion in the Appendix of their 1984 Bible & KIT. To this day I have not seen critics been successful proving otherwise. The critics somehow have to cling to the argument that qualitative expressions run counter to indefinite ones, a claim that does not hold up under scrutiny.

    To start, the Coptic of John 1.1c uses an indefinite expression. Wright & Ricchuiti in their review, declared that this choice indicates its qualitativeness. There is no other way for Trinitarians to say it. Will they admit that Jesus could be "a god" instead of "The God"? No, they can't help themselves. I suspect you too can come up with extensive arguments against the indefinite sense of the Coptic in writing. But there it is.

    Do mainstream scholars say unequivocably that a qualitative expression rules out a potential indefinite sense in many cases? Some may claim so, but their research does not hold up. The list the WT provided in the Appendix cannot be undone. One can claim that the samples provided do not support such conclusion at John 1.1. Really? If so, grammar would not be the issue, theology, then would.

    For instance, Wallace cites John 4.19 [found in NW list], and concludes: “Although the translation [of John 4:19] is most naturally ‘Sir, I perceive that you are a prophet,’ the sense may be better characterized as indefinite-qualitative.” (Bold letters added)

    Is he correct here? Yes, he is. In English, the translator sometimes has no choice but to use the indefinite article when the original expression may be understood in a qualitative sense. Dixon admitted: “Often, the only way to effectively communicate a qualitative noun in the English idiom is by prefacing the noun with ‘a.’” This is so because translating John 4.19 literally, "I am beholding that prophet are you" is not standard English. So the "a" before "prophet" must be supplied in English to smooth things out. Incidentally, adding the indefinite article in this text is not necessary in Latin-based languages, although some translators do so.

    Example: The French translation by Segond reads in John 4.19, “Je vois que tu es prophète," with no indefinite article. It sounds just fine without one. But another French translation (by Darby) adds the indefinite meaning, “Je vois que tu es un prophète."

    In John 6.70, where the Greek says, "and out of you [in plural] one devil is," Dixon says: “It is best, therefore, to take διάβολος qualitatively. A good rendering might be: ‘one of you is a devil.’”

    And in John 8.44, the Greek reads: "That (one) man-killer was from beginning." On this, Robert Hanna wrote: The second segment of this verse should be translated 'your father the Devil was a murderer from the beginning.'" Note: Hanna makes it clear here that grammatically adding the indefinite article is acceptable. In John 1.1, he denies it, althought the construction is similar. Why the difference? Obviously, theology is the culprit. Can you see that?

    There is an axiom in Koine Greek. "Nouns without the definite article are either indefinite or qualitative.” (J. Harold Greenlee) If so, don't you think it is dishonest to claim that John 1.1 can, or should only be translated in the traditional way?

    One more thing, you and some other posters often bring out that most scholars render John 1.1, "and the Word was God." True! However, most Bible readers on the internet seem to be unaware that those translating otherwise, like, "the Word was divine," the Word was a god (or, "a divine being"), the Word was godlike, etc., are no small number.

    It is estimated that there are some 500 English versions of the Bible. The number of translators rendering John 1.1 in a non-traditional way in various languages tally up to upwards of 150 translators, most with university credentials. In English alone there are dozens of them. No hard rule, but the number may be, perhaps, one in ten English tanslations that deviate from the norm. How significant is this? That's for the reader to decide. What is clear is that a translator's PhD cannot settle this issue. More relevant to the subject is what the Bible itself teaches. We would do better in accepting that the Gospel of John was purposedly written to have the people the world over believe "that Jesus is the Christ the Son of God," not The God. (John 20.31)


  • aqwsed12345
    aqwsed12345
    @Wonderment

    Your attempt to separate the theoretical conclusions of grammarians like Wallace, Harner, and Dixon from the “real world practice” of translators misstates the issue. While it is certainly possible for translators to make theological choices, the overwhelming majority of modern, critical editions and major translations—across a wide range of confessions—do not render John 1:1c as “a god,” because the consensus of scholarship, including Wallace’s own detailed analysis, finds the qualitative force primary and contextually determinative. Wallace’s Greek Grammar Beyond the Basics (1996) provides a rigorous analysis of Koine Greek, concluding that anarthrous pre-verbal predicate nominatives, like θεός in John 1:1c, are primarily qualitative, emphasizing the nature or essence of the subject (Wallace, p. 269). His acknowledgment that disagreement is possible reflects academic humility, not a lack of confidence in his findings.

    Your assertion that Trinitarians “evade” the indefinite sense for theological reasons oversimplifies the issue. Translation practice reflects target-language idiom, not a rejection of this principle. Scholars like Wallace, Philip Harner, and Paul Dixon base their qualitative interpretation on detailed grammatical and statistical analyses, not merely theological presuppositions. For instance, Harner’s study analyzed over 250 such constructions, finding that the qualitative force predominates, particularly in theological contexts like John 1:1c. Similarly, Dixon’s thesis found that 94% of pre-copulative anarthrous predicate nominatives in John’s Gospel are qualitative, not indefinite. These studies, grounded in empirical data, contradict your claim that Trinitarian scholars rely solely on credentials to “seal the matter.” These are not theological assertions but syntactic observations, widely accepted beyond Trinitarian circles.

    The listed verses in the NWT’s 1984 Appendix involve nouns describing roles or characteristics within creation (e.g., prophets, devils, thieves), which differ significantly from the theological weight of θεός in John 1:1c, where the Word is introduced as pre-existent and the creator of all things (John 1:3). For example, in John 4:19 (“you are a prophet”), the Samaritan woman identifies Jesus as a member of the class of prophets, a contextually appropriate indefinite rendering. In contrast, John 1:1c’s context—establishing the Word’s eternal relationship with God and role in creation—demands a qualitative interpretation, as θεός describes the divine nature, not membership in a class of gods. Dixon’s statistical analysis confirms this, showing that in John’s Gospel, pre-verbal anarthrous predicate nominatives are overwhelmingly qualitative (94%), not indefinite (17%). Your reliance on these verses ignores their distinct contextual and theological roles, rendering the comparison invalid.

    In Mark 6:49, φάντασμα conveys a qualitative nature, not an indefinite class; John 4:19’s προφήτης emphasizes Jesus’s prophetic essence. John 1:1c’s θεός, tied to the Logos’s divine role (John 1:3), differs fundamentally from these category nouns, supporting a qualitative, not indefinite, sense. The Coptic ou.noute in John 1:1c, which you cite as indefinite, is better understood qualitatively per scholars like Layton, especially given the definite article for the Son in John 1:18, undermining a lesser-deity reading.

    Hence, in all these instances, the nouns in question refer to categories (a kind of being among others), and the context does not carry the unique monotheistic and Christological weight that John 1:1 bears. Wallace and others do indeed acknowledge that, for ordinary nouns, English sometimes must supply an indefinite article for reasons of idiom and clarity. However, the context of John 1:1 is entirely different: θεός in the Prologue is not a category noun, but a monadic noun: a term for the unique divine nature; and the anarthrous construction is, as all studies show, overwhelmingly qualitative in pre-verbal predicate nominative position, especially when referring to persons already introduced as definite. The point is not a “theological dodge,” but a syntactic and semantic reality observed across Greek literature.

    Brian J. Wright and Tim Ricchuiti’s study in The Journal of Theological Studies (2011) demonstrates that the Coptic indefinite article (ou-) in John 1:1c conveys a qualitative sense, meaning “divine” or “of God’s nature,” rather than an indefinite sense implying a separate deity. Coptic scholars like Bentley Layton and Ariel Shisha-Halevy confirm that the indefinite article in Coptic often functions differently from English, frequently denoting quality or essence, especially with abstract nouns like noute (god). For instance, Layton notes that the Coptic structure can predicate either a quality (“is divine”) or an entity (“is a god”), with the reader deciding based on context, but in John 1:1c, the qualitative sense is more likely. Your claim that Trinitarians “cannot” admit "a god" ignores that the qualitative interpretation aligns with the Coptic’s intent, as seen in other passages like John 1:18, where the Son is rendered with a definite article (pnoute), reinforcing his divine status.

    The Sahidic indefinite article does not function identically to the English indefinite article, and frequently carries a qualitative rather than an indefinite sense, especially with abstract nouns or nouns denoting nature or essence. The fact that ou.noute (“a god”) can mean “divine” or “of the same nature as God” in Coptic is not only acknowledged by most Coptologists, but also recognized by those same Dallas Seminary scholars. The context of John 1:1c is not introducing an additional member of a class of “gods,” but characterizing the Logos as sharing the very nature of the one God. That the Coptic version sometimes renders the Greek anarthrous θεός with the indefinite article in other passages (including, at times, for the Father himself) only further demonstrates that it does not, in Coptic idiom, always indicate an indefinite or lesser being. The evidence you cite is therefore not only inconclusive, but actually undercuts the Watchtower’s position when fully examined. The Coptic translation thus undermines, rather than supports, the NWT’s henotheistic implication.

    Scholars like Harner, Dixon, and Wallace do not deny that an indefinite sense is grammatically possible in some contexts, but they argue that in John 1:1c, the qualitative sense is contextually demanded. Harner’s analysis shows that the structure θεός ν λόγος emphasizes the Word’s shared divine nature with God, avoiding both modalism (equating the Word with the Father) and polytheism (implying multiple gods). Your reference to J. Harold Greenlee’s axiom—that “nouns without the article are either indefinite or qualitative”—is itself incomplete, as it fails to note the relative probabilities and contextual constraints that grammar and discourse impose. The crucial question is not whether a qualitative or indefinite sense is possible in isolation, but which is required or best supported by the context.

    In John 1:1c, where the Logos has just been distinguished from “the God” (i.e., the Father) and is then described as θεός in a creedal, ontological statement about his nature, the qualitative reading is not only grammatically probable, but contextually demanded. This is why virtually every major scholarly translation, including those by non-Trinitarians, have abandoned the indefinite “a god” rendering, in favor of “the Word was God,” “the Word was divine,” or “what God was, the Word was”—all of which preserve the qualitative sense without polytheistic or subordinationist implications. In John 1:1c, the context of the prologue, which establishes the Word’s pre-existence and role as creator (John 1:3), precludes an indefinite rendering, as it would introduce a secondary deity, contradicting biblical monotheism (e.g., Deuteronomy 32:39, Isaiah 44:6). The NWT’s list, while showing instances where indefinite articles are used, does not address the unique theological context of John 1:1c, where θεός carries a monotheistic weight absent in terms like “prophet” or “devil.”

    Wallace and Dixon acknowledge that English translations may use an indefinite article to convey a qualitative sense due to idiomatic constraints, but this does not mean the Greek noun is indefinite. For instance, in John 4:19, “prophetes” is qualitative, describing Jesus’ prophetic nature, but the indefinite article is used in English for clarity, not to denote a generic class member. Similarly, in John 6:70, διάβολος (devil) is qualitative, emphasizing Judas’ devilish character, not his membership in a class of devils. In John 8:44, “man-killer” (νθρωποκτόνος) is qualitative, describing the devil’s nature, not an indefinite instance. Your claim that Hanna’s acceptance of an indefinite article in John 8:44 but rejection in John 1:1c shows theological bias ignores the contextual differences. John 1:1c’s theological weight—introducing the Word’s divine role—demands a rendering that reflects its monotheistic context, unlike the descriptive roles in the cited verses. Wallace means the Greek is qualitative, with “a” added in English for clarity, not that indefiniteness is equally valid. Dixon’s point reinforces this. In John 6:70 and 8:44, διάβολος and νθρωποκτόνος are qualitative, with English “a” reflecting idiom, not grammar. John 1:1c’s context—a creedal statement—demands a qualitative θεός, not “a god,” which clashes with John’s monotheism (John 10:30). Greenlee’s axiom that anarthrous nouns are indefinite or qualitative requires contextual adjudication; here, qualitative prevails.

    The tally of non-traditional English versions you cite is, in fact, misleading and irrelevant to the question of grammatical correctness or theological coherence. Many such renderings arise from obscure, paraphrastic, or idiosyncratic translations, often outside the mainstream of linguistic or theological scholarship. The standard academic authorities in Koine Greek and biblical translation (including BDAG, Liddell-Scott, Wallace, Harner, Porter, and others) support the qualitative interpretation. The French translations (e.g., Segond’s “tu es prophète” vs. Darby’s “tu es un prophète”) illustrate that language-specific idioms vary, but the Greek remains qualitative, not indefinite.

    While it’s true that some translations deviate from “the Word was God,” many, like “the Word was divine,” aim to capture the qualitative sense of θεός, not to support the NWT’s implication of a lesser deity. The scholarly consensus, as seen in standard translations (e.g., NIV, ESV, NASB) and works by Metzger, Bruce, and Mantey, overwhelmingly favors the qualitative interpretation, aligning with “the Word was God.” Your estimate of 500 English Bible versions with one in ten deviating is speculative and lacks citation, but even if accurate, it represents a minority view often driven by theological agendas, as with the NWT. The majority of mainstream translations and scholars reject “a god” due to its polytheistic implications, which conflict with John’s monotheistic theology.

    Finally, your citation of John 20:31 (“that you may believe that Jesus is the Christ, the Son of God”) does not advance your case. Nowhere does this statement imply that Jesus is not fully divine; rather, in Johannine theology, the unique Sonship of Christ is itself a claim to full participation in the Father’s nature (see John 5:18; 10:30–33; 14:9–11; 17:5, 21–24). The Prologue’s explicit affirmation of the Logos as the Creator of all things, the possessor of life, and the one in the closest relationship with the Father (John 1:3–4, 18) all make sense only if the Word is not a lesser “god,” but is fully, qualitatively divine. In John’s Gospel, "Son of God" is not a subordinate title but a claim to divine equality, as seen in John 5:18, where Jesus’ claim to be God’s Son prompts accusations of making himself equal with God, and John 10:33, where his divine claims lead to charges of blasphemy. John 20:28, where Thomas calls Jesus “my Lord and my God,” serves as the Gospel’s climactic affirmation of Jesus’ deity, directly linking to John 1:1c. The purpose of John’s Gospel is to affirm Jesus as the divine Messiah, not a lesser god, making the NWT’s rendering inconsistent with the text’s intent.

    Your claim that Trinitarian translations are driven by theology while the NWT aligns with linguistic data is a false dichotomy. All translations involve theological considerations, but the NWT’s “a god” is explicitly shaped by Arian theology, which denies Jesus’ full divinity. The traditional translation, supported by scholars like Wallace, Harner, and Dixon, balances grammar and theology, aligning with the monotheistic framework of the New Testament. Your suggestion of monolatristic henotheism—where one God is worshipped among others—is untenable, as John’s Gospel, rooted in Jewish monotheism (Deuteronomy 6:4), rejects any notion of multiple gods. The qualitative interpretation of θεός in John 1:1c affirms the Word’s full divinity while maintaining personal distinction from the Father, as articulated in Trinitarian theology and early Christian creeds like the Nicene Creed.

    The use of nomina sacra in ancient manuscripts further undermines the NWT’s rendering. In texts like Papyrus 46, θεός and κύριος (Lord) are abbreviated as nomina sacra (e.g., ΘΣ, ΚΣ) when referring to both the Father and the Son, indicating their equal divine status. In John 1:1c, θεός is consistently treated as a nomen sacrum, suggesting that early Christians viewed the Word as fully divine, not a lesser god. This practice, evident in the earliest manuscripts, contradicts your claim that the indefinite rendering was normative.

    Your defense of the NWT’s “and the Word was a god” is grammatically and contextually unsustainable. The qualitative interpretation of θεός in John 1:1c, supported by Harner, Dixon, Wallace, and others, aligns with the Greek syntax, Johannine theology, and early Christian tradition. The NWT’s rendering introduces polytheism, contradicts monotheistic principles, and lacks consistent application across similar constructions. The Coptic translation, when properly understood, supports the qualitative sense, not the indefinite.

    In sum, the case for rendering John 1:1c as “a god” is both grammatically and contextually weak, and has been repeatedly refuted by the best scholarship in the field. The consensus, based on detailed analysis of Greek grammar, context, and theological coherence, is that the anarthrous θεός in John 1:1c is qualitative: it asserts the full, essential divinity of the Logos, not his membership in a class of lesser deities. To claim otherwise is not only to misunderstand the syntax, but to sever the Prologue from its deeply Jewish monotheistic context and its place in the development of Christian doctrine. The scholarly consensus, historical evidence, and textual data overwhelmingly favor the traditional translation, as the most accurate reflection of John’s intent.

    Thus, the traditional rendering—“the Word was God”—remains the most accurate, both grammatically and theologically.

  • Wonderment
    Wonderment
    aqwsed: "The tally of non-traditional English versions you cite is, in fact, misleading and irrelevant to the question of grammatical correctness or theological coherence. Many such renderings arise from obscure, paraphrastic, or idiosyncratic translations, often outside the mainstream of linguistic or theological scholarship."

    Your answer above indicates a prejudicial conclusion. I have examined many of the translations of John 1.1 with non-traditional renderings - and their scholarly background, and what it reveals is in stark contrast with your claim above. Many of those versions come from Catholic & Protestant sources from reputable universities at that, from the world over. So your statement is not factual. The majority of versions siding with the trinitarian view of John 1.1 are not necessarily more accurate than those coming from other sources with a different exegesis. Bias goes both ways.

    You keep making the very same basic mistake,i.e., that a qualitative noun is altogether different in scope and reach, exclusive from the indefinite notion. Not true. You cite Colwell, Wallace & Dixon. They as Trinitarians came up with these studies to counter the distribution of Bible readings going against the current renderings of John 1.1. They were just as biased from the start as any other. They do concede that the subject is no clear matter. The qualitative and indefinite concept can often overlap, period. Wallace admitted so. He himself produced diagrams in the book where the concept of both were connected. Why deny this?

    More importantly, most translators do not follow his rules to the dot. Both Wallace & Dixon say that the indefinite notion is the least popular of the three (incl., definite). Nonsense! It is a trinitarian play.

    This is what they want the Trinitarian masses to believe, but in practice translators often render pre-verbal anarthrous predicates in the nominative with the indefinite article. Why don't you take the time to verify or deny these occurrences by using the Zondervan Greek-English Concordance?

    Take Acts 28.4 literally for instance: "By all means murderer is the man this"

    How do most translators deal with this construction? Do they translate this indefinite-qualitative predicate like Trinitarians do with John 1.1 and end up with: By all means, this man is Murderer? No. The NIV reads: “This man must be a murderer.” Is it wrong to just say that this man is murderer? No just like it is acceptable to say that "the Word was god," in a qualitative sense.

    And this brings me to the last point. You claim that the Word was God is the most accurate rendering. Really? If you believe this, then you are not making any distinction whatsoever between the articular theos and the anarthrous theos in the same verse. Verse 2 argues for a distinction between the invisible God the Word was with, and the Word who was in his presence, now arriving in flesh before mankind. It is this only-begotten Logos who can explain the invisible Father like no other, by virtue of being in close association with God.

    In English, the Word was God leads the reader wrongly to identify the Word with God, and does not bring out the qualitative force as god, or divine does in this descriptive predicate. Thus various translators render, "the Word was god (or, divine, a god)." Barclay explains this matter simply.


  • Blotty
    Blotty

    AQ, Wonderment and Earnest are far more credible than you are.. Wallace is a grammarian - you Ai text generate and cant cite a source to most of your claims...

    Is there an option to ban AQ from this website or atleast character limit him to a single sentence responce? if they cant be constructive or helpful why do we allow such nonsense?

  • aqwsed12345
    aqwsed12345
    @Wonderment

    Your response insists that the existence of a broad range of non-traditional English translations of John 1:1c—many from Catholic and Protestant sources—discredits the argument for scholarly consensus behind the mainstream rendering. However, this claim misconstrues the very nature of scholarly consensus in biblical studies. The mere presence of alternative translations does not invalidate the overwhelming alignment of major critical scholars and standard reference works with the qualitative interpretation, nor does it demonstrate equivalence between mainstream academic translation and the often idiosyncratic choices of marginal or paraphrastic versions. To measure consensus by mere quantity or denominational label is methodologically unsound: scholarship is not advanced by counting versions, but by analyzing the linguistic, literary, and theological arguments undergirding translation choices. The mainstream rendering persists because it most faithfully reflects the grammatical, contextual, and theological data.

    he core issue of this debate revolves around the interpretation of the Greek text of John 1:1c and what it implies about the relationship between the Father and the Son. Specifically, the question is whether this verse:

    • Asserts an identity of quiddity (essential nature) between the Father and the Son,
    • Neither asserts nor denies such an identity, or
    • Explicitly negates it.

    To answer this, we must first examine whether the theological framework of the Gospel of John—and the NT as a whole—allows for the concept of a “lesser god” or reflects a form of monolatristic henotheism (the worship of one god while acknowledging the existence of others). The NT, including John’s Gospel, emerges from the context of Second Temple Judaism, a period marked by a strong commitment to monotheism. This is rooted in the Hebrew Scriptures, particularly in texts like Isaiah 43:10-11 and 44:6, where God declares, “Before me no god was formed, nor shall there be any after me,” and “I am the first and I am the last; besides me there is no god.” These anti-henotheistic statements reject the idea of multiple gods or lesser divine beings sharing God’s nature or status.

    In this light, John 1:1c must be understood. The absence of the definite article before θεός has led some, like Arian apologists, to argue that the Logos (the Son) is a “lesser god” rather than fully divine in the same sense as the Father. However, the broader theological framework of the NT does not support this. There is no evidence of monolatristic henotheism—where multiple gods are acknowledged but only one is worshiped—in this framework.

    Arian apologists often cite Exodus 7:1 (“See, I have made you like God [elohim] to Pharaoh”) and Psalm 82:6 (“I said, ‘You are gods’”) to argue that the Bible permits the concept of lesser divine beings. However, these OT texts, written centuries earlier in Hebrew, do not align with the theological context of the Second Temple period, when the apostles lived and wrote. In Exodus 7:1 Moses is appointed as God’s representative, not a literal lesser deity. The term elohim is used functionally, not ontologically. Psalm 82 likely refers to human judges or heavenly beings (e.g., angels) under God’s authority, not independent gods. The Septuagint, the Greek translation of the Hebrew Bible widely used in the apostolic era, reflects this shift. For instance, in Psalm 8:5 (8:6 in the LXX), elohim (“gods” or “heavenly beings”) is translated as aggeloi (“angels”), a choice echoed in Hebrews 2:7. This shows a deliberate move away from language that could suggest henotheism or multiple gods.

    By the time of the NT, such terminology had been “purged” or reinterpreted to align with strict monotheism. The apostles, operating in this milieu, show no precedent for reviving a henotheistic understanding of divine beings.

    The Greek text of John 1:1c asserts an identity of quiddity between the Father and the Son, affirming that the Word shares the same divine essence as God. The theological framework of John and the NT rejects the notion of a “lesser god” or monolatristic henotheism. While Arian apologists point to earlier texts like Exodus 7:1 and Psalm 82, these do not reflect the Second Temple context, where such ideas were reinterpreted or abandoned, as seen in the LXX and apostolic writings. There is no evidence from the apostolic era to support a henotheistic reading. Instead, John 1:1c stands as a clear declaration of the Son’s full divinity.

    The core issue lies in the anarthrous use of θεός in John 1:1c, which lacks the definite article ho, unlike τν θεόν in John 1:1b. You claim that my reliance on scholars like Daniel Wallace is misplaced, arguing that grammarians and translators diverge, with Wallace's work allegedly biased by theology, as admitted in his preface. However, this mischaracterizes Wallace's Greek Grammar Beyond the Basics, which is an empirical study of Koine Greek syntax, not a theological treatise. Wallace concludes that pre-verbal anarthrous predicate nominatives, such as θεός here, are primarily qualitative, indicating the nature or essence of the subject, based on extensive analysis of NT and Septuagint data. His preface acknowledges scholarly disagreement, but this does not undermine his finding that such constructions are overwhelmingly qualitative, especially in theological contexts like John 1:1c.

    Your argument conflates the qualitative and indefinite readings of anarthrous predicate nouns, asserting that there is significant overlap and that this undercuts the distinction insisted on by Wallace, Harner, and Dixon. Yet this is a fundamental misunderstanding of their analysis. The central question is not whether some overlap exists in certain constructions (which all grammarians recognize), but which sense is required in the specific context of John 1:1c. The studies you reference are not trinitarian polemics, but careful examinations of Greek syntax across the NT and contemporaneous literature. Both Harner and Dixon, after exhaustive statistical studies, demonstrate that pre-copulative, anarthrous predicate nominatives are overwhelmingly qualitative in Johannine and broader Koine usage, especially in creedal or ontological statements. Wallace’s categories explicitly note that while indefiniteness is sometimes possible, in John 1:1c the qualitative sense is contextually and grammatically demanded by the literary context and by the structure of the Prologue. The diagram you allude to in Wallace (p. 269) is not an admission that the indefinite and qualitative readings are interchangeable, but a mapping of their possible intersections in Greek grammar, with context always remaining decisive.

    Your assertion that Trinitarians evade the indefinite sense for theological reasons, appealing to Colwell, Dixon, and Harner, ignores the linguistic rigor of these studies. Colwell's Rule states that definite predicate nominatives preceding the verb often lack the article, suggesting θεός can be definite or qualitative, not necessarily indefinite. Harner's study analyzed over 250 examples and found that the qualitative sense predominates, particularly in pre-verbal positions, with the indefinite sense rare and contextually excluded. Dixon's dissertation similarly supports this, noting that English translations may use an indefinite article for readability, but the Greek conveys essence. These conclusions are not theological fiat but empirical observations, accepted across confessional lines, including by non-Trinitarian linguists.

    Your claim that Wallace, Dixon, and Colwell are biased from the start, conducting studies to counter non-traditional readings, is unsubstantiated. Their work is peer-reviewed and grounded in syntactic patterns, not theological agendas. Wallace's diagrams show overlap between qualitative and indefinite senses in some contexts, but he explicitly states that in John 1:1c, the qualitative force is primary, given the context of the prologue distinguishing the Logos from “the God” (John 1:1b) while affirming its divine role (John 1:3-4). This overlap does not imply equivalence; rather, it acknowledges that English translation may require “a” for idiom, but the Greek remains qualitative.

    You cite the NWT's 1984 Appendix, listing anarthrous predicate nouns (e.g., Mark 6:49, “a spirit”; John 4:19, “a prophet”) where translators use an indefinite article, suggesting this supports “a god” in John 1:1c. However, this conflates translation choices with grammatical function and ignores contextual distinctions. In Mark 6:49, φάντασμα conveys a qualitative nature (“having the nature of a spirit”), not an indefinite class; John 4:19's προφήτης emphasizes Jesus's prophetic essence, with “a” added in English for readability, not because the Greek is indefinite. These examples differ from John 1:1c, where θεός is not a category noun but a term for the unique divine nature, introduced in a creedal prologue asserting the Logos's role as Creator (John 1:3) and intimate relation with “the God” (John 1:1b). The NWT's list illustrates translation variability, but it does not establish that an indefinite sense is grammatically or contextually appropriate for John 1:1c, where the qualitative interpretation—“the Word was fully God in nature”—aligns with the prologue's monotheistic framework.

    You challenge me to verify these occurrences using the Zondervan Greek-English Concordance, citing Acts 28:4 (“By all means murderer is the man this”) as an example where translators render it as “This man must be a murderer” (NIV), suggesting an indefinite-qualitative sense. However, this comparison is flawed. In Acts 28:4, the noun “murderer” (νθρωποκτόνος) describes a category, and the context (the islanders' inference based on a snakebite) supports an indefinite reading (“a murderer among others”). In contrast, John 1:1c is a metaphysical statement about the Logos's essence, not a descriptive category, set within a prologue exalting its divine status. The NIV's “a murderer” reflects English idiom, not a rejection of qualitative force, and does not parallel John 1:1c, where the context demands a qualitative θεός.

    Hence, the examples you cite are not analogous to the use of θεός in John 1:1c. The use of the indefinite article there is natural and expected in English, given the idiom and the type of noun involved. By contrast, θεός in John 1:1c is a monadic, abstract noun situated in an ontological, not classificatory, assertion about the Word’s nature, immediately after the Word has been distinguished from the Father by the articular usage in 1:1b. Thus, the analogy fails: translating θεός ν λόγος as “the Word was a god” introduces an unwarranted category distinction, suggesting the existence of multiple gods, a meaning John’s monotheistic context does not allow.

    Your claim that translators often render pre-verbal anarthrous predicates with an indefinite article, contradicting Wallace and Dixon's rules, is overstated. Both scholars note that the indefinite sense is the least popular, but you dismiss this as a “Trinitarian play.” This ignores that translation practices reflect target-language needs, not a repudiation of Greek grammar. For instance, in John 4:19, Wallace describes the sense as “indefinite-qualitative,” meaning the Greek is qualitative, with “a” added in English for clarity, not because it is indefinite. Dixon admits English may require “a” for qualitative nouns, but this does not mean the Greek is indefinite; it reflects translation idiom, as seen in French translations like Segond (“Je vois que tu es prophète”) versus Darby (“Je vois que tu es un prophète”).

    You argue that “the Word was God” wrongly identifies the Word with God, failing to distinguish between the articular θεός (John 1:1b, “the God”) and the anarthrous θεός (John 1:1c), citing verse 2 to argue for a distinction between the invisible God and the Word. However, this misunderstands the Greek. The anarthrous θεός in 1:1c is qualitative, emphasizing shared essence, not identity of person. John 1:2 (“He was in the beginning with God”) reinforces this, maintaining the distinction in personhood (“with [the] God”) while affirming shared full divinity (“was God”).

    Hence, the absence of the article with θεός in 1:1c is precisely what signals qualitative predication: the Word is everything that God is by nature, but remains distinct in person from “the God” ( θεός) whom he is with. This is the point of the syntactic construction. The traditional translation “the Word was God” is, in English idiom, the most effective way of conveying this qualitative identity—though not identity of person—without introducing the misleading implication of a lesser, subordinate, or additional deity. Alternative translations like “the Word was divine” or “what God was, the Word was” have been adopted by some, not to suggest a lesser being, but to clarify the qualitative force. None of these renderings support the indefinite, category-member sense that the NWT projects.

    Barclay, whom you cite, himself states: “John does not say that the Word was the God (ho theos); to have said that would have been to identify the Word with God. He says that the Word was theos—without the article—which means that the Word was, as we would say, of the very same character and quality and essence and being as God.” This is precisely the qualitative force articulated by Wallace, Harner, and the vast majority of critical scholars: not that the Word is “a secondary, lesser god,” but that the Word shares fully in the nature of God. The very distinction between the articular and anarthrous θεός in John 1:1 is not obliterated by the traditional translation; rather, it is the result of careful exegetical and grammatical analysis, as found in every major critical commentary.

    The assertion that English “the Word was God” is ambiguous or misleading is not supported by mainstream English usage, in which “God” functions both as a proper noun and as a title denoting divine nature. The context, especially when interpreted alongside the rest of the Prologue (John 1:3, 1:18, etc.), makes it clear that the Word is not identical with the Father in person but is fully and truly God in essence. The translation “the Word was God” reflects this distinction as well as the language allows, and any further clarification belongs in commentary, not in a distortion of the translation itself.

    You also claim that the NWT’s rendering is more faithful because it reflects the anarthrous nature of the Greek noun. But as demonstrated by every major Greek grammar, including those by non-Trinitarians and even secular scholars, anarthrous predicate nouns in this syntactic position are typically qualitative, not indefinite. The mere absence of the article does not imply indefiniteness; it is the context, especially in ontological and theological statements, that is determinative.

    Your invocation of a global plurality of English translations, some of which differ from the traditional rendering, does nothing to advance your argument. The overwhelming weight of scholarly authority—BDAG, Wallace, Harner, Dixon, Porter, Robertson, and every standard lexicon and grammar—aligns with the qualitative interpretation. The proliferation of alternative translations often arises from confessional bias or from a desire to clarify the qualitative sense for lay readers, not from grammatical necessity. Indeed, translations that opt for “divine” or “what God was, the Word was” do so out of concern for clarity, not because they accept the indefinite “a god.” Your examples, like the 1808 Improved Version (“and the Word was a god”) or Moffatt's 1924 translation (“the Logos was divine”), reflect minority views, often driven by theological agendas, not grammatical necessity. William Barclay clarifies that Jesus is of the same essence as God but distinct, supporting the qualitative interpretation.

    Finally, the very logic of your position leads to a form of henotheism or monolatry, in which more than one being can properly be called “god.” This is utterly foreign to the context of John’s Second-Temple Jewish monotheism, in which the Shema (Deut 6:4) remains the fundamental confession. The Prologue of John, from 1:1 to 1:18, affirms that the Word is not a creature, not a member of a class of gods, but the very source of all that exists (John 1:3), the light and life of men, the one who reveals the Father. Such predication cannot be made of a subordinate, lesser deity. It is the qualitative force that alone does justice to the context, the syntax, and the theology of the Gospel.

    In summary, your arguments—that my appeal to Wallace is misplaced, that qualitative and indefinite senses overlap, and that non-traditional translations undermine the traditional rendering—are refuted by the grammatical evidence, scholarly consensus, and theological context. Wallace, Harner, and Dixon support a qualitative θεός in John 1:1c, aligned with the monotheistic framework of John. The NWT's “a god” relies on a misreading of syntax and translation practice, while the Coptic and variant translations cited fail to overturn the consensus. Theology informs interpretation, but here it aligns with, rather than distorts, the linguistic data, affirming “the Word was God” as the most accurate rendering.

    Your argument fails on grammatical, contextual, and theological grounds. The consensus of critical scholarship, based on exhaustive studies of Greek grammar and NT usage, is that John 1:1c is best rendered with a qualitative sense—“the Word was God,” or, if further clarity is desired, “the Word was fully divine.” The indefinite “a god” is not supported by the syntactic construction, the context of the Prologue, the usage in the rest of the Gospel, or the monotheistic setting of the NT. The qualitative force affirms the unique and eternal deity of the Word, maintaining distinction of persons without division of essence. Any rendering that suggests otherwise is linguistically, historically, and theologically mistaken.

  • stan livedeath
    stan livedeath

    does anyone read all this shit ?

Share this

Google+
Pinterest
Reddit