January 21, 2016 4 Comments
Grammatical-Syntactical Analysis of John 5:27b
We now turn specifically to the anarthrous PN-CV construction in 5:27b:
ὅτι υἱὸς ἀνθρώπου ἐστίν
ὅτι [PN:] υἱὸς ἀνθρώπου [CV:] ἐστίν
hoti [PN:] huios anthrōpou [CV:] estin
because huios anthrōpou He [the Son of God] is
It will be necessary to begin the discussion with a ‘rule’ put forth by E. C. Colwell, since it tends to be misapplied, even partly by Colwell himself in the establishment of his own rule.87 Knowing Colwell’s methodology will prove helpful in determining how to properly implement his findings.
Colwell surveyed NT syntactical constructions in which the predicate nominative is without the article and precedes its copulative verb (anarthrous PN-CV constructions) in which the PN was, by his estimation, “indubitably definite.”88 Importantly, Colwell did not consider either indefinite or qualitative PNs as part of his final analysis. In his study, he barely mentions indefinite PN-CVs at all; however, he specifically excludes all qualitative PNs because they are “not definite” and declares that the total amount in the NT is “small” anyway.89 Yet, as noted above, definite and indefinite nouns can sometimes be simultaneously qualitative.
Colwell also considered anarthrous CV-PN constructions. Of the 123 total “indubitably definite” anarthrous PNs he counted in the NT,90 97 were PN-CV (79%), while 26 were CV-PN (21%).91 What we do not know via Colwell is: (a) how many total anarthrous PN-CV constructions there are in the NT;92 (b) of this total how many were determined to be (primarily) qualitative or indefinite as opposed to definite; and, (c) the relative distribution of the three compared to each other.
Colwell also noted that there are incidences of arthrous PN-CV constructions. Recall that all arthrous nouns are definite, no matter the context. Colwell added these 15 arthrous PN-CV constructions to his 97 “indubitably definite” anarthrous PN-CVs, concluding that 87% of the time a definite PN does not have the article when it precedes the CV,93 thus resulting in his general rule: A definite predicate nominative has the article when it follows the verb, it does not have the article when it precedes the verb.94 Of course, this percentage is only valid to the extent of Colwell’s accuracy in predetermining definite anarthrous PN-CV constructions. He concedes this point, though he explains that he endeavored “to exclude all nouns as to whose definiteness there could be any doubt.”95 From his analysis Colwell claims that, with respect to translation and interpretation, he has
show[n] that a predicate nominative which precedes the verb cannot be translated as an indefinite or a “qualitative” noun solely because of the absence of the article if the context suggests that the predicate is definite, it should be translated as a definite noun in spite of the absence of the article.96
This statement is not untrue considering his important conditional clause “if the context suggests that the predicate is definite.” Yet, in building his case for the incidences of definite anarthrous PN-CV constructions, he compared the arthrous – and hence, definite – the Son of Man (ho huios tou anthrōpou) in the CV-PN construction of Matthew 13:37 to the anarthrous PN-CV huios anthrōpou of John 5:27b, ‘proving’ that John 5:27b is definite.97 This exemplifies at least one methodological error. Colwell assumed that the definite usage of a PN in one context necessarily renders that same PN definite in another, in this case partially upon his presumption that the expression in both texts is a “title.”98 Moreover, it may be that he presupposed definiteness because of the anarthrous PN-CV construction (and perhaps because of the myriad instances of the Son of Man in the Gospels). That is, he may not have adequately assessed the context, merely imposing his hypothesis upon the text, for, on the surface, it does not appear he could declare John 5:27b “indubitably definite.” If true, he would have been assuming the converse of his own rule in determining definiteness, a practice found elsewhere in his work.99 Many others have done this very thing, citing Colwell and erroneously presuming that ‘an anarthrous PN is definite when it precedes the CV’100 – even though Colwell did not examine all such NT constructions. But, as Dixon points out, this is demonstrably false:
[T]he converse of Colwell’s rule . . . is not a valid inference . . . From the statement “Articular nouns are definite,” it is not valid to infer “Definite nouns are articular.” Likewise, from the statement “Definite predicate nominatives preceding the verb are anarthrous,” it is not valid to infer “Anarthrous predicate nominatives preceding the verb are definite.”101
To rephrase, as we noted above, sometimes definite nouns lack the article, so it would be fallacious to state ‘definite nouns are arthrous.’ Similarly, sometimes when a PN precedes its CV the noun is other than definite. The point here is that Colwell’s findings merely allow the possibility that an anarthrous PN-CV construction be definite. ‘Colwell’s rule’ states “nothing about the probability of definiteness.”102 Hence, context must be the first consideration. If the context suggests definiteness, then it is grammatically permissible to render it definite, per a proper interpretation of Colwell’s work.
Yet, according to the analysis of Harner and Dixon, in the Gospel according to John, an anarthrous PN is most times qualitative when it precedes its CV.103 Specifically, Harner interprets these constructions as primarily qualitative, and this qualitativeness “may be more important than the question whether the predicate noun itself should be regarded as definite or indefinite.”104 In his article he concludes:
. . . [A]narthrous predicate nouns preceding the verb may be primarily qualitative in force yet may also have some connotation of definiteness. The categories of qualitativeness and definiteness, that is, are not mutually exclusive, and frequently it is a delicate exegetical issue for the interpreter to decide which emphasis a Greek writer had in mind.105
In fact, of the 53 occurrences of anarthrous PN-CV constructions in John’s Gospel, Harner found 40 primarily qualitative in force.106 Dixon, on the other hand, deemed 45 of the 53 to be with a qualitative emphasis,107 with another five “probably qualitative.”108 While Harner does not count John 5:27b as qualitative,109 this clause is specifically identified by Dixon as “probably qualitative, but could be definite.”110 Assessing the work of Harner and Dixon along with his own analysis, Wallace states a general rule, thus revising Colwell: “An anarthrous pre-verbal PN is normally qualitative, sometimes definite, and only rarely indefinite.”111
Given the exegetical findings of the previous section in conjunction with syntactical probabilities, the position taken here is that υἱὸς ἀνθρώπου in John 5:27b is primarily qualitative and secondarily definite, revealing the μονογενὴς (monogenēs, unique, one-of-a-kind; 1:14, 18) Son of God’s incarnational status. Rather than making a one-to-one correspondence (because He [the Son of God] is the Son of Man), the focus in this verse is upon His incarnational humanity over against His deity.112 Thus, it should be interpreted: because the Son of God is son of man, i.e. because the Son of God is human. In other words, this clause should be understood: ‘because the Son of God (also) possesses all the qualities and characteristics consistent with being human.’ This seems the best way to understand John 5:27b in view of its immediate context, its contrast with yet connection to ὁ υἱὸς τοῦ ἀνθρώπου throughout John’s Gospel as illustrated in the previous sections above, its agreement with parts of the larger Johannine corpus (Revelation 1:13 and 14:14), and its allusions and references to the OT (especially in relation to Daniel 7:13-14, 7:26, 12:1c-2, but also to include general usage of υἱὸς ἀνθρώπου in the LXX113).
Some commentators perceive the qualitative force in John 5:27b. Barrett, e.g., recognizes the qualitative emphasis here, with an underlying definiteness,114 quoting Schlatter in agreement in this regard:
It is unnecessary here to use the articles because ‘in this context his uniqueness is perfectly clear. It arises out of the uniqueness of his status as Son of God . . . But here the emphasis lies upon the fact that he belongs to humanity as he who took the measure of life appointed to men.’115
Hare, also, seems to affirm a qualitative-definite force noting, importantly, that while the anarthrous form is “not identical” to the arthrous, he sees the expression in 5:27 as combining both:
[H]uios anthrōpou does not serve as a name but expresses a quality or status, yet its connotative force appears to be the same as that of the fuller appellative. Both forms of the phrase can refer to the humanity of the Word that became flesh for our salvation.116
Aune, following Hare, also asserts that the anarthrous υἱὸς ἀνθρώπου in 5:27 “expresses a quality or status, i.e., the incarnate status of the Son.”117 Additionally, Augustine affirms the view adopted here, as “[h]e is the one who will come and it is in the form [of man] that he will come [to judge].”118 As well, Westcott construes a qualitative-definite rendering, with the understanding that this is due to Christ’s role as Judge:
The prerogative of judgement is connected with the true humanity of Christ (Son of man), and not with the fact that He is the representative of humanity (the Son of man). The Judge, even as the Advocate (Heb. 2:18), must share the nature of those who are brought before Him. The omission of the article concentrates attention upon the nature and not upon the personality of Christ.119
Similarly, Godet recognizes that it’s necessarily from Christ’s humanity that He judges mankind, asserting a qualitative priority in John 5:27b:
It emphasizes the relation between the character of judge and that of Son of man. What is this relation? . . . The term, Son of man, without the article sets forth simply the quality of man which He shares with all other men . . . The quality of man is made prominent here for the purpose of explaining, not the dignity of Saviour, but that of judge. The judgment of humanity is a homage rendered to the holiness of God; but this homage, in order really to make reparation for the outrage committed, must proceed from the [human] race itself which has committed the offense. Judgment, in this view, is exactly on the same line with expiation, of which it serves as the complement. Expiation is the reparation freely offered believing humanity; judgment is the satisfaction which God takes from humanity which has refused Him this reparation. In the one, as in the other, of these acts, a man must preside.120
As evidenced by the foregoing, our position that John 5:27b should be understood as qualitative-definite finds theological and contextual validity, syntactical grounding, and support from some commentators. Strict definiteness, while viable, seems outweighed by contextual and theological considerations. Moreover, if the Gospel writer wished to make it clear that his intention in 5:27b was to indicate definiteness he could simply have utilized both articles and/or placed an arthrous PN after the CV – the latter construction used 66 times in his Gospel.121 Indefiniteness, though grammatically possible, is “the most poorly attested” of the three options involving a PN-CV construction,122 and seems unlikely in this context, for it could imply that a mere man is qualified to be eschatological judge.
87 E. C. Colwell, “A Definite Rule for the Use of the Article in the Greek New Testament,” Journal of Biblical Literature 52 (1933): 12-21. See Wallace, Grammar, on how Colwell himself erroneously applied his own ‘rule’ (pp 258-259; cf. 259-262).
88 Colwell, “Definite Rule,” p 17. As Dixon, “Anarthrous Predicate Nominative in John,” observes: “Thus, what is to be asserted is not definiteness, but articularity” (p 13). Cf. Wallace, Grammar, p 262.
89 Colwell, “Definite Rule,” p 17.
90 Colwell, “Definite Rule,” p 17. Colwell also specifically excludes relative clauses (pp 16-17). He states that this 123 total is not necessarily 100% accurate – he may have missed a few – and this fact does not necessarily materially affect his results (“Definite Rule,” p 16 nt 10). I’ll agree that if the total were actually, say, 125 as opposed to his 123 that this would not detract from his analysis on this particular point. It’s what he concludes with this data that is problematic, as we will see.
91 Colwell, “Definite Rule,” p 17.
92 Philip B. Harner, “Qualitative Anarthrous Predicate Nouns: Mark 15:39 and John 1:1” (Journal of Biblical Literature 92 (1973): 75-87) counts 53 such constructions in John’s Gospel alone, with another eight in Mark (p 82). However, Harner construed most of these as emphasizing a qualitative force. More on this below.
93 Colwell, “Definite Rule,” p 17.
94 Colwell, “Definite Rule,” p 13. Emphasis added.
95 Colwell, “Definite Rule,” p 17.
96 Colwell, “Definite Rule,” p 20. Emphasis added
97 Colwell, “Definite Rule,” p 14.
98 Colwell, “Definite Rule,” p 14. Dixon, “Anarthrous Predicate Nominative in John,” identifies this sort of error (pp 19-22). Moreover, as Hare, Son of Man, illustrates, Matthew 13:37 is the only CV-PN construction of the expression in the entire NT, and the context is within a parable, which, given its allegorical nature, necessitates this form, in his opinion; cf. Gal 4:24f (p 151).
99 See Colwell’s (“Definite Rule”) exegesis of John 1:1c (p 21), which is discussed specifically below (see note 126 and corresponding text). Wallace (Grammar) recounts how he learned that Colwell considered the converse of his rule as valid as the rule itself (p 259, esp. ftnt 11).
100 As but one example, in an effort to refute the Arianism inherent in the Jehovah’s Witnesses’ New World Translation rendering of John 1:1c (“and the Word was a god”), Walter Martin [The Kingdom of the Cults: The Definitive Work on the Subject, Revised, Updated, and Expanded Edition, gen. ed. R. Zacharias (Bloomington, MN: Bethany House, 2003)] misapplied Colwell’s Rule, which, ironically, resulted in his implied support of Sabellianism (modalism) – a heresy he had staunchly opposed. Citing Colwell, Martin dogmatically declared: “the Greek grammatical construction leaves no doubt whatsoever that this [and the Word was God] is the only possible rendering . . . Colwell formulated a rule that clearly states that a definite predicate nominative (in this case theos – God) never takes an article when it precedes the verb (was), as we find in John 1:1[c] . . . [T]herefore . . . no article is needed . . . and to translate it ‘a god’ is both incorrect grammar and poor Greek” (p 108). The reason this can support Sabellianism is that in John 1:1b (and the Word was with [the] God (ho theos)) God (ho theos) can be understood contextually as the Father, and by claiming that the PN is definite in which “no article is needed” Martin affirmed – whether intentional or not – a fully convertible A = B / B = A proposition: the Word = God (the Father) and God (the Father) = the Word. See A. T. Robertson, [A Grammar of the Greek New Testament in the Light of Historical Research, 4th ed. (Nashville, TN: Broadman, 1934 (1914)), pp 767-768] who references this very clause, adding: “[W]hen the article occurs with the subject . . . and predicate, both are definite, treated as identical, one and the same, and interchangeable” (p 768). Cf. Wallace, Grammar, pp 258, 268. [More on John 1:1c in the next section below.]
Martin went further astray in his statement that a definite noun in a PN-CV construction “never takes an article,” for, per Colwell, 15 out of 112 are arthrous (Colwell, “Definite Rule,” p 17). Hence, not only did Martin erroneously affirm the converse of Colwell’s rule, he mistakenly asserted that there can never be an articular PN-CV construction.
101 Dixon, “Anarthrous Predicate Nominative in John,” (hereafter simply “Dixon”) pp 11-12.
102 Dixon, p 18.
103 Harner, “Qualitative Anarthrous Predicate Nouns,” pp 75-87 (hereafter simply “Harner”); Dixon, pp 1-61.
104 Harner, p 75.
105 Harner, p 87. Emphasis added.
106 Harner, pp 82-83.
107 Dixon, p 32.
108 Dixon, p 34. These are identified as 1:49, 5:27, 9:5, 17:17, and 19:21.
109 This verse is not enumerated in his list of 40 in which “the qualitative force of the predicate is more prominent than its definiteness or indefiniteness” (Harner, pp 83, 83 nt 20). While Harner specifically engages some anarthrous PN-CV constructions in John, 5:27b is not one of them, and no specific reason was given for the non-inclusion of this verse in his list of 40. However, he does state: “Some degree of subjectivity is unavoidable . . . and the interpretation of some examples is uncertain” (p 83). Perhaps this means John 5:27b is, in his opinion, one of those that are “uncertain.”
110 Dixon, p 33; cf. 56.
111 Wallace, Grammar, p 262. Italics in original, bold added.
112 It is the surrounding context that is expressing Jesus’ deity via His divine functioning, not the Son of God idiom itself, as discussed above.
113 To include especially Psalm 8:4, 80:17, and 144:3 in their larger contexts. Leung, Kingship-Cross Interplay, states that “Son of Man” in Psalms 8:4, 80:17, and 144:3 are referenced in the post-NT Targums as “evidently messianic” (p 70). Yet the author includes John 5:27 among these “son of man” references, assuming it is definite along with all 12 other particularized usages of the term in John’s Gospel. This illustrates (a) the very point we’ve been making here that the understanding of a messianic the Son of Man only came after the NT had been written; and, (b) that this author, like many others, imposes particularity and definiteness upon 5:27b.
114 Barrett, Gospel According to St. John, p 262. The commentator also states: “Everywhere else in John both articles are used . . . because the phrase is here anarthrous it has been suggested that its meaning is not ‘the Son of man’; Jesus is qualified and authorized to judge because he has shared the experiences of men as one of themselves” (p 262).
115 Barrett, Gospel According to St. John, p 262 (emphasis added), who cites Adolf Schlatter, Der Evangelist Johannes, wie er spricht denkt und glaubt: Ein Kommentar zum vierten Evangelium (Stuttgart: Calwer, (1930) 1958), p 152. [Translation: John the Evangelist, As He Speaks, Thinks, and Believes: A Commentary on the Fourth Gospel.]
116 Hare, Son of Man, p 96; cf. pp 90-96.
117 Aune, Revelation 1-5, p 90.
118 Augustine of Hippo, “Sermon 127.10,” in Joel C. Elowsky, ed. Ancient Christian Commentary on Scripture, New Testament, IVa: John 1-10, Thomas Oden, gen. ed. (Downers Grove, IL: InterVarsity, 2006), p 199. Brackets in original.
119 Westcott, Gospel According to St. John, 1, p 194; parentheses in original.
120 Godet, Commentary on the Gospel of John, Vol. 1, pp 477-478; italics in original.
121 For this quantity, see Harner (p 82) and Dixon (p 24). Delbert R. Burkett [The Son of the Man in the Gospel of John, (JSNTSup, 56; Sheffield: Sheffield Academic Press, 1991)], who claims the usage here is indefinite (p 43), asserts that there are two other syntactical choices the Gospel writer could have used to mark the appellative as definite – though he also calls it a title – structures the biblical author utilized elsewhere in the Gospel, as “[t]he second article in [PN-CV] constructions is regularly retained . . .” (p 42). These two options are: (a) employing an anarthrous υἱός before the CV and τοῦ ἀνθρώπου after; (b) placing υἱὸς τοῦ ἀνθρώπου before the CV (p 42). For (a) he appeals to 1:49, 9:5, 10:2, 12:31, and 19:21; for (b) he compares to 8:39 and 10:36. Regarding Burkett’s supporting texts for (b), both Dixon (p 56) and Harner (p 83 ftnt 20), in contrast to Burkett, find the two to be primarily qualitative. I’m inclined to agree.
As to the supporting verses for (a), Harner explicitly agrees with the position taken by Burkett regarding 1:49 and 9:5 (pp 83-84) and implicitly agrees regarding 19:21 (p 83); however, Dixon finds these three verses “probably qualitative, but, possibly definite” (pp 19-22, cf. 41-44). 10:2 is determined to be qualitative by both Dixon (p 56) and Harner (p 83 ftnt 20). 12:31 does not appear to fit the pattern and is not assessed by either Harner or Dixon. Regarding 10:2, we concur with Harner and Dixon. On the other three verses we remain ambivalent: we can see the argument of Harner, and we can sympathize with the position of Dixon.
While it may be true that the Gospel writer would retain the second article in such constructions, given the evidence, to include sample size, I’m not so sure we can make any definitive conclusions as to the biblical author’s intentions regarding in/definiteness or qualitativeness, contra Burkett. In fact, it seems likely that Burkett deems these constructions definite on their face merely because he presupposes they are all titles.
122 Wallace, Grammar, p 267.