² tc The witnesses for the shorter reading (in which the verb “submit” is only implied) are minimal (Ì⁴⁶ B Cl Hier^mss), but significant and early. The rest of the witnesses add one of two verb forms as required by the sense of the passage (picking up the verb from v. 21). Several of these witnesses have ὑποτασσέσθωσαν (Jupotassesqwsan), the third person imperative (so א A I P Ψ 0278 33 81 1175 1739 1881 al lat co), while other witnesses, especially the later Byzantine cursives, read ὑποτάσσεσθε (Jupotassesqe), the second person imperative (D F G Ï sy). The text virtually begs for one of these two verb forms, but the often cryptic style of Paul’s letters argues for the shorter reading. The chronology of development seems to have been no verb – third person imperative – second person imperative. It is not insignificant that early lectionaries began a new day’s reading with v. 22; these most likely caused copyists to add the verb at this juncture.

² tc The verb λατρεύω (latreuw; here the participial form, λατρεύοντες [latreuonte"]) either takes a dative direct object or no object at all, bearing virtually a technical nuance of “worshiping God” (see BDAG 587 s.v.). In this text, πνεύματι (pneumati) takes an instrumental force (“by the Spirit”) rather than functioning as object of λατρεύοντες. However, the word after πνεύματι is in question, no doubt because of the collocation with λατρεύοντες. Most witnesses, including some of the earliest and best representatives of the Alexandrian, Western, and Byzantine texts (א* A B C D² F G 0278^vid 33 1739 1881 Ï co Ambr), read θεοῦ (qeou; thus, “worship by the Spirit of God”). But several other important witnesses (א² D* P Ψ 075 365 1175 lat sy Chr) have the dative θεῷ (qew) here (“worship God by the Spirit”). Ì⁴⁶ is virtually alone in its omission of the divine name, probably due to an unintentional oversight. The dative θεῷ was most likely a scribal emendation intended to give the participle its proper object, and thus avoid confusion about the force of πνεύματι. Although the Church came to embrace the full deity of the Spirit, the NT does not seem to speak of worshiping the Spirit explicitly. The reading θεῷ thus appears to be a clarifying reading. On external and internal grounds, then, θεοῦ is the preferred reading.

⁴ tc ‡ Judging by the superior witnesses for the first person pronoun ἡμῶν (Jhmwn, “us”; Ì⁴⁶ א* A B D* F G 326* 1505 al) vs. the second person pronoun ὑμῶν (Jumwn, “you”; found in א² C D¹ Ψ 075 33 1739 1881 Ï lat sy co), ἡμῶν should be regarded as original. Although it is possible that ἡμῶν was an early alteration of ὑμῶν (either unintentionally, as dittography, since it comes seventeen letters after the previous ἡμῶν; or intentionally, to conform to the surrounding first person pronouns), this supposition is difficult to maintain in light of the varied and valuable witnesses for this reading. Further, the second person is both embedded in the verb ἐμάθετε (emaqete) and is explicit in v. 8 (ὑμῶν). Hence, the motivation to change to the first person pronoun is counterbalanced by such evidence. The second person pronoun may have been introduced unintentionally via homoioarcton with the ὑπέρ (Juper) that immediately precedes it. As well, the second person reading is somewhat harder for it seems to address Epaphras’ role only in relation to Paul and his colleagues, rather than in relation to the Colossians. Nevertheless, the decision must be based ultimately on external evidence (because the internal evidence can be variously interpreted), and this strongly supports ἡμῶν.

¹ tc Some of the better representatives of the Alexandrian and Western texts have a passive verb here instead of the active ἀποκατήλλαξεν (apokathllaxen, “he has reconciled”): ἀποκατηλλάγητε (apokathllaghte) in (Ì⁴⁶) B, ἀποκατήλλακται [sic] (apokathllaktai) in 33, and ἀποκαταλλαγέντες (apokatallagente") in D* F G. Yet the active verb is strongly supported by א A C D² Ψ 048 075 [0278] 1739 1881 Ï lat sy. Internally, the passive creates an anacoluthon in that it looks back to the accusative ὑμᾶς (Juma", “you”) of v. 21 and leaves the following παραστῆσαι (parasthsai) dangling (“you were reconciled…to present you”). The passive reading is certainly the harder reading. As such, it may well explain the rise of the other readings. At the same time, it is possible that the passive was produced by scribes who wanted some symmetry between the ποτε (pote, “at one time”) of v. 21 and the νυνὶ δέ (nuni de, “but now”) of v. 22: Since a passive periphrastic participle is used in v. 21, there may have a temptation to produce a corresponding passive form in v. 22, handling the ὑμᾶς of v. 21 by way of constructio ad sensum. Since παραστῆσαι occurs ten words later, it may not have been considered in this scribal modification. Further, the Western reading (ἀποκαταλλαγέντες) hardly seems to have arisen from ἀποκατηλλάγητε (contra TCGNT 555). As difficult as this decision is, the preferred reading is the active form because it is superior externally and seems to explain the rise of all forms of the passive readings.

¹ tc The words ἐπὶ τοὺς υἱοὺς τῆς ἀπειθείας (epi tou" Juiou" th" apeiqeia", “on the sons of disobedience”) are lacking in Ì⁴⁶ B b sa, but are found in א A C D F G H I Ψ 075 0278 33 1739 1881 Ï lat sy bo. The words are omitted by several English translations (NASB, NIV, ESV, TNIV). This textual problem is quite difficult to resolve. On the one hand, the parallel account in Eph 5:6 has these words, thus providing scribes a motive for adding them here. On the other hand, the reading without the words may be too hard: The ἐν οἷς (en |oi") of v. 7 seems to have no antecedent without υἱούς already in the text, although it could possibly be construed as neuter referring to the vice list in v. 5. Further, although the witness of B is especially important, there are other places in which B and Ì⁴⁶ share errant readings of omission. Nevertheless, the strength of the internal evidence against the longer reading is at least sufficient to cause doubt here. The decision to retain the words in the text is less than certain.

³ tc ‡ Most witnesses (A D Ψ 048 1739 1881 Ï sy) have κυρίου (kuriou, “Lord”) instead of θεοῦ (qeou, “God”) here, while a few have Χριστοῦ (Cristou, “Christ”; 206 {429 1758}). θεοῦ, however, is well supported by א C F G I 614 629 630 1175 al. Internally, the Pastorals never elsewhere use the expression ἐνώπιον κυρίου (enwpion kuriou, “before the Lord”), but consistently use ἐνώπιον θεοῦ (“before God”; cf. 1 Tim 2:3; 5:4, 21; 6:13; 2 Tim 4:1). But this fact could be argued both ways: The author’s style may be in view, or scribes may have adjusted the wording to conform it to the Pastorals’ universal expression. Further, only twice in the NT (Jas 4:10 [v.l. θεοῦ]; Rev 11:4 [v.l. θεοῦ]) is the expression ἐνώπιον κυρίου found. That such an expression is not found in the corpus Paulinum seems to be sufficient impetus for scribes to change the wording here. Thus, although the external evidence is somewhat on the side of θεοῦ, the internal evidence is on the side of κυρίου. A decision is difficult, but κυρίου is the preferred reading.

¹ tc πάντες (pantes, nominative plural in “you all know”) is read by א B P Ψ sa. A C 049 33 1739 Ï latt sy bo have the accusative πάντα (panta, “you know all things”). The evidence favors the nominative reading, but it is not overwhelming. At the same time, the internal evidence supports the nominative for a variety of reasons. A scribe would naturally tend to give the transitive verb a direct object, especially because of the parallel in the first half of the verse. And intrinsically, the argument seems to be in balance with v. 19: The “all” who have gone out and are not “in the know” with the “all” who have an anointing and know that they are true believers. Further, as R. E. Brown points out, “the fact of their knowledge (pantes), not the extent of its object (panta), seems best to fit the reassurance” (Epistles of John [AB], 349). Brown further points out the connection with the new covenant in Jer 31 with this section of 1 John, esp. Jer 31:34 – “they all [pantes] shall know me.” Since 1 John alludes to Jer 31, without directly quoting it, this is all the more reason to see the nominative as original: Allusions are often overlooked by scribes (transcriptional evidence), but support the intrinsic evidence. Thus, the evidence is solidly (though not overwhelmingly) behind the nominative reading.

¹ tc ‡ Some mss read “taken off” (ἐκδυσάμενοι, ekdusamenoi) instead of “put on” (ἐνδυσάμενοι, endusamenoi). This alternative reading would change the emphasis of the verse from putting on “our heavenly house” to taking off “our earthly house” (see the following note regarding the specification of the referent). The difference between the two readings is one letter (ν or κ), either of which may be mistaken for the other especially when written in uncial script. ἐνδυσάμενοι enjoys strong support from the Alexandrian text (Ì⁴⁶ א B C 33 1739 1881), Byzantine witnesses, versions (lat sy co), and Clement of Alexandria. The Western text is the only texttype to differ: D*^,c reads ἐκδυσάμενοι, as does ar f^c Mcion Tert Spec; F and G read εκλ for εκδ which indirectly aligns them with D (and was surely due to confusion of letters in uncial script). Thus “put on” has the oldest and best external attestation by far. Internal evidence also favors this reading. At first glance, it may seem that “after we have put on our heavenly house we will not be found naked” is an obvious statement; the scribe of D may have thought so and changed the participle. But v. 3 seems parenthetical (so A. Plummer, Second Corinthians [ICC], 147), and the idea that “we do not want to be unclothed but clothed” is repeated in v. 4 with an explanatory “for.” This concept also shows up in v. 2 with the phrase “we desire to put on.” So the context can be construed to argue for “put on” as the original reading. B. M. Metzger argues against the reading of NA²⁷, stating that ἐκδυσάμενοι is “an early alteration to avoid apparent tautology” (TCGNT 511; so also Plummer, 148). In addition, the reading ἐνδυσάμενοι fits the Pauline pattern of equivalence between apodosis and protasis that is found often enough in his conditional clauses. Thus, “put on” has the mark of authenticity and should be considered original.

² tc ‡ The unusual order καὶ κυρίου ἡμῶν (kai kuriou Jhmwn), which produces the reading “our Lord Jesus Christ” instead of “God our Father,” is read by Ì^46,51vid B D F G H 1739 1881 Ï sy sa, while the more normal ἡμῶν καὶ κυρίου (Jhmwn kai kuriou) is found in א A P Ψ 33 81 326 365 2464 pc. Thus, the reading adopted in the translation is more widespread geographically and is found in the two earliest witnesses, along with several good representatives of the Alexandrian, Western, and Byzantine texttypes. Internally, there would be a strong motivation for scribes to change the order: “from God our Father and the Lord Jesus Christ” is Paul’s normal greeting; here alone is the pronoun attached to “Jesus Christ” (except in the pastorals, though the greeting in these letters is nevertheless unlike the rest of the corpus Paulinum). Intrinsically, the chosen reading is superior as well: Scribes would be prone to emulate Paul’s regular style, while in an early letter such as this one his regular style was yet to be established (for a similar situation, cf. the text-critical discussion at 1 Thess 1:1). Hence, there is a strong probability that the reading in the translation is authentic. Although B. M. Metzger argues that “the apostle’s stereotyped formula was altered by copyists who, apparently in the interest of Christian piety, transferred the possessive pronoun so it would be more closely associated with ‘Lord Jesus Christ’” (TCGNT 520), one might expect to see the same alterations in other Pauline letters. That this is not the case argues for “our Lord Jesus Christ” as the authentic reading here.

² tc Most witnesses, including some important ones (א² A Ψ [33] 1739 1881 Ï ar vg sy bo), read “holy” before “brothers [and sisters]” (ἁγίοις ἀδελφοῖς, Jagioi" adelfoi"). It is possible that ἁγίοις dropped out by way of homoioteleuton (in uncial script the words would be written agioisadelfois), but it is equally possible that the adjective was added because of the influence of ἁγίῳ (Jagiw) in v. 26. Another internal consideration is that the expression ἅγιοι ἀδελφοί ({agioi adelfoi, “holy brothers”) is not found elsewhere in the corpus Paulinum, though Col 1:2 comes close. But this fact could be argued either way: It may suggest that such an expression is not Pauline; on the other hand, the unusualness of the expression could have resulted in an alteration by some scribes. At the same time, since 1 Thessalonians is one of the earliest of Paul’s letters, and written well before he addresses Christians as saints (ἅγιοι) in 1 Corinthians for the first time, one might argue that Paul’s own forms of expression were going through something of a metamorphosis. Scribes insensitive to this fact could well impute later Pauline collocations onto his earlier letters. The internal evidence seems to support, albeit slightly, the omission of ἁγίοις here. Externally, most of the better witnesses of the Alexandrian and Western texts (א* B D F G 0278 it sa) combine in having the shorter reading. Although the rating of “A” in UBS⁴ for the omission seems too generous, this reading is still to be preferred.

⁴ tc The reading σειραῖς (seirai", “chains”) is found in Ì⁷² P Ψ 33 1739 Ï vg sy, while σιροῖς (sirois [or σειροῖς, seirois], “pits”) is found in א A B C 81 pc. The evidence is thus fairly evenly divided. Internally, the reading adopted here (σειραῖς) is a rarer term, perhaps prompting some scribes to replace it with the more common word. However, this more common term is not a synonym and hence does not follow the normal pattern of scribes. As well, the use of the genitive ζόφου (zofou) in “chains of darkness” is a bit awkward (a rare genitive of place), perhaps prompting some scribes to change the imagery to “pits of darkness” (in which case ζόφου is an attributive genitive). A further point that complicates the issue is the relationship of 2 Peter to Jude. Jude’s parallel (v. 6) has δεσμοῖς (desmois, “chains”). Apart from the issue of whether 2 Peter used Jude or Jude used 2 Peter, this parallel suggests one of two possibilities: either (1) since these two books obviously have a literary relationship, σειραῖς is original, or (2) early scribes, recognizing that these two books shared their material, changed σειροῖς to σειραῖς to conform the wording, at least conceptually, to Jude 6. On balance, σειραῖς looks to be original because scribes were not prone to harmonize extensively between books other than the Gospels (although 2 Peter and Jude do display some of this harmonizing). Further, such harmonization is often, if not usually, verbally exact, but δεσμοῖς is not a variant here.

³ tc The reading “set free” (λύσαντι, lusanti) has better ms support (̹⁸ א A C 1611 2050 2329 2351 Ï^A sy) than its rival, λούσαντι (lousanti, “washed”; found in P 1006 1841 1854 2053 2062 Ï^K lat bo). Internally, it seems that the reading “washed” could have arisen in at least one of three ways: (1) as an error of hearing (both “released” and “washed” are pronounced similarly in Greek); (2) an error of sight (both “released” and “washed” look very similar – a difference of only one letter – which could have resulted in a simple error during the copying of a ms); (3) through scribal inability to appreciate that the Hebrew preposition ב can be used with a noun to indicate the price paid for something. Since the author of Revelation is influenced significantly by a Semitic form of Greek (e.g., 13:10), and since the Hebrew preposition “in” (ב) can indicate the price paid for something, and is often translated with the preposition “in” (ἐν, en) in the LXX, the author may have tried to communicate by the use of ἐν the idea of a price paid for something. That is, John was trying to say that Christ delivered us at the price of his own blood. This whole process, however, may have been lost on a later scribe, who being unfamiliar with Hebrew, found the expression “delivered in his blood” too difficult, and noticing the obvious similarities between λύσαντι and λούσαντι, assumed an error and then proceeded to change the text to “washed in his blood” – a thought more tolerable in his mind. Both readings, of course, are true to scripture; the current question is what the author wrote in this verse.

⁴ tc The earliest and best mss (Ì^18vid א A C P 2050 al lat sy co) lack the term “God” (θεοῦ, qeou) between “from” (ἀπό, apo) and “he who is” (ὁ ὤν, Jo wn). Its inclusion, as supported by the bulk of the Byzantine witnesses, is clearly secondary and a scribal attempt to achieve two things: (1) to make explicit the referent in the passage, namely, God, and (2) to smooth out the grammar. The preposition “from” in Greek required a noun in the genitive case. But here in Rev 1:4 the words following the preposition “from” (ἀπό) are in another case, i.e., the nominative. There are two principal ways in which to deal with this grammatical anomaly. First, it could be a mistake arising from someone who just did not know Greek very well, or as a Jew, was heavily influenced by a Semitic form of Greek. Both of these unintentional errors are unlikely here. Commenting on this ExSyn 63 argues: “Either of these is doubtful here because (1) such a flagrant misunderstanding of the rudiments of Greek would almost surely mean that the author could not compose in Greek, yet the Apocalypse itself argues against this; (2) nowhere else does the Seer [i.e., John] use a nom. immediately after a preposition (in fact, he uses ἀπό 32 times with the gen. immediately following).” The passage appears to be an allusion to Exod 3:14 (in the LXX) where God refers to himself as “he who is” (ὁ ὤν), the same wording in Greek as here in Rev 1:4. Thus, it appears that John is wanting to leave the divine name untouched (perhaps to allude to God’s immutability, or as a pointer to the Old Testament as the key to unlocking the meaning of this book), irrespective of what it “looks” like grammatically. The translation has placed the “he who is” in quotation marks to indicate to the reader that the syntactical awkwardness is intentional. (For further comments, see ExSyn 63).

¹ tc Χριστόν (Criston, “Christ”) is attested in the majority of mss, including many important witnesses of the Alexandrian (Ì⁴⁶ 1739 1881) and Western (D F G) texttypes, and other mss and versions (Ψ latt sy co). On the other hand, some of the important Alexandrian witnesses have κύριον (kurion, “Lord”; א B C P 33 104 1175 al). A few mss (A 81 pc) have θεόν (qeon, “God”). The nomina sacra for these readings are quite similar (cMn, kMn, and qMn respectively), so one might be able to account for the different readings by way of confusion. On closer examination, the variants appear to be intentional changes. Alexandrian scribes replaced the highly specific term “Christ” with the less specific terms “Lord” and “God” because in the context it seems to be anachronistic to speak of the exodus generation putting Christ to the test. If the original had been “Lord,” it seems unlikely that a scribe would have willingly created a difficulty by substituting the more specific “Christ.” Moreover, even if not motivated by a tendency to overcorrect, a scribe might be likely to assimilate the word “Christ” to “Lord” in conformity with Deut 6:16 or other passages. The evidence from the early church regarding the reading of this verse is rather compelling in favor of “Christ.” Marcion, a second-century, anti-Jewish heretic, would naturally have opposed any reference to Christ in historical involvement with Israel, because he thought of the Creator God of the OT as inherently evil. In spite of this strong prejudice, though, {Marcion} read a text with “Christ.” Other early church writers attest to the presence of the word “Christ,” including {Clement of Alexandria} and Origen. What is more, the synod of Antioch in a.d. 268 used the reading “Christ” as evidence of the preexistence of Christ when it condemned Paul of Samosata. (See G. Zuntz, The Text of the Epistles, 126-27; TCGNT 494; C. D. Osburn, “The Text of 1 Corinthians 10:9,” New Testament Textual Criticism: Its Significance for Exegesis, 201-11; contra A. Robertson and A. Plummer, First Corinthians [ICC], 205-6.) Since “Christ” is the more difficult reading on all accounts, it is almost certainly original. In addition, “Christ” is consistent with Paul’s style in this passage (cf. 10:4, a text in which {Marcion} also reads “Christ”). This text is also christologically significant, since the reading “Christ” makes an explicit claim to the preexistence of Christ. (The textual critic faces a similar dilemma in Jude 5. In a similar exodus context, some of the more important Alexandrian mss [A B 33 81 pc] and the Vulgate read “Jesus” in place of “Lord.” Two of those mss [A 81] are the same mss that have “Christ” instead of “God” in 1 Cor 10:9. See the tc notes on Jude 5 for more information.) In sum, “Christ” has all the earmarks of authenticity here and should be considered the original reading.

² tc The earliest and most important mss omit “in Ephesus” (Ì⁴⁶ א* B* 6 1739 [Mcion^T,E]), yet the opening line of this epistle makes little sense without the phrase (“to the saints who are and are faithful…”? or perhaps “to the saints who are also faithful,” though with this sense the οὖσιν [ousin] is redundant and the καί [kai] is treated somewhat unnaturally). What is interesting is Marcion’s canon list which speaks of the letter to the Laodiceans among Paul’s authentic epistles. This, coupled with some internal evidence that the writer did not know his audience personally (cf. 1:15; 3:2; absence of personal names throughout), suggests that Ephesians was an encyclical letter, intended for more than one audience. Does this mean that the shorter reading is to be preferred? Yes and no. A plausible scenario is as follows, assuming Pauline authorship (though this is strongly contested today; for arguments on behalf of Pauline authorship, see M. Barth, Ephesians [AB 34], 1:36-50; P. T. O’Brien, Ephesians, 4-47; and H. W. Hoehner, Ephesians, 2-61): Paul sent the letter from Rome, intending it first to go to Ephesus. At the same time, Colossians was dispatched. Going counterclockwise through Asia Minor, this letter would first come to Ephesus, the port of entry, then to Laodicea, then Colossae. Tychicus’ instructions may well have been for each church to “fill in the blank” on the address line. The church at Ephesus would have certainly made the most copies, being Paul’s home base for nearly three years. Hence, most of the surviving copies have “in Ephesus” in v. 1 (so א² A B² D F G Ψ 0278 33 1881 Ï latt sy co). But one might expect a hint of evidence that Laodicea also made a few copies: Both Marcion’s list and Col 4:16 may well imply this. What is to account for the early Alexandrian evidence, then? These mss were perhaps made from a very early copy, one reflecting the blank line before each church filled it in. Although it is of course only speculation (as is necessary in a historical investigation lacking some of the pieces to the puzzle), this scenario accounts for all of the data: (1) “in Ephesus” in most mss; (2) Laodicea in Marcion’s list and Col 4:16; (3) the lack of an addressee in the earliest witnesses; (4) why the earliest witnesses’ reading must be rejected as too hard; and (5) why the author seems not to know the readership. In sum, is “in Ephesus” original? Yes and no. Some address belongs there; ἐν ᾿Εφέσῳ (en Efesw) is the predominant address, but several other churches also received this circular letter as their own. For this reason the phrase has been placed in single brackets in the translation. NA²⁷ also lists the words in brackets, indicating doubt as to their authenticity.