³ tc The variant ἤπιοι (ēpioi, “gentle”) has fair support (א^c A C² D² Ψ^c 0278 33 1241 1739 1881 M), but νήπιοι (nēpioi, “little children”) has significantly stronger backing (P⁶⁵ א* B C* D* F G I Ψ* it bo). It is not insignificant that the earliest Alexandrian and Western witnesses in support of ἤπιοι are actually not Alexandrian or Western; they are the second correctors of Alexandrian and Western mss. Such correctors generally follow a Byzantine Vorlage. The reading νήπιοι is thus superior externally. Further, νήπιοι is much harder in this context, for Paul mixes his metaphors (“we became little children in your midst…Like a nursing mother…”). Thus, the scribes would naturally alter this reading to the softer ἤπιοι (“we became gentle…”). Paul is not known for his consistency of figures, however (cf., e.g., Gal 4:19); hence, the intrinsic evidence points to νήπιοι as autographic. On the other hand, it is possible that νήπιοι was caused by dittography with the preceding -μεν (-men). It is even possible that νήπιοι was caused by an error of hearing right from the beginning: The amanuensis could have heard the apostle incorrectly. But such a supposition cuts both ways; further, Paul would no doubt have corrected the reading in the ms before it was sent out. If so, one would surely have expected both earlier witnesses on the side of ἤπιοι and perhaps a few first correctors to have this reading. The reading “little children” thus stands as most probably original. (For an extended discussion of this problem, see J. A. D. Weima, “‘But We Became Infants Among You’: The Case for NHPIOI in 1 Thess 2.7,” NTS 46 [2000]: 547-64; T. B. Sailors, “Wedding Textual and Rhetorical Criticism to Understand the Text of 1 Thessalonians 2.7,” JSNT 80 [2000]: 81-98.)

¹ tc Some of the better representatives of the Alexandrian and Western groups have a passive verb here instead of the active ἀποκατήλλαξεν (apokatēllaxen, “he has reconciled”): ἀποκατηλλάγητε (apokatēllagēte) in P⁴⁶ B, ἀποκατήλλακται [sic] (apokatēllaktai) in 33, and ἀποκαταλλαγέντες (apokatallagentes) in D* F G. Yet the active verb is strongly supported by א A C D² Ψ 048 075 0278 1175 1505 1739 1881 2464 M al lat sy. Internally, the passive creates an anacoluthon in that it looks back to the accusative ὑμᾶς (humas, “you”) of v. 21 and leaves the following παραστῆσαι (parastēsai) 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 others. 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 been a temptation to produce a corresponding passive form in v. 22, so that the ὑμᾶς of v. 21 functioned as subject 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.

³ tc The witnesses display a bewildering array of variants here. Instead of “the Holy One of God” (ὁ ἅγιος τοῦ θεοῦ, ho hagios tou theou), Tertullian has ὁ Χριστός (ho Christos, “the Christ”); C³ Θ* ƒ¹ 33 565 lat read ὁ Χριστὸς ὁ υἱὸς τοῦ θεοῦ (ho Christos ho huios tou theou, “the Christ, the Son of God”); two versional witnesses (b sy^c) have ὁ υἱὸς τοῦ θεοῦ (“the Son of God”); the Byzantine text as well as many others (Ψ 0250 ƒ¹³ 33 M) read ὁ Χριστὸς ὁ υἱὸς τοῦ θεοῦ τοῦ ζῶντος (ho Christos ho huios tou theou tou zōntos, “the Christ, the Son of the living God”); and P⁶⁶ as well as a few versions have ὁ Χριστὸς ὁ ἅγιος τοῦ θεοῦ (“the Christ, the Holy One of God”). The reading ὁ ἅγιος τοῦ θεοῦ is, however, well supported by P⁷⁵ א B C* D L W as well as versional witnesses. It appears that Peter’s confession in the Synoptic Gospels (especially Matt 16:16) supplied the motivation for the variations. Although the witnesses in Matt 16:16; Mark 8:29; and Luke 9:20 vary considerably, the readings are all intra-synoptic, that is, they do not pull in “the Holy One of God” but reflect various permutations of “Christ”/“Christ of God”/“Christ, the Son of God”/“Christ, the Son of the living God.” The wording “the Holy One of God” (without “Christ”) in significant witnesses here is thus unique among Peter’s confessions, and best explains the rise of the other readings.

¹ tc Most mss (A C³ Ψ 33 M) have τὸ εἰ δύνασαι πιστεῦσαι (to ei dunasai pisteusai, “if you are able to believe”), instead of τὸ εἰ δύνῃ (to ei dunē, “if you are able”; supported by א B C* L N* Δ ƒ¹ 579 892). Others have εἰ δύνῃ (or δυνάσαι) πιστεῦσαι (“if you are able to believe”; so D K Θ ƒ¹³ 28 565 al), while still others have τοῦτο εἰ δύνῃ (touto ei dunē, “if you can [do] this”; so [P⁴⁵] W). The reading that best explains the rise of the others is τὸ εἰ δύνῃ. The neuter article indicates that the Lord is now quoting the boy’s father who, in v. 22, says εἴ τι δύνῃ (ei ti dunē, “if you are able to do anything”). The article is thus used anaphorically (see ExSyn 238). However, scribes could easily have overlooked this idiom and would consequently read τὸ εἰ δύνῃ as the protasis of a conditional clause of the Lord’s statement. As such, it would almost demand the infinitive πιστεῦσαι, producing the reading τὸ εἰ δύνασαι πιστεῦσαι (“if you are able to believe, all things are possible…”). But the article here seems to be meaningless, prompting other scribes to modify the text still further. Some dropped the nonsensical article, while others turned it into the demonstrative τοῦτο and dropped the infinitive. It is clear that scribes had difficulty with the initial text here, and made adjustments in various directions. What might not be so clear is the exact genealogy of the descent of all the readings. However, τὸ εἰ δύνῃ is both a hard saying, best explains the rise of the other readings, and is supported by the best witnesses. It thus rightly deserves to be considered authentic.

¹ tc The MT reads אֲעִידֵךְ (ʾaʿidekh), Hiphil imperfect first person common singular + second person feminine singular suffix from עָדָה (ʿadah, “to testify”): “[How] can I testify for you?” However, Latin Vulgate comparabo te reflects the reading אֶעֱרָךְ (ʾeʿerakh), Qal imperfect first person common singular from עָרַךְ (ʿarakh, “to liken”): “[To what] can I liken [you]?” The verb עָרַךְ (ʿarakh) normally means “to lay out, set in rows; to get ready, set in order; to line up for battle, set battle formation,” but it also may denote “to compare (as a result of arranging in order), to make equal” (e.g., Pss 40:6; 89:6 [7 HT]; Job 28:17, 19; Isa 40:18; 44:7). The BHS editors suggest the emendation, which involves simple orthographic confusion between ר (resh) and ד (dalet), and deletion of י (yod), which the MT could have added to make sense of the form. The variant is favored based on internal evidence: (1) it is the more difficult reading because the meaning “to compare” for עָרַךְ (ʿarakh) is less common than עָדָה (ʿadah, “to testify”), (2) it recovers a tight parallelism between עָרַךְ (ʿarakh, “to liken”) and דָּמָה (damah, “to compare”) (e.g., Ps 89:6 [7 HT]; Isa 40:18), and (3) the MT reading, “How can I testify for you?” makes little sense in the context. Nevertheless, most English versions hold to the MT reading: KJV, RSV, NRSV, NASB, NIV, TEV, and CEV. This textual emendation was first proposed by J. Meinhold, “Threni 2, 13, ” ZAW 15 (1895): 286.

² tc The translation follows the reading of the Syriac version. The Hebrew text reads, “have grazed [= “shaved” ?] your skulls [as a sign of disgracing them].” Note that the reference shifts from third person, “him,” to second person, “you,” which is common in Hebrew style. The words “people of Israel” have been supplied in the translation to help identify the referent and ease the switch. The reading presupposes יְרֹעוּךְ (y^eroʿukh) a Qal imperfect from the verb רָעַע (raʿaʿ; see BDB 949 s.v. II רָעַע Qal.1, and compare usage in Jer 15:2; Ps 2:9). The MT reads יִרְעוּךְ (yirʿukh), a Qal imperfect from the root רָעָה (raʿah; see BDB 945 s.v. I רָעָה Qal.2.b, for usage). The use of the verb in the MT is unparalleled in the sense suggested, but the resultant figure, if “graze” can mean “shave,” is paralleled in Jer 47:5; 48:37; Isa 7:20. The reading of the variant is accepted on the basis that it is the rarer root; the scribe would have been more familiar with the root “graze” even though it is unparalleled in the figurative nuance implied here. The noun “head/skull” is functioning as an accusative of further specification (see GKC 372 §117.ll, and compare usage in Gen 3:8), i.e., “they crack you on the skull” or “they shave you on the skull.” The verb is a prefixed form and in this context is either a preterite without vav (ו) consecutive or an iterative imperfect denoting repeated action. Some modern English versions render the verb in the future tense, “they will break [or shave] your skull.”

¹ tc The meaning of the MT is unclear. The Kethib (consonantal text) reads יַשִּׁימָוֶת עָלֵימוֹ (yashimavet ʿalemo, “May devastation [be] upon them.”). The proposed noun יַשִּׁימָוֶת occurs only here and perhaps in the place name Beth Jeshimoth in Num 33:49. The Qere (marginal text) has יַשִּׁי מָוֶת עָלֵימוֹ (yashi mavet ʿalemo). The verbal form יַשִּׁי is apparently an alternate form of יַשִּׁיא (yashiʾ), a Hiphil imperfect from נָשַׁא (nashaʾ, “deceive”). In this case one might read “death will come deceptively upon them.” This reading has the advantage of reading מָוֶת (mavet, “death”) which forms a natural parallel with “Sheol” in the next line. The present translation is based on the following reconstruction of the text: יְשִׁמֵּם מָוֶת (y^eshimmem mavet). The verb assumed in the reconstruction is a Hiphil jussive third masculine singular from שָׁמַם (shamam, “be desolate”) with a third masculine plural pronominal suffix attached. This reconstruction assumes that (1) haplography has occurred in the traditional text (the original sequence of three mems [מ] was lost with only one mem remaining), resulting in the fusion of originally distinct forms in the Kethib, and (2) that עָלֵימוֹ (ʿalemo, “upon them”) is a later scribal addition attempting to make sense of a garbled text. The preposition עַל (ʿal) does occur with the verb שָׁמַם (shamam), but in such cases the expression means “be appalled at/because of” (see Jer 49:20; 50:45). If one were to retain the prepositional phrase here, one would have to read the text as follows: יַשִּׁים מָוֶת עָלֵימוֹ (yashim mavet ʿalemo, “Death will be appalled at them”). The idea seems odd, to say the least. Death is not collocated with this verb elsewhere.

³ tc The Byzantine text along with a few other witnesses (א³ A^c C² D² Ψ [88] 1241 1505 1739 1881 M al vg^ms) read θεός (theos, “God”) for ὅς (hos, “who”). Most significant among these witnesses is 1739; the second correctors of some of the other mss tend to conform to the medieval standard, the Byzantine text, and add no independent voice to the textual problem. At least two mss have ὁ θεός (69 88), a reading that is a correction on the anarthrous θεός. On the other side, the masculine relative pronoun ὅς is strongly supported by א* A* C* F G 33 365 1175 Did Epiph. Significantly, D* and virtually the entire Latin tradition read the neuter relative pronoun, ὅ (ho, “which”), a reading that indirectly supports ὅς since it could not easily have been generated if θεός had been in the text. Thus, externally, there is no question as to what should be considered the Ausgangstext: The Alexandrian and Western traditions are decidedly in favor of ὅς. Internally, the evidence is even stronger. What scribe would change θεός to ὅς intentionally? “Who” is not only a theologically pale reading by comparison; it also is much harder (since the relative pronoun has no obvious antecedent, probably the reason for the neuter pronoun of the Western tradition). Intrinsically, the rest of 3:16, beginning with ὅς, appears to form a hymn with six strophes. As such, it is a text that is seemingly incorporated into the letter without syntactical connection. Hence, not only should we not look for an antecedent for ὅς (as is often done by commentators), but the relative pronoun thus is not too hard a reading (or impossible, as Dean Burgon believed). Once the genre is taken into account, the relative pronoun fits neatly into the author’s style (cf. also Col 1:15; Phil 2:6 for other places in which the relative pronoun begins a hymn, as was often the case in poetry of the day). On the other hand, with θεός written as a nomen sacrum, it would have looked very much like the relative pronoun: q-=s vs. os. Thus, it may have been easy to confuse one for the other. This, of course, does not solve which direction the scribes would go, although given their generally high Christology and the bland and ambiguous relative pronoun, it is doubtful that they would have replaced θεός with ὅς. How then should we account for θεός? It appears that sometime after the 2nd century the θεός reading came into existence, either via confusion with ὅς or as an intentional alteration to magnify Christ and clear up the syntax at the same time. Once it got in, this theologically rich reading was easily able to influence all the rest of the mss it came in contact with (including mss already written, such as א A C D). That this reading did not arise until after the 2nd century is evident from the Western reading, ὅ. The neuter relative pronoun is certainly a “correction” of ὅς, conforming the gender to that of the neuter μυστήριον (mustērion, “mystery”). What is significant in this reading is (1) since virtually all the Western witnesses have either the masculine or neuter relative pronoun, the θεός reading was apparently unknown to them in the 2nd century (when the “Western” text seems to have originated, though its place of origination was most likely in the east); they thus supply strong indirect evidence of ὅς outside of Egypt in the 2nd century; (2) even 2nd century scribes were liable to misunderstand the genre, feeling compelled to alter the masculine relative pronoun because it appeared to them to be too harsh. The evidence, therefore, for ὅς is quite compelling, both externally and internally. As TCGNT 574 notes, “no uncial (in the first hand) earlier than the eighth or ninth century (Ψ) supports θεός; all ancient versions presuppose ὅς or ὅ; and no patristic writer prior to the last third of the fourth century testifies to the reading θεός.” Thus, the cries of certain groups that θεός has to be original must be seen as special pleading. To argue that heretics tampered with the text here is self-defeating, for most of the Western fathers who quoted the verse with the relative pronoun were quite orthodox, strongly affirming the deity of Christ. They would have dearly loved such a reading as θεός. Further, had heretics introduced a variant to θεός, a far more natural choice would have been Χριστός (Christos, “Christ”) or κύριος (kurios, “Lord”), since the text is self-evidently about Christ, but it is not self-evidently a proclamation of his deity. (See ExSyn 341-42, for a summary discussion on this issue and additional bibliographic references.)

¹ tc The text and consequent meaning of these first two lines are uncertain. The Masoretic reads literally, “Against let him string. Let him string, the one who strings his bow, and against let him raise himself up in his coat of armor.” This makes absolutely no sense, and the ancient versions and Hebrew mss did not agree in reading this same text. Many Hebrew mss and all the versions, as well as the Masoretes themselves (the text is left unpointed with a marginal note not to read it), delete the second “let him string.” The LXX (or Greek version) left out the words “against” at the beginning of the first two lines. It reads, “Let the archer bend his bow, and let the one who has armor put it on.” The Lucianic recension of the LXX and some Targum mss supplied the missing object “it” and thus read, “Let the archer ready his bow against it, and let him array himself against it in his coat of mail.” This makes good sense but does not answer the question of why the Hebrew text left off the suffix on the preposition twice in a row. Many Hebrew mss and the Syriac, Targum, and Vulgate (the Latin version) change the pointing of “against” (אֶל [ʾel]) to “not” (אַל [ʾal]) and thus read, “Let the archer not string the bow, and let him not array himself in his armor.” However, many commentators feel that this does not fit the context because it would apparently be addressed to the Babylonians, not the enemy, which would create a sudden shift in addressee with the second half of the verse. However, if it is understood in the sense taken here, it refers to the enemy not allowing the Babylonian archers to get ready for the battle, i.e., a surprise attack. This sense is suggested as an alternative in J. Bright, Jeremiah (AB), 346, n. u-u, and J. A. Thompson, Jeremiah (NICOT), 747, n. 5, and is the interpretation adopted in TEV, and probably also in NIrV.

⁴ tc ‡ Most mss (א M^K) do not add the words “[as] their God” (αὐτῶν θεός, autōn theos) after “he will be with them.” The mss with these words include A 2030 2050 2329 al. The Andreas group (M^A) also has the words, but in a different arrangement with the preceding (ἔσται μετ᾿ αὐτῶν θεὸς αὐτῶν, estai met’ autōn theos autōn). Not only do the words float, but scribes may have been motivated to make a connection here more directly with Isa 7:14; 8:8; Jer 24:7; 31:33; Zech 8:8. In light of sufficient external evidence as well as the possibility that the longer reading is theologically motivated, the shorter reading is preferred. NA²⁸ places the words in brackets, indicating doubts as to their authenticity.

⁷ tc Most later witnesses (P^vid Ψ 5 88 1175 1611 1735 2492 M sy) have θεόν (theon, “God”) after δεσπότην (despotēn, “master”), which appears to be a motivated reading in that it explicitly links “Master” to “God” in keeping with the normal NT pattern (see Luke 2:29; Acts 4:24; 2 Tim 2:21; Rev 6:10). In patristic Greek, δεσπότης (despotēs) was used especially of God (cf. BDAG 220 s.v. 1.b.). The earlier and better witnesses (P^72,78 א A B C 0251 33 81 323 436 442 1241 1243 1739 2344 al co) lack θεόν; the shorter reading is thus preferred on both internal and external grounds.

³ tc A number of variants are generated from the simple τὸν ᾿Ιησοῦν (ton Iēsoun, “Jesus”), some of which turn the expression into an explicit object-complement construction. ᾿Ιησοῦν κύριον (Iēsoun kurion, “Jesus as Lord”) is found in א, ᾿Ιησοῦν Χριστόν (Iēsoun Christon, “Jesus as Christ”) is read by 5 (442 1175) 1243 1735 2492 M, τὸν Χριστόν (“the Christ”) is the reading of 1846, τὸν Χριστὸν ᾿Ιησοῦν (“Christ Jesus” or “Jesus as the Christ”) is the wording of 307, and ᾿Ιησοῦν without the article is found in 1881 2464. But τὸν ᾿Ιησοῦν is well supported by A B Ψ 33 81 436 1611 1739 1852 2344, and internally best explains the rise of the others. It is thus preferred on both external and internal grounds.

² tc The expression “consciousness/conscience of God” (συνείδησιν θεοῦ; suneidēsin theou) is unusual, occurring only here in the NT. Because θεοῦ was liable to misinterpretation, several witnesses altered the text, either replacing it with ἀγαθήν (agathēn; C Ψ 323 442 614 630 945 1175 1241 1243 1505 1611 1739 1852 2492 sy) or expanding the expression by adding ἀγαθήν before θεοῦ (P⁷² [A* 33 2344] 81). Replacing θεοῦ with ἀγαθήν conforms to other NT phrases, notably in this same letter (Acts 23:1; 1 Tim 1:5, 19; 1 Pet 3:16, 21), suggesting that such a reading is motivated. The reading θεοῦ, however, has superior support (א A^c B P 049 5 436 1735 M lat co), and best explains the rise of the others.

⁵ tc A few mss (A B 1852 vg) lack καθαρᾶς (katharas, “pure”) and read simply καρδίας (kardias, “from the heart”) ”) or καρδίας ἀληθινῆς (kardias alēthinēs, “from a true heart,” found in א² vgms), but there is excellent ms support (P⁷² א* C P Ψ 33 1739 M co) for the word. The omission may have been accidental. In the majuscule script (kaqaras kardias) an accidental omission could have happened via homoioteleuton or homoioarcton. καθαρᾶς should be considered the initial reading. The NA²⁸ prints καθαρᾶς καρδίας with a diamond, indicating that the decision was a toss-up or, in the words of the preface, “there are two variants which in the editors’ judgement could equally well be adopted in the reconstructed initial text.”

² tc Most later mss (P M) have διὰ πνεύματος (dia pneumatos, “through the Spirit”) after ἀληθείας (alētheias, “truth”), while the words are lacking in a broad spectrum of early and significant witnesses (P⁷² א A B C Ψ 33 81 323 945 1241 1739 vg sy co). On external grounds, the shorter reading cannot be easily explained if it were not autographic. The longer reading is clearly secondary, added to show more strongly God’s part in man’s obedience to the truth. But the addition ignores the force that the author gives to “purified” and ruins the balance between v. 22 and v. 23 (for in v. 23 the emphasis is on God’s part; here, on mankind’s).

¹ tc Several witnesses, many of them early and significant (א A C D* P Ψ 0243 0278 33 1739 1881 al lat co), have at the end of v 7, “You have given him dominion over the works of your hands.” Other mss, not quite as impressive in weight, lack the words (P⁴⁶ B D² M). In spite of the impressive external evidence for the longer reading, it is most likely a scribal addition to conform the text of Hebrews to Ps 8:6 (8:7 LXX). Conformity of a NT quotation of the OT to the LXX was a routine scribal activity, and can hardly be in doubt here as to the cause of the longer reading.

³ tc ‡ ὑμῖν (humin, “you”) is found in many valuable witnesses (P⁶¹ א F G P 0278 33 1505 1739 1881 al sy co). The witnesses for ἡμῖν (hēmin, “us”) are not as strong (A C D K L P Ψ 048^vid 1241 M), but nevertheless represent a broad base. Internally, ἡμῖν could be favored because of second person pronouns surrounding it, making it the harder reading. On the other hand, the last second person plural pronoun was in v. 3, and the next one will not show up until v. 22, a fact which tends to counter the internal argument on behalf of ἡμῖν. Although a decision is difficult, with the internal evidence being capable of favoring either reading, our preference is based on the external evidence; ὑμῖν is thus slightly preferred.

² tc Most mss (A D F G Ψ 1175 1241 1505 M al lat sy) read ἁμαρτίας (hamartias, “of sin”) here, but several significant mss (א B 0278 6 81 1739 1881 2464 al co) read ἀνομίας (anomias, “of lawlessness”). Although external support for ἁμαρτίας is broader, the generally earlier and better witnesses are on the side of ἀνομίας. Internally, since ἁμαρτία (hamartia, “sin”) occurs nearly ten times as often as ἀνομία (anomia, “lawlessness”) in the corpus Paulinum, scribes would be expected to change the text to the more familiar term. At the same time, the mention of ἀνομία in v. 7 and ὁ ἄνομος (ho anomos, “the lawless one”) in v. 8, both of which look back to v. 3, may have prompted scribes to change the text toward ἀνομίας. The internal evidence is thus fairly evenly balanced. Although a decision is difficult, ἀνομίας has slightly greater probability of authenticity than ἁμαρτίας.

¹ tc ‡ δέ (de, “now”) is found in א² B D 0226 6 1505 1739 1881 2464 al, but lacking in א* A F G 33 it. γάρ (gar, “for”) is the reading of the Byzantine text and a few other witnesses (Ψ 0278 1175 1241 M al). Although normally the shorter reading is to be preferred, the external evidence is superior for δέ (being found in the somewhat better Alexandrian and Western witnesses). What, then, is to explain the γάρ? Scribes were prone to replace δέ with γάρ, especially in sentences suggesting a causal or explanatory idea, thus making the point more explicit. Internally, the omission of δέ looks unintentional, a case of homoioarcton (otandelegwsin). Although a decision is difficult, in this instance δέ has the best credentials for authenticity.

² tc If the name Nympha is accented with a circumflex on the ultima (Νυμφᾶν, Numphan), then it refers to a man; if it receives an acute accent on the penult (Νύμφαν), the reference is to a woman. Scribes that considered Nympha to be a man’s name had the corresponding masculine pronoun αὐτοῦ here (autou, “his”; so D [F G] Ψ [1505] M), while those who saw Nympha as a woman read the feminine αὐτῆς here (autēs, “her”; B 0278 6 1739^[*^] 1881 sa). Several mss (א A C P 075 33 81 104 326 1175 2464 bo) have αὐτῶν (autōn, “their”), perhaps because of indecisiveness on the gender of Nympha, perhaps because they included ἀδελφούς (adelphous, here translated “brothers and sisters”) as part of the referent. The harder reading is certainly αὐτῆς, and thus Nympha should be considered a woman.