In an effort to defend the assertion that two distinct persons are in view in Titus 2:13 and 2 Peter 1:1, the New World Translation (NWT) committee approvingly quotes Abbot’s appeal to two supposedly parallel article-substantive-kaiv-substantive (TSKS) constructions in the New Testament.1 Of course, Abbot’s theory encounters a slight problem here, since neither of his supposed parallels fall within the scope of Granville Sharp’s rule.2 First, his example in Matthew 21:12 of two distinct groups buying and selling in the temple employs the use of plural participles. This is not parallel to Titus 2:13 or 2 Peter 1:1 where two singular nouns are in view, and it ignores the fact that not all plural participles in the TSKS construction require identical referents.3 Abbot’s second example in 2 Thessalonians 1:12 of a distinction between the Father and Jesus Christ also falls outside the scope of Sharp’s rule, since kurivou =Ihsou' Cristou' is a common title possessing the qualities of a proper name.4
In addition to his two proposed parallel TSKS constructions, Abbot appealed to Winer’s argument that the insertion of hJmw'n before swth'ro" in Titus 2:13 definitizes the noun, thus explaining the absence of the second article. However, Paul’s reference in Philippians 2:25 to Epaphroditus as toVn ajdelfoVn kaiV sunergoVn kaiV sustratiwvthn mou (“my brother and fellow-worker and fellow-soldier”) clearly demonstrates that the intrusion of a genitive pronoun does not invalidate Sharp’s rule.5 Both Abbot and the NWT appendix show an imprecise understanding of Sharp’s rule, and the latest NWT apologia6 shows a continuation of that trend.
Despite the somewhat complex approach to his argument, Greg Stafford’s discussion of christologically significant TSKS constructions7 can be broken down rather simply. On the one hand, his arguments depend almost exclusively on one area of limited data,8 and on the other hand, he almost completely ignores an overwhelming corpus of contrary evidence.
Stafford’s argument for viewing two persons in the christologically significant TSKS constructions relies almost exclusively on his attempt to find compound proper names. If he can somehow demonstrate the fact that “the Great God” (Titus 2:13) or “Savior Jesus Christ” (Titus 2:13; 2 Pet 1:1) are compound proper names, then the limitations on Sharp’s rule apply and two distinct persons are intended. However, Stafford has not assembled nearly enough primary data to establish his case.9
Stafford’s first attempt at establishing the case for a compound proper name begins with “the Great God” (tou' megavlou qeou') in Titus 2:13. A one paragraph statement was all that Stafford deemed necessary,10 and his sole argument leans on Wallace’s admission that it is just possible to construe similar phrases “the only Father” and “the God and Father” as virtual proper names.11 Running with this remote possibility, Stafford muses that since the title “the Great God” also appears frequently in the LXX,12 it too might be considered a compound proper name.
It is our contention that frequency of usage in the Old Testament does not automatically prohibit qeoV" mevga" from being applied to Christ in the New Testament, and thus does not constitute a proper name. This can be demonstrated with a brief examination of another title frequently applied to Yahweh in the LXX, and which we believe is securely applied to Christ in the New Testament.
The title “Almighty God” (qeoV" oJ pantokravtwr) refers specifically to the Father over a dozen times in the LXX.13 The same construction occurs nine times in the New Testament, all confined to the Book of Revelation.14 We would submit that within the pages of the Apocalypse the title is applied to both the Father and the Son.
It is not our intention to offer an explanation for each occurrence of qeoV" oJ pantokravtwr in the New Testament, but merely to offer some observations which we believe make it likely that Christ is referred to as “God Almighty” in the Book of Revelation. Stafford agrees that Revelation 1:17 and 2:8 make clear reference to Christ as “the First and the Last” (oJ prw'to" kaiV oJ e[scato"), and he recognizes that the LXX version of Isaiah 44:6 provides an exact parallel in reference to the Father. Nevertheless, he is unwilling to apply the full force of this phrase to Christ in Revelation 22:13.
First, Stafford objects to Christ being the speaker in Revelation 22:12-15, arguing that the appearance of the first person singular pronoun in 22:16 signals a shift in speaker. He cites two supposedly similar shifts in the same book:
The first example is Revelation 1:9, which reads according to the NASB: ‘I, John, your brother and fellow partaker in the tribulation and kingdom and perseverance [which are] in Jesus, was on the island called Patmos, because of the Word of God and the testimony of Jesus.’ Here John, right after ‘the Alpha and the Omega,’ finishes speaking in verse 8, refers to himself in the first-person singular, followed by an explicit identification through the use of his name. Surely no one will argue that this means John is ‘the Alpha and the Omega’ of the previous verse!15
No, we would not argue that John is “the Alpha and the Omega.” But we would argue that the explicit phrase levgei kuvrio" oJ qeov" (“says the Lord God”) in 1:8 clearly identifies the previous speaker. Thus, the sudden change poses no problem in 1:9. The shift is rather obvious, and there is no room for confusion as to the identification of the speaker in either verse. Stafford continues:
The second example is found in Revelation 22:8: ‘I, John, am the one who heard and saw these things. And when I had heard and seen them, I fell down to worship at the feet of the angel who had been showing them to me.’ (NIV) Will anyone conclude based on the opening of this verse that John is the one ‘coming quickly’ in verse 7?16
No, we would not conclude that John is the one “coming quickly.” But we would conclude that John’s admission to having just heard and seen the things of the previous verse clearly distinguishes him from the one speaking those words and showing those things. It seems pretty clear that the examples set forth by Stafford involve contextual indictors which plainly signal a shift in the speaker. However, when we come to Revelation 22:12-16, that is not the case. Once “the Alpha and the Omega, the first and the last, the beginning and the end” (v. 13) begins to speak in verse 12, there are no contextual indicators to hint at a shift in speakers at verse 16. Thus, there is no difficulty with viewing Christ as the one who is speaking for the duration of 22:12-16,17 and he may be understood as identifying himself as the “Almighty God” of 1:8.18
If “Almighty God” (qeoV" oJ pantokravtwr) can be used of the Father frequently in the LXX and yet applied to Christ in the New Testament, then recurrent LXX usage cannot be offered as a reason for denying Christ the title “the Great God” (tou' megavlou qeou').19 This becomes highly significant in Stafford’s treatment of christologically significant texts in the TSKS construction, since Titus 2:13 does not allow him the concession of calling Christ qeov" in some qualified sense. The title “Great God” can imply nothing less than full deity.
Continuing his quest to find a compound proper name in the aforementioned TSKS constructions, Stafford does some spade work with the epithet “Savior Jesus Christ” (swthvr jIhsou'" Cristov") and digs up the following argument:
It is similar with 2 Thessalonians 1:12, where the compound name ‘Lord Jesus Christ’ does not require the article to be considered a second subject. In Titus 2:13 and 2 Peter 1:1 the use of swthvr, together with ‘Jesus Christ,’ puts these example [sic] outside the general category of article-noun-kaiv-noun constructions, which do not have the equivalent of a proper name in either the first or second position.20
As with his first attempt to find a compound proper name, Stafford’s handling of the primary data is found wanting. In fact, in all of his discussion regarding swthvr jIhsou'" Cristov" as a proper name, Stafford does not provide one argument from primary data. He does a good job, however, of locating secondary sources that agree with his position.21
Since we prefer not to give secondary sources primary authority, a closer look at the two pertinent christological titles is in order. According to the current version of acCordance,22 the title “Lord Jesus Christ” (kuvrio" jIhsou'" Cristov") occurs a total of sixty-two times in sixteen different New Testament letters.23 It is particularly significant to note that the title occurs early24 and is evenly distributed throughout the New Testament (see Appendix). Such early, frequent and evenly spread attestation gives abundant evidence for treating the title as a proper name.
On the other hand, the title “Savior Jesus Christ” (swthvr jIhsou'" Cristov") is scarcely attested. It occurs only five times, and is confined to two later writings25 produced in roughly the same time period (see Appendix).26 Considering the fact that its semantic pattern is so far removed from that of kuvrio" jIhsou'" Cristov", it is quite unlikely that swthvr jIhsou'" Cristov" assumes the former’s semantic force. In other words, there simply is not enough primary data to make a case that swthvr jIhsou'" Cristov" functions as a compound proper name.
Interestingly, swthvr jIhsou'" Cristov" never occurs outside of the TSKS construction, and three of its five occurrences undisputedly indicate shared identity with the head noun (cf. 2 Pet 1:11; 2:20; 3:18). This would seem to indicate that Stafford assumes proper nouns can fit the contours of Sharp’s rule,27 despite the fact that Wallace did not find a single reference in which proper nouns in the TSKS pattern were identical.28
In sum, Stafford has not dealt adequately with the primary data, and his arguments fly in the face of all TSKS constructions examined to date. There are not sufficient grounds for viewing qeoV" mevga" or swthvr jIhsou'" Cristov" as compound proper names, and there is yet more compelling evidence that qeov" should be inseparably linked to swthvr in the TSKS construction. We will now briefly turn our attention to this evidence which Stafford has largely ignored, namely, the well-established idiom qeoV" kaiv swthvr.
The cohesiveness of the time-worn expression qeoV" kaiv swthvr has been noted by more than a few scholars,29 but perhaps Harris has most succinctly captured its early force and accompanying significance for TSKS constructions of christological import:
The expression qeoV" kaiv swthvr was a stereotyped formula common in first-century religious terminology (see Wendland), was (apparently) used by both Diaspora and Palestinian Jews in reference to Yahweh, and invariably denoted one deity, not two. If the name jIhsou'" Cristov" did not follow the expression, undoubtedly it would be taken to refer to one person; yet jIhsou'" Cristov" is simply added in epexegesis.30
If the idiom is found to be early, frequent and uniformly asserting singular identity, we would expect such a force to naturally accompany usage in the New Testament.31 Thus, it is not surprising that Stafford is careful to emphasize the admitted lateness of examples Wallace found in The Oxyrhynchus Papyri,32 while forgetting to mention that “there are earlier uses of the phrase circulating in Hellenistic circles—and not a few which antedate the New Testament” (emphasis added).33 A deceptively small footnote in Wallace’s dissertation is loaded with evidence for early attestation:
Cf. the references in BAGR, s.v. swthvr, dating back to the Ptolemaic era [332-30 BCE]. Cf. also L. R. Taylor, The Divinity of the Roman Emperor (Middletown, CN: American Philological Association, 1931), who gives a helpful list in her ‘Appendix III: Inscriptions recording Divine Honors,’ 267-83. Frequently, and from very early on, the inscriptions honor the Roman emperors as qeov", swthvr, and eujergevth". Almost invariably the terms are in a TSKS construction (among the earliest evidence, an inscription at Carthage, 48-47 BCE, honors Caesar as toVn qeoVn kaiV aujtokravtora kaiV swth'ra; one at Ephesus honors him as toVn . . . qeoVn ejpifanh' kaiV . . . swth'ra; Augustus is honored at Thespiae, 30-27 BCE, as toVn swth'ra kaiV eujergevthn; and in Myra he is called qeovn, while Marcus Agrippa is honored as toVn eujergevthn kaiV swth'ra).34
Stafford seems unimpressed by the evidence that qeoV" kaiv swthvr was a stereotyped formula commonly used in the first century, and assumes that “this is quite beside the point,”35 since Paul clearly applied the title swthvr to both the Father and Christ. He continues by saying that “we must not arbitrarily assume that just because the two titles ‘God’ and ‘Savior’ are used together in such close proximity that they ipso facto apply to one person.”36 Stafford’s use of rhetoric is a bit much here, since the abundance of evidence is far from arbitrary. The fact is, Stafford has not adequately addressed the evidence for the idiomatic force of qeoV" kaiv swthvr in extra-biblical literature. That the title swthvr is applied by Paul to both the Father and to Christ does not make the idiom beside the point—it is the point.
Stafford seems to realize the implications of this, and offers the concession that even if qeoV" kaiv swthvr were an established idiom, it would merely provide yet another qualified reference to Jesus as qeov".37 This qualified notion is difficult to maintain in light of the apparent interchangeability of qeov" and kuvrio" as swthvr (cf. Titus 2:10, 13; 3:4, 6),38 and does not accord with other uses of qeov" in the TSKS construction where one person is clearly in view.39
Evidence presented suggests swthvr is best understood as a title and its historical inclusion in the idiomatic construction qeoV" kaiv swthvr makes separation of the combination difficult in Titus 2:13 or 2 Peter 1:1. The fact that such an idiom antedates the New Testament and always deifies one person40 squarely places the burden of proof on one wishing to break the construction in the New Testament.
Although a few possible exceptions to Sharp’s rule were found by Wallace in extra-biblical literature,41 he notes that “the phrase oJ qeoV" kaiV swthvr (Titus 2:13 and 2 Pet 1:1) admitted of no exceptions—either in Christian or secular writings.”42 Thus, the evidence presented requires that those who wish to see two persons43 in Titus 2:13 and 2 Peter 1:1 produce contrary evidence on two fronts. First, they must demonstrate that qeov" or swthvr should be excluded from Sharp’s rule because they appear in compound proper names. Second, they must provide evidence that oJ qeoV" kaiV swthvr ever refers to more than one person in Hellenistic Greek. Until such evidence is produced, there is little left to discuss.
2 Wallace offers the following restatement of Granville Sharp’s rule: “In native Greek (i.e., not translation Greek), when a single article modifies two substantives connected by kaiv (thus, article-substantive-kaiv-substantive), when both substantives are (1) singular (both grammatically and semantically), (2) personal, (3) and common nouns (not proper names or ordinals), they have the same referent.” Daniel B. Wallace, “The Article with Multiple Substantives Connected by Kaiv in the New Testament: Semantics and Significance” (Ph.D. dissertation, Dallas Theological Seminary, 1995) 134-35.
7 Though Stafford discusses the TSKS construction in Ephesians 5:5, 2 Thessalonians 1:12, Titus 2:13 and 2 Peter 1:1, our interest lies solely in the final two passages. The first two passages involve proper names and thus do not indicate identity between the two nouns. Nonetheless, the TSKS construction always indicates some sort of unity between two nouns, and may even connote equality. See Wallace, Exegetical Syntax, 270. Thus, Ephesians 5:5 and 2 Thessalonians 1:12 may very well be implicit references to Christ’s deity.
8 We are referring to Stafford’s attempt to establish swthvr jIhsou'" Cristov" as a compound proper name in Titus 2:13 and 2 Peter 1:1. Less effort is spent arguing for patristic exceptions to Sharp’s rule and qeoV" mevga" as a proper name in Titus 2:13, and we address those arguments in turn. Stafford’s only other argument appears under 2 Peter 1:1, where he suggests two persons should be in view since New Testament epistles typically make an opening reference to both the Father and the Son. See Witnesses Defended, 246. However, a simple comparison of the opening verses in 1 and 2 Peter make this unlikely. 1 Peter 1:1 makes an opening reference to Christ, and no mention of the Father occurs until the following verse. So, it is not so strange that Peter’s second epistle would mention Christ alone in 1:1, and wait to mention the Father until verse 2.
9 With Wallace we agree that, “Any significant statements as to the semantics of a given construction must be based on a large number of examples,” and that, “Any judgment as to the semantics of the disputed passages must be based on clear examples that parallel, in all the essentials, the semantic situation of the target construction.” Wallace, Exegetical Syntax, 1-2. We will see that Stafford’s argument for compound proper names in Titus 2:13 and 2 Peter 1:1 are lacking in this regard.
11 Wallace rejected this view on the grounds that one would then also expect phrases such as “the God over all” (oJ ejpiV pavntwn qeov") and “Almighty God” (qeoV" oJ pantokravtwr) to be used only of the Father. However, these phrases are not restricted to the Father, and do occasionally refer to Christ in the patristic writings. See Wallace, “Multiple Substantives,” 268-69.
12 Stafford cites Abbot’s list of examples where qeoV" mevga" occurs in the LXX, Witnesses Defended, 239 n. 70. See (as numbered in the LXX) Deut 7:21; 10:17; 2 Chron 2:4; Neh 1:5; 8:6; 9:32; Ps 77:13; 85:10; Dan 2:45; 9:4.
14 See Rev 1:8; 4:8; 11:17; 15:3; 16:7, 14; 19:6, 15; 21:22.
17 Of course, Stafford would find a difficulty based upon theological a priori; i.e., Jesus cannot be called Almighty God because he cannot be Almighty God. It is interesting to note that for all of Stafford’s complaints regarding the misrepresentation of Jehovah’s Witnesses, he himself misrepresents what Trinitarians are actually saying. This is evident by his statement that “To teach Christ has two-natures in one person is not only against the teaching of Scripture, but it flies in the face of logic and reason,” Witnesses Defended, 65. We know of no responsible Trinitarians who attempt to explain the Trinity under the guise of logic and reason. To the contrary, Trinitarians try to deal faithfully with the text of Scripture while remembering that our finite minds cannot fully grasp the nature of our infinite God (cf. Isa 55:8-9).
While we cannot produce a one-to-one correspondence between the reality of God’s nature and human experience, this does not mean that our beliefs regarding God’s nature are illogical. If that were the case, then it would be unreasonable to conclude that God is an eternal being, since our finite minds cannot comprehend a person who has no beginning. Nonetheless, we confidently affirm God’s eternality because it is something which he has revealed about himself. In our discussion of God’s triune nature, it is important to distinguish between that which defies logic and that which is merely incomprehensible to the human mind.
18 It seems clear that Christ is speaking through the angel beginning in 22:12, and this raises a rather interesting point. Beginning with the announcement that the Apocalypse is the “Revelation of Jesus Christ” communicated “by His angel” (cf. Rev 1:1), one is hard pressed to find the angel ever speaking as a conduit of the Father. Thus, it is all the more likely that Christ is “the Alpha and the Omega” speaking in 22:13.
Furthermore, the grammar of Revelation 22:13 suggests that Christ is the one who is speaking: “There can be little doubt that these are the words of Jesus since (i) the phrase =Idou e]rcomai tacuv recalls Jesus’ words e]rcomai tacuv in 2.16 and 3.11, and (ii) in 22.20 the wordsnaiv e]rcomai tacuv are followed by the response, =Amhn, e]rcou kuvrie =Ihsou`. There is no reason to suppose that the subject of 22.13 is different from that of 22.12.” Peter R. Carrell, Jesus and the Angels: Angelology and the Christology of the Apocalypse of John (Cambridge: Cambridge University Press, 1997) 116.
24 The first occurrence is found in the direct discourse of Acts 11:17. Schaff places the conversion of Cornelius as early as 37 CE, thus opening the possibility that the title was used before 40 CE. See Philip Schaff, History of the Christian Church, 3rd rev. ed. (Peabody: Hendrickson, 1996) 1:278.
25 The title occurs late enough for one author to view 2 Peter 1:1 as the evolutionary climax of swthvr being applied to Christ in ways previously applied to the Father. The implication is that calling Jesus swthvr becomes just as significant as calling him qeov". See C. H. Moehlmann, The Combination Theos Soter as Explanation of the Primitive Christian Use of Soter as Title and Name of Jesus (Rochester: Du Bois, 1920) 17.
27 Indeed, Stafford explicitly states that “In Titus 2:13 and 2 Peter 1:1 the use of swthvr, together with ‘Jesus Christ,’ puts these examples outside the general category of article-noun-kaiv-noun constructions, which do not have the equivalent of a proper name in either the first or second position. This is not to say that such constructions cannot describe one person with two nouns, for, clearly, in the case of 2 Peter 1:11, 2:20, 3:18 and Jude 4, they do,” Witnesses Defended, 247-48.
31 It is interesting to note that the occurrence of other proper names immediately following swthvr in the TSKS construction never break the force of singular identity. A search of PHI Documentary CD #7 (Los Altos, CA: Packard Humanities Institute, 1991-96) surfaced ten examples, four of which contain no significant textual ambiguities. See oJ [b]asileuV" kaiV swthVr Ptolemai/'o" in Aegean Islands [general] Inscriptiones Graecae XII,7 [Amorgos] document 506, a, 11; tw'i ktivsthi kaiV swth'ri Aujtokravtori JAdrianw'/ inAttica, Inscriptiones Graecae II et III [2789-5219] document 3370, 2; toi/'" . . .qeoi'" kaiV Swth'ri Ptolemaivwi inDelos, Inscriptiones Graecae XI,4 [510-1349] document 1038, 25; tw'/ eujergevth/ kaiV swth'ri JAdrianh'" Moyouestiva" th'" Kilikiva" in Italy and the Occident, Inscriptiones graecae urbis Romae [IGUR I-IV] document 1:24, 8.
32 Stafford rejects the papyrological examples, arguing that they were written with the Trinitarian concept of God in mind, Witnesses Defended, 241 n. 76. However, it is significant that the Latin fathers did not see Titus 2:13 or 2 Peter 1:1 as necessary formulae for the deity of Christ, largely because the article does not exist in Latin. So to argue that the Greek papyri are influenced by Trinitarianism does not explain why the examples were packaged in the TSKS construction in the first place. Thus, “the uniformity in the Greek fathers was probably due to Greek syntax, not to nascent creedalism,” Wallace, “Multiple Substantives,” 252.
38 As with Paul’s apparent interchangeability between qeov" and kuvrio" as modifiers of swthvr (cf. Titus 2:10, 13; 3:4, 6), Peter may have been using qeov" as a variant for kuvrio" in 1:1. Harris notes a further interchangeability between hJmevra kurivou in 3:10 and tou' qeou' hJmevra" in 3:12. See Harris, Jesus as God, 232.
40 Harris, Jesus as God, 178-79; 233-34; Wallace, “Multiple Substantives,” 255. It is also significant that Titus 2:14 refers back with o}" e[dwken eJautovn as if only one person were in view. See G. W. Knight, The Pastoral Epistles (Grand Rapids: Eerdmans/Carlisle: Paternoster, 1992) 323. Contextually, Christ is shown to be the great God and Savior by means of his sacrifice to achieve redemption. See Harris, Jesus as God, 270.
41 Stafford notes two patristic exceptions in particular, namely, the reference to “the God and Father and Holy Spirit” (tw'/ qew'/ kaiV patriV kaiV aJgivw/ pneuvmati) in Martyrdom of Polycarp 22:1, and “the only Father and Son” (tw'/ monw'/ patriV kaiV uiJw'/) in Clement of Alexandria’s Paedogogus 3.12.101. Wallace points out that these texts are confined to second or early third century patristic literature, involve only members of the Trinity and involve two terms to describe the Father. See Wallace, “Multiple Substantives,” 268. That the early fathers had not yet articulated a full Christology void of occasional modalism is evident. Burgess notes that “Polycarp utters words which serve as the earliest known doxology in which the Holy Spirit is exalted together with the Father and the Son.” He further says that “Clement follows Neoplatonic doctrine which makes heavy use of negative theology,” and thus “does not. . . attempt any formal definition of the Trinity nor of any Member thereof.” It is significant that Burgess ultimately describes Clement’s treatment as “highly speculative.” See Stanley M. Burgess, The Holy Spirit: Ancient Christian Traditions (Peabody, Hendrickson, 1984) 20, 70. For a full discussion of patristic exceptions to Sharp’s rule, see Wallace, “Multiple Substantives,” 267-72.
43 It is important to note that a lack of affirmation regarding a single identity in Titus 2:13 or 2 Peter 1:1 is not the same as a denial of singular identity. Stafford is careful to point out how Moule “agrees that in passages such as Titus 2:13 it is possible that two persons are in view,” Witnesses Defended, 228 n. 28. Two comments regarding this statement are in order. First, in a personally signed letter dated September 22, 1998, Moule stated that, “[Daniel B.] Wallace’s investigation has indeed narrowed the possibility that two persons are referred to in Tit. 213 and 2 Pet. 11.” Second, Trinitarianism does not stand or fall on the interpretation of these verses. To the contrary, Stafford must categorically deny the possibility that either verse has one person in view (noting previous evidence against a qualified sense for qeov") or his entire theological system is jeopardized.