I’ve blogged a fair bit about Eph 5.22 over the years, culminating in my argument for one of the longer readings in NTS back in 2021. Today, Joey McCollum of Australian Catholic University has a new article out in JSNT extending my argument. He gives much more attention than I could to the internal evidence and concludes in favor of the third person plural imperative. It’s especially helpful to have his thoughts on the function of the third person imperative, a question I only barely touched on in my essay but one that commentators especially need to consider.
Here’s the abstract:
This study revisits a contested textual variant concerning the presence, placement, and person of an imperative directed at wives in Eph. 5.22. Most previous treatments of this variant have decided the matter (typically in favor of the reading without an imperative) on the basis of manuscript support and transcriptional arguments about how readers and copyists of the text would have changed it, but the intrinsic probabilities of what the author would have written based on his argument and style have generally been neglected. This study fills this gap by assessing the intrinsic probabilities of the variant readings in Eph. 5.22 using discourse and information structure, the pragmatics of the Greek imperative, and stylistic observations in Ephesians. As a result of this analysis, the reading with the highest intrinsic probability is shown to be τοῖς ἰδίοις ἀνδράσιν ὑποτασσέσθωσαν, which bolsters the recent case made by Gurry (2021) for the same reading.
Thankfully, the article is open access too. Read it here.
I don't agree with Gurry and McCullum on the external evidence. The addition of the verb makes the submission of women more explicit, and is just the kind of thing that we should expect, since it is both clarifying and potentially sexist. The proposed eye skip that would eliminate the verb, has only two letter repeating, with a different letter between the two, so it is surely much less likely.
ReplyDeleteThe main pillar of Gurry's argument is that scribes would fill in the blank in 5:22 with a second person imperative, rather than a third person imperative. However, they may have added the third person imperative in the belief that Paul was addressing the men, telling them to make the women submissive. Also, McCullum provides a reason when he writes, "In particular, the language used for the object of the imperative, τοῖς ἰδίοις ἀνδράσιν, fits a third-person imperative better than a second-person imperative (Gurry 2021: 578). More common phrases like τοῖς ἀνδράσιν ὑμῶν and τοῖς ἑαυτοῖς ἀνδράσιν would be readily available for a second-person imperative." Furthermore, the third person imperative was almost certainly added by Clement of Alexandria. When he quotes 5:21-22 he gives no verb in 5:22. When he quotes 5:22 on its own he is forced to provide a verb, and it is the third person imperative that he supplies. I think Gurry is wrong to suppose that Clement had a manuscript with the verb. I agree with NA28 that Clement supports only the shorter reading. McCullum discusses this in note 10 but does not point out that it undermines Gurry's contention that people would not have added the third person imperative.
I think Gurry is too quick to dismiss the evidence of the varying positions of both forms of the verb. He writes, "That is indicative of nothing more than normal scribal mistakes. After all, if we assumed that every transposition was the result of scribes adding to the text, we would have little Greek New Testament left of which to speak." The DFG ancestor preserves independent ancient readings, and has a tendency to insert additions from the margin to the wrong place in the text. Transpositions often result when shorter readings are "corrected". See JSNT forthcoming.
Let's move on to the internal evidence. Arguing against the shorter reading, Gurry writes, "I suspect no reader of Ephesians in Greek thinks of this book first when it comes to succinct style. To take only one measure, A. van Roon has counted more ‘exceeding long sentences’ (of 14 or more lines) in Ephesians than in any other Pauline epistle." The logic here is flawed because the lack of verb in 5:22 makes the sentence longer, not shorter!
Much of McCullum's article is puzzling to me, though I have read it only twice. I am struggling to see why it matters whether 5:21 connects with 5:20. Can't it connect both with 5:20 and with 5:22?
Nor am I seeing tension between 5:21 and a verbless 5:22. The author is urging mutual submission and highlighting the submission of women to their husbands as an example the he feels is important.
I don't understand why McCullum discusses whether the text "is interrupted by anacoluthon before 5.22". If there was no verb in 5:22, then the implied verb belongs within 5:22, not before it. What am I missing?
McCullum writes, "The idea of submission was already introduced through ὑποτασσόμενοι in 5.21, so it is a known and familiar part of the discourse. Because an imperative conveying the same idea corresponds to presupposed information, it is nonvital and unlikely to be put in a prominent place in the sentence." It is precisely because the verb is nonvital in 5:22 that it may not have been there at all originally.
Where is the weighty argument in McCullum's article? I am not seeing it.
Richard, at the base of our difference is that you seem more confident in your ability to detect the motives of scribes from a millennium and a half ago than I do. Where a simple mechanical explanation works, I think we should prefer it over an intentional explanation. Is the mistake I'm suggesting, given the letters involved, common? Probably not. But I'm not suggesting it was only that it did happen here as it also seems to have happened in P46 in 1 Cor 8.12 where the scribe's eye skipped from συνείδησιν to ἀσθενοῦσαν, omitting the latter. The same could explain the shorter reading in Clement too.
DeleteYes, mechanical errors are the most common. However, this passage concerns the status of women, and such passages often suffer corruptions that move them in a patriarchal direction. Nearly all the NT texts that list women before men have an early textual variant that denies this status to the female(s). See my 2022 CBQ article. Also, whenever someone is defined by their relationship to a female relative, we find variants that conform the text to patriarchal norms. For example, in Mark 6:22 someone changed Herodias's daughter to "his daughter", which makes no sense. See my recent TC article here: https://jbtc.org/v28/TC-2023_Fellows.pdf. All this needs to be taken into account when considering Eph 5:22, but neither you nor McCollum have done so. Your preference for mechanical slips, while often applicable, is misplaced in this case.
DeleteP46 does NOT have a strong tendency to make eye skips. Royse (page 906) shows that P47, P72, and P45 have eye skips more frequently than P46. Only P66 and P75 (of the mss surveyed) have a lower frequency than P46. We should also note that additions had a tendency to spread through mixture, since scribes would include a word when it was in one exemplar but not in another.
Incidentally, we should probably not blame the copyists for the addition of the verb to Eph 5:22. A reader of a manuscript could have inserted the verb, perhaps because he was familiar with it because it had been supplied (of necessity) whenever the verse had been quoted without 5:21. The frequent quotation of 5:22 (on its own) in the patriarchal second or third century would explain how the verb was added.
A study of the issue should include a list of all the sentences in the NT where the verb is implied, and an assessment of how often we find variants that supply the verb.
Having read McCollum's article a third time, I think I can discern two (and only two) arguments that he uses for the presence of the verb in 5:22. Firstly, he argues that 5:21 is part of the same sentence as 5:20, and he assumes that 5:21 cannot also be part of the same sentence as 5:22. He then concludes that 5:22 must have its own verb. But we know that the author of Ephesians liked long sentences, so 5:20–22 can all be part of the same sentence. Later transmitters of the text conformed it to more normal style by adding a verb to 5:22, thus splitting the long sentence.
ReplyDeleteSecondly, he thinks that the verb in 5:22 cannot be implied from 5:21 (rather than written) because in 5:21 it is masculine, whereas the subject in 5:22 is feminine. But when the reader imports the idea of submission from 5:21 to 5:22, he/she is not compelled to also import the number or gender. In French we can say "Les personnes sont allées: les hommes à la guerre", and I don't think anyone will object that "allées" is feminine but "les hommes" are masculine. McCollum needs to do more to show that there is a problem here, otherwise he has no case.
I suppose it’s no surprise that I’m on board with McCollum’s close grammatical analysis, and I emphasize that that is a careful balancing of the probabilities and not a categorical judgment of what can and “cannot” work. The affirmative reasons for taking v.21 with v.20 are more than what’s stated here: (a) it continues participial elaborations beginning in v.19 of the main clause in v.18, and (b) the resulting asyndeton between v.21 and v.22 makes v.22 a fitting beginning of a new section in the text, which the household code is.
DeleteOn the other hand, joining v.21 to v.22 creates a number of pragmatic difficulties (lack of gender concord implying a topic shift appropriate for an independent assertion that’s missing a verb, etc.) Adding v.20 or even v.19 to vv.21-22 does not address these issues. They still remain. The French example is not apposite for the Greek: the French has two finite clause each with its own gender concord, whereas the Greek is involving a participial clause that should ideally fit the expressed subject of the matrix clause in case, number, and gender.
Thanks for commenting, Stephen. I am surprised that agreement in gender would be required (or even considered ideal), given that a) the verb is absent (from 5:22), b) the women are already part of the implied subjects of the verb in 5:21, c) among the parallels listed by Aubrey, Rom 11:18; 1 Cor 9:25; Gal 6:14 demonstrate that agreement in person was not required when the verb is omitted. If there are more exact parallels (e.g. with participles), please supply them.
DeleteWhat evidence is there that the gender mismatch would have been considered non-ideal?
Richard, thank you for the time and thought you've given to my paper! Having read it three times, you've already been extremely generous with your attention, and I hope that my response will adequately engage with your questions and comments. Australian Labour Day has given me enough free time to consider the points you've raised and offer a response, but my continued work on my PhD will probably delay any further responses after today.
DeleteThankfully, Stephen has already covered much of what I would have said (and probably more eloquently than I would have said it). Intrinsic probabilities should be regarded in probabilistic terms. All of the variant readings in this passage are possible, but the question is which readings the author was more likely to have written than others in light of which structure for his discourse best communicates his argument.
The reading without a verb in 5:22 is possible, but I consider it less likely intrinsically than the inclusion of the verb for the affirmative reasons Stephen outlined above. The household code, by virtue of its length, surely must constitute its own section of the epistle; the only question is where the section begins. The section break is marked clearly and effectively if the household code begins with a new sentence in 5:22. This sentence would begin with asyndeton, which often marks the start of the new section, and either a shift in audience marked by the third-person imperative or a vocative address with a second-person imperative. The sequence of participles in 5:19–21, for their part, would form the conclusion to the previous section of general injunctions to the whole congregation, and v. 21 in particular would serve as a "hinge verse" that wraps up the previous sequence while setting up (if somewhat broadly) the theme of the household code in terms of its basic rule and motiviation: "submitting to one another in fear of Christ." If a verb for "submit" is included in v. 22, then this function of v. 21 is further emphasized through the hook word "submit." This, incidentally, would explain why the author would repeat "submit" for the first group he addresses in the household code, while he uses other verbs for different groups.
The structure is obscured in several ways if 5:22 continues the sentence containing 5:21. If 5:21 is read as continuing the sequence of participles in 5:19–20, then there is no section break between the household code and what precedes it. The more preferable alternative is to read 5:21 as the start of a new sentence, and thus the start of the household code. But then the author's introductory statement of the household code reads "Submitting to one another in fear of Christ, the wives to their own husbands, as to the lord." The common masculine gender of ὑποτασσόμενοι and the reciprocal pronoun ἀλλήλοις both set up the expectation that the author will continue, "and husbands to their own wives"—but this is precisely what the author does not do. He instead uses the injunction directed to wives as a point of departure for his elaboration on why wives should submit to their husbands! Even if we grant the author of Ephesians one of his characteristically lengthy digressions, he resumes his original thought in 5:25, and while we would expect him to do so with οἱ ἄνδρες ταῖς γυναιξίν, οἱ ἄνδρες ὑποτασσόμενοι ταῖς γυναιξίν, or οἱ ἄνδρες ὑποτάσσεσθε ταῖς γυναιξίν, he instead writes οἱ ἄνδρες, ἀγαπᾶτε τὰς γυναῖκας (the only textual variation being with regard to the inclusion and placement of ἑαυτῶν or ὑμῶν). In short, if 5:21 marks the start of the household code, then the author's continuation in 5:22 inexplicably subverts the expectation he sets up in the earlier verse.
An instructive point of comparison (and one that I think might serve as one of the "more exact parallels" you requested from Stephen, though I welcome additional or better examples from him) comes from 1 Peter. In 1 Pet 2:13, the author begins with the general imperative ὑποτάγητε, and he particularizes this in long dependent participial clauses in 1 Pet 2:18 (οἱ οἰκέται ὑποτασσόμενοι), 1 Pet 3:1 (ὁμοίως [αἱ] γυναῖκες ὑποτασσόμεναι), and 1 Pet 3:7 (οἱ ἄνδρες ὁμοίως συνοικοῦντες). The participle is explicitly supplied for each group and in the gender matching the noun it modifies, even when the same verb has been supplied earlier (as ὑποτάγητε or ὑποτασσόμενοι). In this passage, as in the Ephesians household code, different verbs are used in the addresses to wives and husbands, even when "submission" is the overarching theme of the section.
DeleteI hope this clarifies some of the arguments I make regarding the intrinsic probabilities of readings in Eph 5:22. Since the focus of my paper is intrinsic probabilities and not transcriptional probabilities or external evidence, I don't devote much discussion to the latter two types of evidence for reasons of space, and for the same reasons, I won't do so here. For now, I'm afraid it will have to suffice for me to say that in my PhD thesis in progress, I do discuss these other classes of evidence. In particular, I regard the clarifying addition of the third-person imperative as a legitimate possibility alongside the clarifying addition of the second-person imperative, and I take note of your remarks on Clement of Alexandria as a potential historical precedent for this change.
Thanks, Joey, for the patient and respectful engagement and clarification.
DeleteSo, the argument is based on the assertion that the author would have wanted to mark the beginning of the household codes as a new section. I agree with you that the author used 5:21 to link to both what came before and what came after. You call it a "hinge verse". I think it shows that the author did not want there to be a break in thought between 5:21 and 5:22. He included 5:21 to create a smooth flow of thought between 5:20 and 5:22. Also, the author's style was to link thoughts together, creating long sentences. Therefore, I think the author would have wanted to avoid a break between 5:21 and 5:22. The asyndeton that occurs when the verb is added to 5:22 creates such an unwanted beak, as does the shift in audience marked by the third person imperative. In contrast, if there is no verb in 5:22, then 5:22 is joined to 5:21, thus reinforcing the "hinge" that we know the author wanted.
Thus, your argument is reversed, or a least neutralized, isn't it?
Yes, 1 Peter 2:18; 3:1; 3:7 include participles. The author had to include them because the imperative in 2:13 was too distant to serve quadruple duty. So this cannot be used as a possible example of a New Testament writer avoiding the omission of a verb because of a mismatch of gender, person, or number. So, my question remains: what is the evidence that the gender of the participle in 5:21 would have been mentally imported, along with the idea of submission, into the gap in 5:22, creating a gender disagreement?
Delete