2 Peter 1: 1 Corruption in the NWT

by Sea Breeze 54 Replies latest watchtower bible

  • aqwsed12345
    aqwsed12345

    What I think it's rather pathetic that you're getting stucked up on a short summarized part of a longer comment, instead of addressing the whole in its entirety. As for Acts 20:28, you can't chase two rabbits and catch them both. Decide whether you want to overturn the majority scholarly opinion on the "Church of Lord" vs. "Church of God" part, or whether you're going to go down the "own blood" vs. "blood of his own [one]" route. Anyway, I explained all about textual variants in the dedicated topic.

  • slimboyfat
    slimboyfat

    You are ridiculous. Your AI posts are full of mistakes and faulty logic. And your response to me pointing out a particularly glaring error is that I am at fault because I should have gone through your entire screed and pointed out all the mistakes. Wow. Where do you get off? Why should I spend hours of my time trying to clear up the AI nonsense it takes you seconds to generate and post? And when errors are pointed out you don’t even acknowledge them, just move on, misdirect and change the conversation.

    Anyway, I explained all about textual variants in the dedicated topic.

    Sure you did, if by “explain” you mean posted a lot of untrue and misleading statements generated by AI then failed to apologise when I pointed out your wildly inaccurate summary on this thread. Thanks for that.

  • aqwsed12345
    aqwsed12345
    @Earnest

    Your latest rejoinder turns on the claim thatὁ θεός in Paul and Peter functions “in the same way” asὁ Βάαλ in Judges, so that Titus2:13 and 2Peter1:1 ought to be barred from the domain to which GranvilleSharp’s canon applies.The comparison depends on two assertions: that Βάαλ in the LXX displays the same grammatical behaviour as θεός in the NT, and that θεός in apostolic prose is therefore effectively a proper name, not a common noun.Neither assertion withstands close inspection of the Greek evidence.

    So the question is whyὁ Βάαλ is a proper name andὁ θεός is not? The assertion that “the Baal” and “the God” share a grammatical similarity sufficient to equate their syntactic behavior in Koine Greek overlooks critical distinctions in their lexical and syntactic properties. In the LXX, “Baal” (Βάαλ) predominantly functions as a proper noun, referring to a specific Canaanite deity or localized manifestations thereof, as seen in passages such as Judges 2:13 Βάαλ κα τας στάρταις). In the LXX the transliterationΒάαλ always designates the West‑Semitic storm‑deity (or, by metonymy, his cult‑image).When Greek writers need to modify that name, they do not treatΒάαλ as an adjective‑bearing common noun; they coin a new compound proper name (Βεελζεβούβ, Βεελφεγώρ, Βεελμαών) or they use a genitive phrase (e.g., τ Βάαλ τ Φαγώρ, Num25:5LXX).The bare sequence ὁ Βάαλκαὶ σωτήρ is therefore unattested; whenever ὁ Βάαλ is followed by another title the article is repeated or the construction is reshaped, precisely because proper names resist being “further described” by a second common noun inside the same determiner phrase.That is why Sharp excluded proper names: Greek article‑syntax itself marks them off from the class of nouns to which his observation applies.

    Thus, your examples do not establish “Baal” as a common noun in Greek usage. In Semitic languages, “baal” indeed carries the general sense of “lord” or “master,” but in the LXX, it is consistently treated as a proper noun designating a specific deity or cultic entity, even when accompanied by epithets or in plural form to denote multiple localized deities (e.g., “the Baals” in Judges 3:7). These modifiers and plural forms reflect the diversity of Canaanite cultic practices rather than a shift to a common noun status. In contrast, θεός is a common noun in Koine Greek, meaning “god” or “deity,” and its reference is determined by context and grammatical markers, such as the definite article. This distinction is crucial for the application of Granville Sharp’s rule, which governs constructions involving singular, personal, common nouns linked by a single article, as in Titus 2:13 and 2 Peter 1:1. The rule does not apply to proper nouns like “Baal” due to their inherent definiteness, which precludes the need for syntactic unification via a single article.

    By contrast, θεός behaves throughout Greek literature as an ordinary count noun.It is pluralized (θεοί, Acts14:11), qualified by adjectives (ὁ μόνος θεός, John5:44), prefixed by genitives (ὁ θες τς ερήνης, Rom15:33, 1Thess 5:13), denied by the privative (οκ ἔστιν θεός, 1Cor8:4), and applied to different referents in a single paragraph (1Cor8:5‑6).A noun that can be plural, modified, denied and re‑applied is by definition a common noun, even when conventionally associated with one person.Sharp’s criterial phrase “common, not proper, noun” retains full force: θεός shares exactly the same syntactic profile as κύριος, σωτήρ, πατήρ, γεμών—titles that furnish the bulk of his database.

    Your comparison of “the Baal and Saviour Jesus Christ” to “the God and Saviour Jesus Christ” misapplies this distinction. In a hypothetical construction like Βάαλ κα σωτρ ησος Χριστός, “Baal” as a proper noun would not fall under Sharp’s rule, as it inherently designates a specific entity without requiring further syntactic clarification. Conversely, θεός and σωτὴρ are common nouns, and their coordination under a single article in Titus 2:13 and 2 Peter 1:1 signals co-referentiality, identifying Jesus Christ as both “God” and “Savior.” Your claim that ὁ θεός functions as an identifier akin to a proper name in Peter and Paul’s writings conflates semantic frequency with syntactic function. While ὁ θεός often refers to the Father, this reflects a theological convention, not a grammatical constraint, and the TSKS construction’s syntactic unity takes precedence in determining reference. The absence of any Koine Greek example where a single-article TSKS construction with singular, personal, common nouns denotes distinct individuals reinforces the reliability of this interpretation.

    Because θεός and σωτήρ are common nouns, the strings τοῦ μεγάλουθεοῦ καὶ σωτρος ἡμν and τοῦ θεοῦ ἡμνκαὶ σωτρος fall squarely inside the evidential field Sharp surveyed and Wallace re‑tested.There is still no example—biblical, epigraphic, papyrological—where two singular, personal, common nouns under a single article undeniably designate different individuals.That uniform distribution is not a “patch” but the dataset; Wallace’s restatement simply clarifies its limits. To remove θεός from the pool one would need to show that it never functions as a common noun in apostolic Greek, a factual impossibility given the dozens of places where Peter and Paul attach adjectives or genitives to ὁ θεός.

    The examples you cite (Baal‑berith, Baal‑zebub, “Baals” in the plural) confirm rather than undermine the distinction.In Hebrew the lexeme בַּעַל can mean “lord/owner,” but once transliterated as Βάαλ in Greek it is frozen as a proper name or embedded in a longer proper compound.None of those compounds occurs in a Greek clause of the form ὁ Βάαλκαὶ X where X is another common noun without a repeated article.The Septuagint’s Greek therefore furnishes no counter‑instance to Sharp’s canon.

    Hence, Βάαλ is consistently a transliterated proper name. It is indeclinable, never takes an article for the purpose of delimiting one member of a wider class, and never appears with qualitative or abstract predicates that convert it into a concept (“a baal,” “some baal,” “true baal,” etc.). When it does carry what look like adjectival elements—Βααλβερίθ, Βααλμυίαν, Βααλφαγώρ—those elements are part of a fixed compound that functions exactly like Zeus κέσιος or Ares ρειοπαγίτης; the compounded epithet is not a syntactic adjective but an embedded genitive or appositive that leaves the name’s definiteness untouched. The handful of plural forms (ο Βάαλιμ, τ Βάαλιμ) mark separate local cult‑centres, not a grammatical shift from proper‑name to common‑noun status. On every test Sharp himself articulated—resistance to adjectival gradation, immunity to qualitative anarthrous use, incapacity to range over a genus—Βάαλ behaves like Κίρρος or πόλλων, not like θεός, κύριος, or σωτήρ.

    θεός, by contrast, is the paradigm common noun. The NT writers can deploy it articularly ( θεός) or anarthrously (θες ν λόγος), pluralise it (θεοί), predicate it (θεός ληθινός), deny it (ο θεός), and even juxtapose it with itself in a single sentence ( τν θεν θεός, 2Macc12:15). This malleability is what makes θεός susceptible to the article’s unifying force in a TSKS string. Where one article governs θεός and a second common noun singular and personal, first‑century Greek readers instinctively treat the two nouns as a double description of one individual—unless some positive contextual factor requires otherwise. No such factor is present in Titus or 2Peter; indeed both letters reinforce the co‑reference by repeating the construction with an intervening ξ and two articles when distinct persons are intended (Titus1:4, 3:4–6; 2Pet1:2).

    If θεός were a proper name, every NT author would need to repeat the article whenever coupling it with another title to avoid implying identity.Peter himself disproves the premise: in 1Peter5:10 he writes ὁ θεςπς χάριτος (single article, two nouns) where no one imagines a second person.The construal test is article‑distribution, not an a‑priori decision about which nouns are “names.”In Titus2:13 and 2Peter1:1 the syntax points in one direction; semantics must follow.

    The standard Greek text of Hippolytus’s De Christo et Antichristo §67 is preserved in Migne’s Patrologia Graeca (PG), Volume 10, column 788. The text includes the Titus 2:13 citation as: τν μακαρίαν λπίδα κα πιφάνειαν τς δόξης το μεγάλου Θεο κα Σωτρος μν ησο Χριστο, which is followed by the gloss το κυρίου μν ησο Χριστο το Βασιλέως, affirming the unitary reading where Jesus Christ is identified as both “our great God and Savior.” The Ante-Nicene Fathers translation corroborates this, rendering it as “looking for that blessed hope, and the glorious appearing of the great God and our Saviour Jesus Christ.”

    Clement of Alexandria’s citation of Titus 2:13 in Protrepticus 1 further supports the unitary reading. The text, available through the Perseus Digital Library, and in the Greek/English edition by George William Butterworth, read: προσδεχόμενοι τὴν μακαρίαν ἐλπίδα καὶ ἐπιφάνειαν τῆς δόξης τοῦ μεγάλou θεοῦ καὶ σωτῆρος ἡμῶν Ἰησοῦ Χριστοῦ. This citation aligns with Hippolytus’s interpretation, reinforcing the early Christian understanding of Titus 2:13 as referring to Jesus Christ as both “God” and “Savior,” with Clement explicitly confessing Jesus as God in this passage by attributing the divine title “the great God” directly to Him.

    The Baal analogy fails because Βάαλ, once imported into Greek, functions grammatically as a proper name, immune to the article’s unifying force; θεός remains a common noun and obeys the ordinary rules of Koine determination. Hence Wallace’s “not proper names” stipulation does not shield Sharp’s canon; it simply restates a linguistic boundary already observed by Sharp and confirmed by subsequent corpora. The passages under debate lie on Sharp’s side of that boundary. Hippolytus’ pre‑Nicene citation, preserved in Greek, shows that early exegesis read them accordingly. Until an authenticated Koine example overturns the distribution, the titles remain fused: Jesus Christ is our great God and Savior.”

    The parallel Βάαλ κα σωτρ ησος Χριστός fails before it can disprove Sharp’s canon: its first noun is excluded a priori as a proper name, just as Παλος or βραάμ would be. θεός in Titus2:13 and 2Peter1:1, on the other hand, is a common noun sitting squarely within the canonical data‑set. The single article before θεός therefore binds it to σωτήρ in a unified description of Jesus Christ, a reading endorsed by the oldest patristic exegesis, by every modern critical edition, and by the absence of any genuine Greek counter‑example. The linguistic evidence resists the attempted equivalence of Baal and God, and Hippolytus’s witness remains both textually and theologically probative.

  • Earnest
    Earnest
    aqwsed12345 : The handful [my italics] of plural forms (ο Βάαλιμ, τ Βάαλιμ) mark separate local cult‑centres, not a grammatical shift from proper‑name to common‑noun status.

    Judges 2:11; 3:7; 8:33; 10:6,10; 1 Samuel 7:4; 12:10; 1 Kings 18:18; 2 Chronicles 17:3; 24:7; 28:2; 33:3; 34:4; Jeremiah 2:23 all have plural forms of Baal.

    aqwsed12345 : The standard Greek text of Hippolytus’s De Christo et Antichristo §67 is preserved in Migne’s Patrologia Graeca (PG), Volume 10, column 788.

    I am not quite sure why you mention this. My query was about "the preserved Greek fragment in the Syriac catena [i.e. commentary]" which you claim mirrors the text of Hippolytus's "Treatise on Christ and Antichrist". If you are unable to help me identify it then please just say so.

    aqwsed12345 : The text [from the link above] includes the Titus 2:13 citation as: τν μακαρίαν λπίδα κα πιφάνειαν τς δόξης το μεγάλου Θεο κα Σωτρος μν ησο Χριστο,

    The text preserved in Migne's Patrologia Graeca which you link to reads : τν μακαρίαν λπίδα καπιφάνειαν το Θεο κα Σωτρος μν.

    In your underlined citation το μεγάλου Θεο κα Σωτρος μν ησο Χριστο the words I have emboldened (μεγάλου ["great"] and ησο Χριστο ["Jesus Christ"]) are simply not there. Just check your own link. So it could arguably be a citation from Titus 2:13 but it doesn't quite match.

    aqwsed12345 : which is followed by the gloss το κυρίου μν ησο Χριστο το Βασιλέως

    There is no gloss in the link you provide to Migne's Patrologia Graeca.

    aqwsed12345 : The Ante-Nicene Fathers translation corroborates this, rendering it as “looking for that blessed hope, and the glorious appearing of the great God and our Saviour Jesus Christ.”

    The Ante-Nicene Fathers translation you link to renders it as “looking for that blessed hope and appearing of our God and Saviour”. Again, you have inserted the words "great" and "Jesus Christ" as well as transposing the possesive pronoun "our".

    I do not wish to climb on the bandwagon but I do agree with slimboyfat that you should not rely on others to proofread your text if you use AI to produce it.

  • aqwsed12345
    aqwsed12345
    @Earnest

    I owe you two things at the outset: an apology for the imprecision of several earlier quotations, and a corrected, tighter account of the evidence that actually matters for the two questions you have pressed—(i) whether the pluralized forms “Baals” in the LXX convertΒάαλ into a true common noun comparable toθεός, and (ii) what precisely Hippolytus quotes in DeChristo et Antichristo67 and how that quotation bears on Titus2:13.

    The plural Βάαλιμ in the Septuagint does not recategorize Βάαλ. The texts you list—Judg2and3; 1Sam7 and12; 1Kgs18; 2Chr17,24,28,33,34; Jer2:23—do indeed reproduce the Hebrew plural baʿalîm. What matters, however, is the morphological behaviour of the loan‑word once it enters Greek. Βάαλ remains indeclinable: the plural is not Βάαλοι or Βάαλα (which would mark a naturalised Greek noun) but simply the Hebrew consonantal form prefaced by an article (ο / τ Βάαλιμ). Nothing else that signals a shift to common‑noun status ever appears. We never meet comparative or qualitative uses (“a true baal, a great baal”), never see adjective agreement, never find the anarthrous plural in distributive sense, and never encounter the noun declined for case. Grammatically, therefore, Βάαλ behaves just like the transliterated plural Σεραφείμ in Isa6:2LXX or Χερουβίμ in Exod25:18LXX—borrowed forms that can be counted but do not enter the ordinary article–noun–adjective system of Greek.

    While the number of examples is indeed significant, this does not undermine the argument that these plurals denote separate local cult centers rather than a grammatical shift to common noun status. In the Septuagint, “Baal” (Βάαλ) remains a proper noun, even in its plural manifestations, referring to specific deities or their cultic representations across different locales. This usage parallels other proper names in Greek that can be pluralized to indicate multiple entities bearing the same name, such as “the Caesars,” without altering their inherent definiteness. The plural forms thus reflect the diversity of Canaanite worship practices, not a reclassification as a common noun, and this distinction preserves the inapplicability of Granville Sharp’s rule to “Baal,” which applies only to singular, personal, common nouns.

    By contrast, θεός is the paradigm common noun of Koine prose. It is freely declined, pluralised, qualified ( μόνος θεός, ληθινς θεός), denied (οκ στιν θεός), and even re‑applied to different entities within a single discourse (1Cor8:5‑6). That malleability is precisely why GranvilleSharp’s observation—one article followed by two singular, personal, common nouns yields a single referent—can apply to θεός but not to Βάαλ. A proper name, whether singular or plural, is already definite; the article does not unite it with a following title the way it unites two common nouns.

    What Hippolytus actually quotes—and why the variation is immaterial. The paragraph printed at PG10, col.788 (derived from the Baroccianus catena tradition) indeed cites Titus2:13 in a shorter form:

    τν μακαρίαν λπίδα κα πιφάνειαν το Θεο κα Σωτρος μν.

    The epithet μεγάλου and the appositive ησο Χριστο are absent in this recension. I apologize for the confusion. The mention of PG 10 was intended to provide a related standard text, you correctly point out that my underlined citation, “τν μακαρίαν λπίδα κα πιφάνειαν τς δόξης το μεγάλου Θεο κα Σωτρος μν ησο Χριστο,” includes terms not present in the PG text. I regret this error and acknowledge that the addition of “μεγάλου” (great) and “ησο Χριστο” (Jesus Christ) was an inadvertent overextension, possibly influenced by variant readings or contextual assumptions. The Ante-Nicene Fathers translation (Volume 5, p. 218) similarly renders it as “looking for that blessed hope and appearing of our God and Saviour,” without “great” or “Jesus Christ,” further confirming the discrepancy. However, the contextual gloss in Hippolytus’s work, such as “For that was the expression he used” and the subsequent discussion of Christ’s appearing, indicates that “our God and Saviour” refers to Jesus, supporting the unitary reading despite the textual variance.

    The absence of the mentioned gloss in the cited PG 10, column 788, is another valid critique, and I apologize for asserting its presence. Upon re-examination, this gloss does not appear immediately following the citation in the referenced text, though it may reflect a broader interpretive tradition within Hippolytus’s work. The core citation itself, however, remains consistent with early Christian exegesis identifying Jesus as the referent, and the lack of the gloss does not negate this interpretation.

    Yet whichever catena you consult, the grammatical fact does not change: Hippolytus quotes a clause in which a single article (το) governs both θεός and σωτήρ. Thus—even in the shortest text—the two nouns are syntactically fused, and the surrounding comment removes all doubt that Hippolytus heard the fusion as a statement about Christ.

    The English rendering in Ante‑Nicene Fathers 5.218 (“looking for that blessed hope and appearing of our God and Saviour”) follows the same Greek line; its omission of μεγάλου and ησο Χριστο simply reflects the shorter catena text. What it does preserve is the co‑ordinated pair “God and Saviour” under one article, precisely the structure we are testing.

    Why these clarifications leave the central conclusion untouched

    1. A plural that never declines or takes qualitative modifiers does not revoke the proper‑name status of Βάαλ. Sharp’s canon still excludes proper names and still includes θεός.
    2. Hippolytus’ citation, whether in short or long form, still unites θεός and σωτήρ under one article and applies the whole expression to Jesus. The presence or absence of μεγάλου and the explicit ησο Χριστο does not alter that syntactic unity; it merely shows the normal abbreviation that occurs when catenae copyists reduce biblical lemmata.
    3. Consequently Titus2:13 and 2Peter1:1 remain secure examples of a single‑referent TSKS construction, and the earliest Greek exegesis we possess treats them exactly so. The attempt to evade Sharp’s rule by appealing to Βάαλ’s plural forms or by tightening Hippolytus’ text therefore fails on linguistic and textual grounds alike.

    I regret the earlier conflations and thank you for prompting the corrections. They strengthen, rather than weaken, the case that in both epistles Jesus Christ is grammatically identified as “our God and Saviour.”

Share this

Google+
Pinterest
Reddit