Monday, July 09, 2018

John DelHousaye: Ephesians 1:1 and the Most Elegant Readings

It’s my pleasure to introduce a guest post from my New Testament colleague, John DelHousaye. John is Associate Professor of New Testament at Phoenix Seminary where he has taught for over 15 years. Many of those included Greek exegesis courses based on Ephesians. Now, he has distilled the fruit of that labor into a new guidebook for students. Since John took special note of textual variants in his book, I asked if he would share some of that with us. (By the way, you can enter to win a copy of the book here.)

In Engaging Ephesians: An Intermediate Reader and Exegetical Guide (GlossaHouse, 2018), I had the opportunity to review many of the recorded variants from our witnesses to the letter. Educated with a bias against the Byzantine tradition, I wanted to be more sympathetic, especially because of the softening to the readings in the Nestle-Aland 28th edition, but concluded they comprised glosses and expansions. In the end, I did not adopt a single uniquely Byzantine reading. Ephesians is also attested in the papyri:

Eph 1:1–23 P46
Eph 1:11–13, 19–21 P92
Eph 2:1–7, 10–22 P46
Eph 3:1–21 P46
Eph 4:1–32 P46
Eph 4:16–29, 31–32 P49
Eph 5:1–6, 8–33 P46
Eph 5:1–13 P49
Eph 6:1–6, 8–18, 20–24 P46

I was not especially impressed by these witnesses: the transcription is often sloppy (see, for example, the P46 reading at 4:30). Also problematic were the variants in Codex Claromontanus (see 3:1). Variants in Claromontanus were also derivative (see 3:1). In my opinion, Codex Sinaiticus, our oldest complete copy of the New Testament, is the least derivative. I may be wrong, but I do not judge that the evidence is on the side of those seeking another point of departure. Of course, Sinaiticus seems to be filled with derivative readings and transcription errors.

My bias was confirmed, but then immediately challenged in the opening line. Sinaiticus reads:
Παῦλος ἀπόστολος ΙΥ ΧΥ διὰ θελήματος θεοῦ τοῖς ἁγίοις τοῖς οὖσιν καὶ πιστοῖς ἐν Χριστῷ Ἰησοῦ
The Nestle-Aland 28 reads:
Παῦλος ἀπόστολος Χριστοῦ Ἰησοῦ διὰ θελήματος θεοῦ τοῖς ἁγίοις τοῖς οὖσιν [ἐν Ἐφέσῳ] καὶ πιστοῖς ἐν Χριστῷ Ἰησοῦ·
The edited text dispenses with the nomina sacra and places Χριστοῦ before Ἰησοῦ; the preposition phrase ἐν Ἐφέσῳ is included with brackets.

Concerning the first variant, variation in the order of Jesus’s name and primary epithet, the Christ (Messiah), is very common in the manuscripts. We also find both orders in the immediate context without variants. On the one hand, Sinaiticus and most of the Byzantine tradition read Ἰησοῦ χριστοῦ. On the other, most printed editions (Nestle-Aland, SBLGNT) follow the earlier reading in P46 (c. 175–225), which is also reflected in Vaticanus, Claromontanus, and other witnesses. I favor this reading because it is earlier and (slightly) more difficult—“Jesus Christ” being the default. The order also fits a couplet pattern:
  • Χριστοῦ Ἰησοῦ (1:1)
    Χριστῷ Ἰησοῦ (1:1)
  • Ἰησοῦ χριστοῦ (1:2)
    Ἰησοῦ χριστοῦ (1:3)
We also find the inverse order without recorded variants in the final greeting (Ἰησοῦν Χριστὸν, 6:24), which is significant because Ephesians appears to be structured as a chiasm.

A         Opening Greeting (1:12)
            1) The faithful (1:1)
            2) Prayer for “grace” and “peace”
            3) Directed to “God our Father and the Lord Jesus Christ” (1:2)
            B         Prayer on “mystery” of God’s will and spiritual warfare (1:323)
                        C         Christ’s Family: Gentiles and Jews (2:13:21)
                                    D         Walking with God (4:15:21)
                                                            1. “Walk” (4:1, 17; 5:2, 8, 15)
                                                            2. Trinity (4:16 // 5:1820)
                        C'        Christ’s Family: Spouses, Children, Slaves (5:216:9)
            B'        Spiritual warfare and prayer on “mystery” of the gospel (6:1020)
A'        6:2123:          Closing Greeting
            1') The faithful Tychicus (6:21)
            2') Prayer for “peace” and “grace” (6:2324, inverse order)
            3') Directed to “God our Father and the Lord Jesus Christ” (6:23)

Therefore, I judge the reading in Sinaiticus to be derivative and not a contender for the earliest reading in our manuscript tradition. In any case, the sense is little affected.

But the second variant, [ἐν Ἐφέσῳ], is more meaningful and difficult to resolve. Our earliest manuscripts (P46, the original hand of Sinaiticus, Vaticanus, Origen) do not mention the city. ἐν Ἐφέσῳ is written by a different hand in a different color of ink in the margin of Sinaiticus; the words are also in the margin of Vaticanus.[2] We may interpret these facts at least two ways. The scribes noted an omission or made an emendation.

On the one hand, some argue the letter was intended to be an encyclical.[3] The content of the letter is more general, suggesting a broader application, perhaps with Tychicus serving as its carrier and reader throughout Asia Minor (6:21–22). Perhaps a scribe (or pseudonymous author) wanted to harmonize the letter with Colossians, which is closely related in style and substance and reads ἐν Κολοσσαῖς (1:1). If we omit τοῖς οὖσιν ἐν Ἐφέσῳ, we have a seemingly more harmonious τοῖς ἁγίοις καὶ πιστοῖς (see Col 1:2).

On the other, the first copyist of Sinaiticus, or more likely because of the absence of ἐν Ἐφέσῳ in the earlier P46 someone else, might have omitted the phrase by accident or in order to generalize the letter. A clearer example of this pattern is the omission of τοῖς ἐν Ῥώμῃ at Rom 1:14 in some later witnesses. Do these readings presuppose an early catholicity to the Pauline letters?

The use of οὖσιν without a predicate is anomalous (see Rom 1:7; Phil 1:1; 1 Cor 1:2; 2 Cor 1:1). Origen, a native speaker, struggled to make sense of it.[4] Also, as William Larkin notes, all extant manuscripts have the superscript or heading ПРОΣΕΠΕΣΙΟΥΣ.[5] The association with the city is not arbitrary. Our witnesses do not provide an alternative location.[6]

The wording and arrangement also suggest a parallelism, conveying a dual citizenship:
τοῖς     ἁγίοις τοῖς οὖσιν      ἐν Ἐφέσῳ
καὶ      πιστοῖς                       ἐν Χριστῷ Ἰησοῦ
Although the reading ἐν Ἐφέσῳ is uncertain, I find no compelling reason to drop it.

My study of Ephesians led to a deeper appreciation of the letter’s elegance. Overall, Codex Sinaiticus is the best witness to this rhetorical beauty. In some cases, along with other considerations, the more elegant reading should be preferred. Of course, the original copyist of Sinaiticus inherited a flawed text and added further infelicities. Nevertheless, in my opinion, this codex remains the best point of departure. 


  1. Adapted from see John Paul Heil, Ephesians: Empowerment to Walk in Love for the Unity of All in Christ (Atlanta: Society of Biblical Literature, 2007), 38-42.
  2. Eadie, Ephesians, xviii.
  3. John Eadie attributes the theory to Usher (A Commentary on the Greek Text of The Epistle of Paul to The Ephesians, 2nd ed. [New York: Robert Carter and Brothers, 1861], xxiv).
  4. Eadie, Commentary, xix.
  5. William J. Larkin, Ephesians: A Handbook on The Greek Text (Waco, TX: Baylor University Press, 2009), 2.
  6. Marcion may have claimed the letter was intended for the Laodiceans, but no extant witness supports this.


  1. Hiya!

    The phrase τοῖς ἁγίοις τοῖς οὔσιν καὶ πιστοῖς can also be treated as an unusual TSKS construction, can it not?


    1. Hi, Jake,

      Yes. See also Eph 2:20; 3:5; and 4:11.

    2. Under Granville Sharp's own limitations, TSKS is not claimed to apply with plurals (although I would not rule such out in all instances).

    3. A fair enough qualification and not a little debate!

    4. Hi All. Here are some thoughts from my commentary writing on Ephesians on "in Ephesus." It has been stated that the phrase τοῖς ἁγίοις τοῖς οὔσιν καὶ πιστοῖς without ἐν Ἐφέσῳ is ungrammatical, or leaning that way. Yet, it wasn't thought ungrammatical by a good number of scribes who often correct these things. Alternatively, there are numerous instances where the verb εἰμί “to be” is found in the same grammatical construction (article + participle) meaning “which/who are…” or “which is…” before a modifier (see, e.g., Gen 19:11; 41:31; 49:32; Lev 18:27; Deut 3:25; 16:14 [LXX]; 2 Cor 11:31). Thus, here the adjective πιστοῖς “faithful” is stressed by the ascensive καί “also” and can be translated “also faithful in Christ Jesus.” Why is "faithful" stressed? Well, there is an important social intertextural dimension of this epistle in its unique description of the letter recipients as “faithful” (πιστός). The descriptor “faithful” (πιστός) had honorific and titular significance in inscriptions and was increasingly used in this official sense to describe Christian workers (1 Cor 4:17; 7:25; Eph 6:21; Col 1:7; 4:7, 9; 1 Pet 5:12). Also, faithfulness is a philosophical topos in the Discourses of Epictetus (late first-century); the adjective “faithful” (πιστός) and its noun “faithfulness” (πίστις) are found prominently as virtues of the Stoic contemplative life (I have a table listing these). The nature of Epictetus’ “faithfulness” includes one’s God-given self-integrity, one’s fidelity in relationship with spouse and neighbor, and the imitation of the faithfulness of the deity. In Ephesians, the topos of faith(fullness) is directed to Jesus as Messiah and Lord (1:1, 15; 3:17; 4:5, 13; cf. 6:21). So, I conclude for these and other reasons (including manuscript external reasons) that "in Ephesus" is not original. Instead, our missive "Ephesians" is a circular letter for believers in Asia Minor occasioned by Paul's arrest in Jerusalem and Roman escort to Caesarea Maritima where he stayed and wrote it and Colossians and Philemon.

    5. I am not exactly convinced about "it wasn't thought ungrammatical by a good number of scribes who often correct these things".

      Given that (a) the "good number" amounts at most to only 6 out of 604 MSS (per Text und Textwert), with at least the original hand of two of these (Aleph/B) apparently reflecting an earlier common archetype, and (b) that MSS from all other textual types and clusters retain "Ephesus" in one form or another, along with the Latin, Syriac, and Coptic versions...

      It would seem that there should come a point -- even for eclectic practitioners -- when the external evidence should be decisive as opposed to any suggested internal possibilities. To me this appears to be one such case.

  2. Thanks for this post. Eph 1:1 was one of the first variants found in the margins of my NIV Student bible that got me first interested in textual criticism all the way back in highschool. (That, and Whether Jesus wrote "the sins of them" in the dirt with his finger).

    I remember thinking on the one hand, plenty of letters with individual addressees were nevertheless circulated generally, so it didn't make sense to me that, seeing an original "in Ephesus" someone would remove that just to circulate it generally.

    On the other hand, it made even less sense to me that, seeing an original txt with the omission, someone would have to add "in Ephesus" just to send it to those in Ephesus - since a text with no individual addressee is already suited to general circulation to places like Ephesus.

    Since those days, I think I still lean towards the second position, and additional arguments like the hanging ousin or the Rom 1:14 comparison do, I think, only strengthen the case.

    1. Ryan,

      Are you saying that the short reading in Rom 1:14 evidences a Tendez among early scribes to fill out addresses? I'd seen the same verse cited as suggesting a Tendez to universalize epistles. But then again I probably shouldn't be surprised: I'm finding that very many data in both OTTC and NTTC are interpreted in exactly opposite ways by different critics.

      (To be clear, I'm not trying to be critical of your position, only to make sure I understand it and secondarily to reflect a bit on it.)

    2. Ryan, thanks for the observations. They seem sound to me. If I had to choose--and I intentionally employed "might" and the rhetorical question in my post because who knows?--a scenario, it would be that in both Eph 1:1 and Rom 1:14 the omission of the city was accidental. The omissions in some manuscripts are the pattern on the page that does not require entering the mind of a scribe. It was something that happened for whatever reason.

  3. Fun stuff!

    Two questions:

    First, by taking 01 as your "point of departure," do you mean that you take it as your base text and depart from it only where very strong evidence to the contrary exists? (I myself had actually considered that approach at one point; in the end I decided against it, but that's just me.)

    Second, by "elegant reading," do you mean a reading that comports so well with intrinsic evidence that it cannot be dismissed even though it is easier than another reading?



    1. Hi, Stephen!

      Thanks for your questions. Yes, with many friends I want to produce a Greek New Testament based on Sinaiticus (like Leningradensis for the Old Testament). Why not do this with the earliest complete copy of the New Testament we have? We would then not be appealing to a hypothetical text but an actual one. For conspiracy theorists, anyone could review the text online. I would make suggested corrections in footnotes.

      I have great respect for the CBGM, but it will still need to be critiqued by intrinsic-rhetorical considerations. I intend the criterion of the more elegant reading to stand in dynamic tension with the more difficult reading. I see great care in the structure of most of the NT writings, from the macro chiasms of entire books, to micro parallelisms and so on. But I rarely see artistry from the hands of copyists!

  4. JDH,
    << In the end, I did not adopt a single uniquely Byzantine reading. >>

    Byz and NA only disagree in Ephesians chapter 1 eleven times. How many "uniquely" Byzantine readings do you say there are in that chapter?

  5. I'm thinking of a blank space solution.

  6. Meyer still makes sense, I do think, on Rom. 1,7 and 15:

  7. It might be useful to consult the The Amsterdam Database of New Testament Conjectural Emendation ( for this verse ( 1:1). The “spatium” option is cj14590.

    1. There is something mildly entertaining about textual critics conjecturing the total absence of anything.

    2. Yes. Although I think we could characterise it as a format conjecture rather than a textual conjecture.

    3. Thanks Jan, that is a very respectable crowd in support. This looks like it might be one of the most popular conjectures in the whole database.