⁴ tc The majority of mss, especially the later Byzantines, include the phrase “to heal the brokenhearted” at this point (A Θ Ψ 0102 ƒ¹ M). The phrase is lacking in several weighty mss (א B D L W Ξ ƒ¹³ 33 579 700 892* lat sy^s co), including representatives from both the Alexandrian and Western text-forms. From the standpoint of external evidence, the omission of the phrase is more likely what the initial text read. When internal evidence is considered, the shorter reading becomes almost certain. Scribes would be much more prone to add the phrase here to align the text with Isa 61:1, the source of the quotation, than to remove it from the initial text.

¹ tc The number and order of the first few names in this verse varies greatly in the mss. The variants which are most likely to be authentic based upon external evidence are Amminadab, Aram (A D 33 565 [1424] pm lat); Amminadab, Aram, Joram (K Δ Ψ 700 2542 pm); Adam, Admin, Arni (P^4vid א* 1241 sa); and Amminadab, Admin, Arni (א² L X [Γ] ƒ¹³). Deciding between these variants is quite difficult. The reading “Amminadab, Aram” is the strongest externally since it is represented by Alexandrian, Western, and Byzantine witnesses, although it is significantly weaker internally because it disrupts the artistic balance of the number of generations and their groups that three names would preserve (see TCGNT 113, fn. 1 for discussion). In this case, the subtle intrinsic arguments that would most likely be overlooked by scribes argues for the reading “Amminadab, Admin, Arni,” although a decision is quite difficult because of the lack of strong external support.

¹ tc The vast majority of witnesses (א A C L W Θ Ψ ƒ^1,13 M lat sy co) have καὶ προσεύχεσθε after ἀγρυπνεῖτε (agrupneite kai proseuchesthe, “stay alert and pray”). This may be a motivated reading, influenced by the similar command in Mark 14:38 where προσεύχεσθε is solidly attested, and more generally from the parallel in Luke 21:36 (though δέομαι [deomai, “ask”] is used there). As B. M. Metzger notes, it is a predictable variant that scribes would have been likely to produce independently of each other (TCGNT 95). The words are not found in B D a c d k. Although the external evidence for the shorter reading is slender, it probably better accounts for the longer reading than vice versa.

³ tc Most mss, especially the later ones (A^[*^] W Θ ƒ^1,13 M sy), have “on the third day” (τῇ τρίτῃ ἡμέρᾳ, tē tritē hēmera) instead of “after three days.” But not only does Mark nowhere else speak of the resurrection as occurring on the third day, the idiom he uses is a harder reading (cf. Mark 8:31; 9:31, though in the latter text the later witnesses also have τῇ τρίτῃ ἡμέρᾳ). Further, τῇ τρίτῃ ἡμέρᾳ conforms to the usage that is almost universally used in Matthew and Luke, and is found in the parallels to this text (Matt 20:19; Luke 18:33). Thus, scribes would be doubly motivated to change the wording. The most reliable witnesses, along with several other mss (א B C D L Δ Ψ 579 892 it co), have resisted this temptation.

³ tc “Those who brought them” (ἐπετιμῶν τοῖς προσφέρουσιν, epetimōn tois prospherousin) is the reading of most mss (A D W [Θ ƒ^1,13] M lat sy), but it is probably a motivated reading. Since the subject is not explicit in the earliest and best witnesses as well as several others (א B C L Δ Ψ 579 892), scribes would be prone to add “those who brought them” here to clarify that the children were not the ones being scolded. Both on external and internal grounds, the shorter reading is strongly preferred. Similar motivations are behind the translation here, namely, “those who brought them” has been supplied to ensure that the parents who brought the children are in view, not the children themselves.

¹ tc Most mss have ὁ θεός (ho theos, “God”) as the explicit subject of ἐποίησεν (epoiēsen, “he made”; A D W Θ Ψ ƒ^1,13 M lat sy), while the most significant witnesses, along with a few others, lack ὁ θεός (א B C L Δ 579 co). On the one hand, it is possible that the shorter reading is an assimilation to the wording of the LXX of Gen 1:27b where ὁ θεός is lacking. However, since it is mentioned at the beginning of the verse (Gen 1:27a) with ἐποίησεν scribes may have been motivated to add it in Mark to make the subject clear. Further, confusion could easily arise in this dominical saying because Moses was the previously mentioned subject (v. 5) and inattentive readers might regard him as the subject of ἐποίησεν in v. 6. Thus, both on internal and external grounds, the most probable wording of the Ausgangstext here lacked ὁ θεός.

³ tc Alexandrian and other witnesses (א B C* L Ψ 0274 892 co) read καὶ πέραν (kai peran, “and beyond”), while Western and Caesarean witnesses (C² D W Δ Θ ƒ^1,13 28 565 579 1241 al) read πέραν (simply “beyond”). It is difficult to decide between the Alexandrian and Western readings here, but since the parallel in Matt 19:1 omits καί the weight is slightly in favor of including it here; scribes may have omitted the word here to harmonize this passage to the Matthean passage. Because of the perceived geographical difficulties found in the earlier readings (omission of the word “and” would make it seem as though Judea is beyond the Jordan), the majority of the witnesses (A M) read διὰ τοῦ πέραν (dia tou peran, “through the other side”), perhaps trying to indicate the direction of Jesus’ travel.

⁴ tc In place of ἠπόρει (ēporei, “he was baffled”) the majority of mss (A C D ƒ¹ 33 M lat sy) have ἐποίει (epoiei, “he did”; cf. KJV’s “he did many things.”) The best mss (א B L [W] Θ co) support the reading followed in the translation. The variation may be no more than a simple case of confusion of letters, since the two readings look very much alike. The verb ποιέω (poieō, “I do”) certainly occurs more frequently than ἀπορέω (aporeō, “I am at a loss”), so a scribe would be more likely to write a more familiar word. Further, even though the reading ἐποίει is the harder reading in terms of the sense, it is virtually nonsensical here, rendering it most likely an unintentional scribal error.

³ tc Most mss (A^c Γ Δ Θ ƒ¹³ 565 579 700 1241 1424 M it sy) have ἀμήν (amēn, “amen”) at the end of v. 20. Such a conclusion is routinely added by scribes to NT books because a few of these books originally had such an ending (cf. Rom 16:27; Gal 6:18; Jude 25). A majority of Greek witnesses have the concluding ἀμήν in every NT book except Acts, James, and 3 John (and even in these books, ἀμήν is found in some witnesses). It is thus a predictable variant. Further, no good reason exists for the omission of the particle in significant and early witnesses such as א A* B D W ƒ¹ 33 al lat sa.

¹ tc The problematic citing of Jeremiah for a text which appears to come from Zechariah has prompted certain scribes to alter it. Codex 22 has Ζαχαρίου (Zachariou, “Zechariah”) while Φ 33 and several versional witnesses omit the prophet’s name altogether. And codex 21 and the Latin ms l change the prophet’s name to “Isaiah,” in accordance with natural scribal proclivities to alter the text toward the most prominent OT prophet. But unquestionably the name Jeremiah is the wording of the original here because it is supported by virtually all witnesses and because it is the harder reading. See D. A. Carson, “Matthew,” EBC 8:562-63, for a discussion of the textual and especially hermeneutical problem.

² tc Most witnesses (C Γ Δ Θ 0102 ƒ^1,13 700 1241 1424 M) have “and plagues” (καὶ λοιμοί, kai loimoi) between “famines” (λιμοί, limoi) and “earthquakes” (σεισμοί, seismoi), while others have “plagues and famines and earthquakes” (L W 33 lat). The similarities between λιμοί and λοιμοί could explain how καὶ λοιμοί might have accidentally dropped out, but since the Lukan parallel (Luke 21:11) has both terms (and W lat have the order λοιμοὶ καὶ λιμοί there too, as they do in Matthew), it seems more likely that scribes added the phrase here. The shorter reading does not enjoy overwhelming support ([א] B D 892 sa, and other Greek and versional witnesses), but it is nevertheless significant; coupled with the internal evidence it should be given preference.

¹ tc Most mss (א*^,2b C L N W Γ Δ 0106 33 565 1241 1424 M) have an expanded introduction here; instead of “For God said,” they read “For God commanded, saying” (ὁ γὰρ θεὸς ἐνετείλατο λέγων, ho gar theos eneteilato legōn). But such expansions are generally motivated readings; in this case, most likely it was due to the wording of the previous verse (“the commandment of God”) that caused early scribes to add to the text. Although it is possible that other witnesses reduced the text to the simple εἶπεν (eipen, “[God] said”) because of perceived redundancy with the statement in v. 3, such is unlikely in light of the great variety and age of these authorities (א^2a B D Θ 073 ƒ^{1, 13} 579 700 892 lat co, as well as other versions and fathers).

¹ tc A few ancient mss and versions lack this verse (א* B L Γ ff¹ k sy^s,c sa), while almost all the rest have it. The witness of א and B is especially strong, but internal considerations override this external evidence. Both v. 46 and v. 47 end with the word λαλῆσαι (“to speak”), so early scribes probably omitted the verse through homoioteleuton. Without v. 47 the passage is nonsensical: v. 46 says Jesus’ family members wanted to speak to him, and v. 48 begins with “to the one who said this.” The antecedent of “the one who said this” must surely be “someone” in v. 47. Thus, the omission of v. 47 is too hard a reading. This verse therefore should be regarded as part of the Ausgangstext.

² tc Most mss (L W Δ Θ 0233 ƒ¹³ 33 565 579 700 1241 1424 M sy sa Didache) read (though some with slight variation) ὅτι σοῦ ἐστιν ἡ βασιλεία καὶ ἡ δύναμις καὶ ἡ δόξα εἰς τοὺς αἰῶνας, ἀμήν (“for yours is the kingdom and the power and the glory forever, amen”) here. The reading without this sentence, though, is attested by generally better witnesses (א B D Z 0170 ƒ¹ lat mae Or). The phrase was probably composed for the liturgy of the early church and most likely was based on 1 Chr 29:11-13; a scribe probably added the phrase at this point in the text for use in public scripture reading (see TCGNT 13-14). Both external and internal evidence argue for the shorter reading.

¹ tc The LXX of Jer 38:15 (31:15 ET) has “lamentation, weeping, and loud wailing”; most later mss (C D L W Γ Δ 0233 ƒ¹³ 33 565 579 700 1241 1424 M) have a quotation in Matthew which conforms to that of the LXX (θρῆνος καὶ κλαυθμὸς καὶ ὀδυρμός; thrēnos kai klauthmos kai odurmos). But such assimilations were routine among the scribes; as such, they typically should be discounted because they are both predictable and motivated. The shorter reading, without “lamentation and,” is thus to be preferred, especially since it cannot easily be accounted for unless it is the reading that gave rise to the other reading. Further, it is found in the better mss along with a good cross-section of other witnesses (א B Z 0250 ƒ¹ lat co).

² tc The MT reads יָגוּרוּ (yaguru, Qal imperfect third person common plural from III גוּר, gur, “to dread”; see BDB 159 s.v. III גוּר 1). This reading is followed by most English versions but is syntactically awkward because III גוּר (“to dread”) is used nowhere else with the preposition לְ (lamed, “they are in dread for…”?). BDB suggests reading יָנוּדוּ (yanudu, Qal imperfect third person common plural from נוּד, nud, “to lament”; BDB 626 s.v. נוּד 2.a), which harmonizes better in the parallelism with אָבַל (ʾaval, “to mourn”) in the following line. The verb נוּד (“to lament”) is used with the preposition לְ in the idiom “to lament for” (e.g., Isa 51:19; Jer 15:5; 16:5; 48:17; Nah 3:7). This involves simple orthographic confusion between ג (gimel) and נ (nun), as well as ר (resh) and ד (dalet), which were often confused by the scribes.

³ tc The MT reads the plural לְעֶגְלוֹת (l^eʿeglot, “for the calves”), while some Greek versions (LXX, Theodotion) and the Syriac reflect the singular לְעֵגֶל (“for the calf [calf idol]”). The singular reading is preferred on the basis of internal evidence: the oracle denounces the calf-idol worship of Samaria. The plural form probably arose due to the ambiguity of the term “calf” when a scribe did not realize that the term was being used as a metonymy for the worship of the Egyptian calf goddess. Most recent English versions adopt the singular form and relate it to the calf goddess cult (RSV, NASB, NIV, NCV, NJPS, TEV, CEV, NLT); however, older English versions follow the MT plural (KJV, ASV).

² tc The MT reads a plural verb לֹא נָשָׂאוּ (loʾ nasaʾu, “they did not lift up”) from נָשָׂא (nasaʾ, “to lift up”); however, the ancient versions (LXX, Aramaic Targum, Latin Vulgate, Syriac Peshitta) have singular verbs, reflecting לֹא נָשָׂא (loʾ nasaʾ, “he did not lift up”). D. R. Hillers suggests that the MT plural is an intentional scribe change, to avoid the appearance that God brought about evil on the priests and elders. It may also be that the third person plural presumes an indefinite subject and the construction is used in place of a passive, but still essentially means, “the priests were not honored” (see following note regarding the idiom). Another alternative would be to revocalize the verb as the rare Qal passive, which would yield the same result.

¹ tc The words “And the women added” are not in the Hebrew text. They are, however, implicit in what is said. They are found in the Syriac version and in one recension of the Greek version. W. L. Holladay (Jeremiah [Hermeneia], 2:279, n. 19a) suggests that these words are missing from the Hebrew text because of haplography, i.e., that the scribe left out וַהַנָּשִׁים אָמְרוּ כִי (vahannashim ʾam^eru khi) because his eye jumped from the ו (vav) at the beginning to the כִּי (ki) that introduced the temporal clause and left out everything in between. It is, however, just as likely, given the fact that there are several other examples of quotes not formally introduced in the book of Jeremiah, that the words were not there and were supplied by these two ancient versions as a translator’s clarification.

² tc The translation follows the reading of the LXX (Greek version). The Hebrew reads, “will dance and be glad, young men and old men together.” The Greek version presupposes a Qal imperfect of a rare verb (יַחְדּוּ [yakhdu] from the verb חָדָה [khadah]; see BDB 292 s.v. II חָדָה Qal), as opposed to the Hebrew text, which reads a common adverb יַחְדָּו (yakhdav). The consonantal text is the same, but the vocalization is different. There are no other examples of the syntax of the adverb used this way (i.e., of a compound subject added to a third subject), and the vocalization of the Hebrew text can be explained on the basis of a scribe misvocalizing the text based on his greater familiarity with the adverb.