2 Peter 1: 1 Corruption in the NWT

by Sea Breeze 23 Replies latest watchtower bible

  • slimboyfat
    slimboyfat

    Another blast of AI text was not warranted. I don’t know what you think you’re achieving.

  • TD
    TD
    Jana Plátová’s article clarifies the status of its fragments, providing a scholarly foundation to challenge the critique’s skepticism

    My literacy in Czechoslovak is practically non existent, but I did manage to find a copy in German. Apparently professor Plátová recognizes 72 fragments which she categorizes into 3 tiers of reliability ranging from a core set to possible/disputed fragments

    Interesting, but it doesn't help an unwashed heathen like myself who would be interested in the content of fragment 38. Worse, not all scholars observe Plátová's numbering system, which makes things even more convoluted, as it's not clear if her fragment 38 is the one we're talking about here.

  • Earnest
    Earnest
    aqwsed12345 : Sharp’s willingness to refine his rule in response to feedback demonstrates intellectual rigor, not desperation. The exceptions he incorporated—proper names and plurals—arose from observable patterns in Greek usage, ensuring the rule’s precision within defined boundaries.

    So there were exceptions to Sharp's initial "rule". Sharp says "many exceptions" when nouns are proper names or in the plural number. If the purpose of Sharp's rule was not theological (he plainly states it was), then he would have added an exception applying when the first and second substantives in a TSKS string have different referents e.g. 2 Peter 1:1; Titus 2:13. As Calvin Winstanley says :

    "There are, you say, no exceptions, in the New Testament, to your rule; that is, I suppose, unless these particular texts be such; which you think utterly improbable. You would argue, then, that if these texts were exceptions, there would be more. I do not perceive any great weight in this hypothetical reasoning.
    aqwsed12345 : Your assumption that θεός here “must” (?) denote the Father alone is a theological assertion, not a grammatical one, and thus begs the question.

    This is the crux determining whether the first substantive "the God" (τοῦ Θεοῦ) and the second substantive "Saviour" (Σωτῆρος) have different referents. The referent of the first substantive is God the Father, the referent of the second substantive is Jesus Christ. The expression "our God" or "the God of us" (τοῦ Θεοῦ ἡμῶν) is used throughout the Hebrew scriptures to refer to Jehovah without exception (Deuteronomy 4:7; 5:2; 6:4 etc) and in the NT it is associated with the Father explicitly (Galations 1:4; Ephesians 5:20; Philippians 4:20). There is no doubt "our God" refers to God the Father here as it does everywhere else.

    The only reason these verses are not recognised as exceptions is a theological one which insists that "our God" and "Jesus Christ" are the same, contrary to the universal use of the expression elsewhere.

  • aqwsed12345
    aqwsed12345
    @Earnest

    Granville Sharp’s first canon is a descriptive statement about how Koine authors habitually deploy the article, not a theological syllogism devised to rescue later dogma. Its evidential base is the set of Greek sentences in which a single article precedes two singular, personal, common nouns joined by καί. Where that syntactic frame occurs in original (non-translation) prose, the nouns invariably prove to be co-referential. Sharp’s own refinement—excluding proper names and plurals—was not damage control but the normal process of delimiting a rule to the linguistic environment in which it is actually attested. Plural nouns are structurally different because they can embrace multiple individuals inside one grammatical slot; proper names do not ordinarily accept a dependence on a preceding article. Removing these statistically deviant categories left a homogenous data-set with no counter-instance then or since. Hundreds of later searches, including Daniel Wallace’s computer-assisted trawl through the papyri and inscriptions, have confirmed the same distribution. That outcome does not vanish because Sharp reviewed objections: intellectual rigor demands that a putative rule be stress-tested and trimmed to the contours of the evidence it is supposed to explain.

    Sharp’s exclusion of proper names and plurals does not indicate a flaw in the rule’s original formulation but rather a deliberate adjustment to align it with observable syntactic patterns in Koine Greek. Proper names and plural nouns exhibit distinct grammatical behaviors in Greek, often functioning independently or requiring additional articles when referring to multiple entities. Sharp’s decision to limit his rule to singular, personal, common nouns reflects a methodological precision aimed at ensuring its predictive accuracy within a defined scope, not an admission of inherent weakness. To suggest that he should have added an exception for cases where substantives have different referents presupposes that such cases exist within the rule’s parameters and that Sharp ignored them. Yet, no clear counterexample from original Greek prose—biblical or otherwise—has been demonstrated where a TSKS construction with singular, personal, common nouns under a single article unambiguously refers to two distinct individuals.

    Winstanley’s dismissal of Sharp’s consistency as “utterly improbable” and his questioning of the lack of additional exceptions do not constitute a substantive rebuttal, as he provides no specific instance where the rule fails within its stated boundaries. Instead, his argument hinges on skepticism about the rule’s universality, which does not suffice to overturn the extensive corpus analysis supporting Sharp’s observations. Winstanley’s reasoning was that if the rule were valid one ought to expect “more exceptions,” yet he could not marshal a single uncontested example from the NT corpus. “Hypothetical” is in fact the apt label for his critique, not for Sharp’s induction.

    The objection that Sharp should have added another “exception” whenever the two nouns happen to refer to different persons mis-states both his aim and his method. A grammatical rule is not disproved by the wish that it behaved otherwise; it is disproved by a sentence that satisfies the rule’s formal conditions yet, indisputably shows a different outcome. No such sentence has been produced. The texts adduced—2 Peter 1:1 and Titus 2:13—do not falsify the canon; they are the very passages whose interpretation depends upon the canon’s force. Citing them as counter-examples merely begs the question: it assumes the point at issue, namely that the nouns do have distinct referents, and then faults Sharp for failing to anticipate an exception that has never been demonstrated on linguistic grounds.

    Your objection—that το Θεο μν in 2 Peter 1:1 “must” by semantic inevitability denote the father and therefore cannot share reference with σωτήρ—confuses customary usage with grammatical structure. It is true that the phrase “our God” regularly designates the covenant God of Israel and is elsewhere applied to the Father. But Koine writers, including Peter, are free to appropriate traditional theistic language in fresh Christological settings. Syntax tells us when they have done so. In the salutation the single article το governs both θεός and σωτήρ; the two nouns are grammatically welded before the proper name ησο Χριστο explicates who is meant. Had the author wished to preserve two subjects he had a straightforward device to do so: repeat the article, exactly as he does in the next sentence, το Θεο κα ησο το Κυρίου μν, where duplication unmistakably separates the referents. The contrast of vv. 1–2 shows that he relied on article distribution, not on inherited Hebrew collocations, to signal either unity or distinction.

    Appeal to OT precedent is therefore irrelevant to Greek sentence grammar. Even within the Septuagint the string “our God and king” (Ps 99:3 LXX) places θεός and βασιλεύς under one article to speak of Yahweh by dual titles—proof that the idiom of shared definiteness could already override the expectation that θεός identifies a different person from the second noun. Peter writes native Greek, not calque; his readership is expected to parse Greek syntax, not to override it with an a priori rule that οποτε λέγ “our God” the Father is necessarily intended.

    This claim suggests that because “our God” consistently refers to Yahweh in the OT, it must also refer to “Jehovah”—understood as the Father alone—in 2 Peter 1:1. However, this reasoning reflects an Arian theological bias, which a priori equates Yahweh exclusively with the person of the Father. In contrast, the Nicene understanding, which represents orthodox Christian theology, holds that Jesus is also Yahweh, though not identical to the Father in person. The assertion that “our God” always refers to “Jehovah” as the Father reveals an underlying Arian presupposition. In Arian theology, Yahweh is solely the Father, and Jesus is a separate, created being who is not divine in the same sense. This leads to the conclusion that any reference to “our God” must exclude Jesus. However, this is not a neutral reading of Scripture but a theological lens that shapes the interpretation. When the argument insists that “our God” in 2 Peter 1:1 cannot refer to Jesus because it must mean the Father, it assumes this Arian framework without justifying it.

    According to the Nicene Creed, Jesus is fully God, co-equal and co-eternal with the Father, yet distinct in person. In this view, Yahweh is not limited to the Father but is the name of the Godhead—the one divine essence shared by the Father, the Son, and the Holy Spirit. Therefore, when 2 Peter 1:1 calls Jesus “our God and Savior,” it does not imply that Jesus is the Father, nor does it exclude Him from being Jehovah. Instead, it affirms His deity as part of the triune God. Saying that “our God” refers to Jesus does not mean it ceases to refer to Yahweh; rather, it reflects the broader Christian understanding that Yahweh includes the Son.

    The Arian argument begins with the premise that Yahweh (or God) refers only to the Father. When Scripture mentions Jesus alongside “God”, they interpret this as evidence that Jesus is a separate, non-divine entity. This logic underpins their claim: if “our God” is Yahweh, and Jesus is mentioned distinctly, then Jesus cannot be God. However, this reasoning relies on their specific theological jargon and assumptions, not on a universal biblical principle. Mainstream Christianity agrees that the God’s name is Yahweh (or Jehovah) and that Jesus referred to the Father as God, but it does not follow that Yahweh/God equals the Father alone. For Christians, the divine name Yahweh designates the Godhead (theotēs, Colossians 2:9) as a whole, encompassing three persons: Father, Son, and Holy Spirit.

    Believers in mainstream Christian churches worship the same God as the JWs, identified as Jehovah or Yahweh, but they understand Him differently. They do not limit Yahweh to the Father but see Him as the triune God: Father, Son, and Holy Spirit. Thus, calling Jesus “our God” in 2 Peter 1:1 aligns with worshiping Yahweh, because Jesus is part of the Godhead. The Watchtower’s practice of speaking of “Jesus and Jehovah” as strictly separate reflects their theological framework, not a definitive biblical truth.

    It’s true that in the OT (e.g., Deuteronomy 4:7; 5:2; 6:4), “our God” refers to Yahweh without explicit Trinitarian distinctions, as the full revelation of the Trinity emerges in the NT. However, the NT expands this understanding by identifying Jesus as divine (e.g., John 1:1; Titus 2:13; Hebrews 1:8). The grammatical structure of 2 Peter 1:1—using a single article for “our God and Savior Jesus Christ”—further supports the view that Jesus is both God and Savior, consistent with His inclusion in the divine identity of Yahweh.

    Furthermore, the immediate context of 2 Peter reinforces the Nicene reading. In the very next verse (1:2), Peter employs a different construction—το Θεο κα ησο το Κυρίου μν (“of God and of Jesus our Lord”)—with separate articles, explicitly distinguishing between God and Jesus. This deliberate variation within the same passage suggests that Peter adjusts his syntax to reflect intended distinctions or unities. The absence of a second article in 1:1, contrasted with its presence in 1:2, undermines the claim that “God” and “Savior” must refer to different persons in the former. If Peter had intended to separate the referents in 1:1, Greek convention would favor repeating the article, as he does in 1:2.

    Your contention that theological bias alone prevents 2 Peter 1:1 and Titus 2:13 from being recognized as exceptions overlooks the robust linguistic basis for Sharp’s rule. The rule’s validity rests not on a desire to equate “our God” and “Jesus Christ” but on a consistent grammatical pattern observed across a wide range of texts. Scholars have scrutinized this construction in both biblical and extra-biblical Greek, finding no definitive exceptions within the rule’s specified parameters. To establish 2 Peter 1:1 as an exception, one would need to demonstrate that the TSKS construction here unambiguously refers to two distinct individuals, a burden unmet by appealing solely to the customary meaning of “our God.” Such an approach risks imposing an external interpretive framework on the text rather than deriving meaning from its grammatical structure.

    Finally, the charge of “theological bias” runs in the opposite direction. The NWT must add an article in English—“of our God and of the Savior”—in order to break the grammatical bond that the inspired text forges. The entire edifice of the dual-referent reading rests on importing into v. 1 the very syntactic marker (a second article) that Peter conspicuously omits. That maneuver is not exegesis but revision.

    In sum, Sharp’s exclusions sharpen the rule; they do not expose it. No example has been found in genuine Greek where a single-article TSKS string of singular personal common nouns names two people. Until such an example is produced, 2 Peter 1:1 and Titus 2:13 remain ordinary instances of the idiom, identifying Jesus Christ alone as “our God and Savior.” Sharp’s refinements to his rule enhance its precision by accounting for syntactic realities in Greek, not by conceding flaws, and Winstanley’s critique lacks the evidential force to discredit it. In 2 Peter 1:1, the TSKS construction grammatically supports a single referent—Jesus Christ as both “our God and Savior”—despite the broader convention of “our God” referring to the Father. This interpretation aligns with the text’s syntax and context, rendering an exception unnecessary and highlighting the primacy of linguistic analysis over presupposed referential norms.

Share this

Google+
Pinterest
Reddit