¹ tc Many important mss (א* B Θ 0281 33 579 892* pc e ff¹ sy^s,c sa) do not include 17:21 “But this kind does not go out except by prayer and fasting.” The verse is included in א² C D L W Ë^1,13 Ï lat, but is almost certainly not original. As B. M. Metzger notes, “Since there is no satisfactory reason why the passage, if originally present in Matthew, should have been omitted in a wide variety of witnesses, and since copyists frequently inserted material derived from another Gospel, it appears that most manuscripts have been assimilated to the parallel in Mk 9.29” (TCGNT 35). The present translation follows NA²⁷ in omitting the verse number as well, a procedure also followed by a number of other modern translations.

¹ tc Although the external evidence for the inclusion of “Jesus” before “Barabbas” (in vv. 16 and 17) is rather sparse, being restricted virtually to the Caesarean text (Θ Ë¹ 700* pc sy^s), the omission of the Lord’s name in apposition to “Barabbas” is such a strongly motivated reading that it can hardly be original. There is no good explanation for a scribe unintentionally adding ᾿Ιησοῦν (Ihsoun) before Βαραββᾶν (Barabban), especially since Barabbas is mentioned first in each verse (thus dittography is ruled out). Further, the addition of τὸν λεγόμενον Χριστόν (ton legomenon Criston, “who is called Christ”) to ᾿Ιησοῦν in v. 17 makes better sense if Barabbas is also called “Jesus” (otherwise, a mere “Jesus” would have been a sufficient appellation to distinguish the two).

¹ tc The words “when they rise again” are missing from several important witnesses (א B C D L W Δ Ψ 33 579 892 2427 pc c r¹ k sy^p co). They are included in A Θ Ë^1,(13) Ï lat sy^s,h. The strong external pedigree of the shorter reading gives one pause. Nevertheless, the Alexandrian and other mss most likely dropped the words from the text either to conform the wording to the parallel in Matt 22:28 or because “when they rise again” was redundant. But the inclusion of these words is thoroughly compatible with Mark’s usually pleonastic style (see TCGNT 93), and therefore most probably authentic to Mark’s Gospel.

⁴ tc The majority of mss, especially the later Byzantines, include the phrase “to heal the brokenhearted” at this point (A Θ Ψ 0102 ˹ Ï). The phrase is lacking in several weighty mss (א B D L W Ξ Ë¹³ 33 579 700 892* pc lat sy^s co), including representatives from both the Alexandrian and Western texttypes. From the standpoint of external evidence, the omission of the phrase is more likely original. 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 original.

³ tc ‡ Most mss, including the majority of later mss (א[* C] A L W Θ Ξ [Ψ] Ë^1,13 33 [1424] Ï [lat sy^c,p,h]) read here, “having spent all her money on doctors.” Uncertainty over its authenticity is due primarily to the fact that certain important witnesses do not have the phrase (e.g., Ì⁷⁵ B [D] 0279 sy^s sa Or). This evidence alone renders its authenticity unlikely. It may have been intentionally added by later scribes in order to harmonize Luke’s account with similar material in Mark 5:26 (see TCGNT 121). NA²⁷ includes the words in brackets, indicating doubt as to their authenticity.

⁴ tc Ì⁷⁵ א* D ˹ 565 as well as several versions and fathers lack the phrase “is superior to all” (ἐπάνω πάντων ἐστίν, epanw pantwn estin). This effectively joins the last sentence of v. 31 with v. 32: “The one who comes from heaven testifies about what he has seen and heard, but no one accepts his testimony.” On the other side, the phrase may have been deleted because of perceived redundancy, since it duplicates what is said earlier in the verse. The witnesses that include ἐπάνω πάντων ἐστίν in both places are weighty and widespread (Ì^36vid,66 א² A B L W^s Θ Ψ 083 086 ˹³ 33 Ï lat sy^s,p,h bo). On balance, the longer reading should probably be considered authentic.

² tc ‡ Important and early witnesses (א B L W 070 it sy^s,p co) lack the conjunction here, while the earliest witnesses along with many others read οὖν (oun, “therefore”; Ì^66,75 C D Θ Ψ 0250 ˹³ 33 Ï). This conjunction occurs in John some 200 times, far more than in any other NT book. Even though the combined testimony of two early papyri for the conjunction is impressive, the reading seems to be a predictable scribal emendation. In particular, οὖν is frequently used with the plural of εἶπον (eipon, “they said”) in John (in this chapter alone, note vv. 13, 39, 48, 57, and possibly 52). On balance, it is probably best to consider the shorter reading as authentic, even though “Then” is virtually required in translation for English stylistic reasons. NA²⁷ has the conjunction in brackets, indicating some doubt as to its authenticity.

¹ tc The translation here follows the reading στήσητε (sthshte, “set up”) found in D W Θ Ë¹ 28 565 2542 it sy^s,p Cyp. The majority of mss here read τηρήσητε (thrhsete; א A L ˹³ 33 Ï co) or τηρῆτε (thrhte; B 2427), both translated “keep.” It is hard to know which reading is best: On the one hand, τηρήσητε/τηρῆτε has much stronger external support, but στήσητε is a more difficult reading. What makes “keep” suspect is that it appears in two different forms, suggesting independent alterations of a difficult reading. Further, scribes may have been influenced by the preceding “commandment of God” to change the text toward “keep” (TCGNT 81), a common enough expression (cf. Matt 19:17; John 14:15; 1 Tim 6:1; 1 John 5:3; Rev 14:12). Thus, the more difficult reading is “set up.” Also, the more natural opposite of “reject” (ἀθεῖτε [aqeite], literally “you set aside”) is “set up.” However, the Western reading may have been influenced by Exod 6:4 or Heb 10:9, but this likelihood seems remote. Thus, “set up” is more likely to be the original wording of Mark here.

¹ tc The earliest mss ([א] B L [W] Δ 0274 Ë^1,13 28* 565 700 pc sy^s sa) have the reading adopted by the translation. Codex Bezae (D) and several Itala read “Every sacrifice will be salted with salt.” The majority of other mss (A C Θ Ψ [2427] Ï lat sy^p,h) have both readings, “Everyone will be salted with fire, and every sacrifice will be salted with salt.” An early scribe may have written the LXX text of Lev 2:13 (“Every sacrifice offering of yours shall be salted with salt”) in the margin of his ms. At a later stage, copyists would either replace the text with this marginal note or add the note to the text. The longer reading thus seems to be the result of the conflation of the Alexandrian reading “salted with fire” and the Western reading “salted with salt.” The reading adopted by the text enjoys the best support and explains the other readings in the ms tradition.

² tc The phrase “or under a basket” is lacking in some important and early mss (Ì^45,75 L Γ Ξ 070 ˹ 700* 1241 2542 pc sy^s sa). It is hard to decide in this case, since the inclusion of “or under a basket” is widely attested by some early and decent witnesses, as well as the overwhelming majority of mss (א A B C D W Θ Ψ Ë¹³ Ï latt). The parallel passage in Luke 8:16 does not include “under a basket.” If the phrase “under a basket” were added as a harmonization with Mark 4:21 and Matt 5:15, it is perhaps surprising that scribes did not add the phrase at Luke 8:16 as well. It seems somewhat more likely that a scribe copying Luke would be inclined to harmonize 11:33 with 8:16 by omitting the phrase here. Thus, the words “or under a basket” seem to have the marks of authenticity.

¹ tc Many important mss (Ì⁷⁵ א¹ B D* W Θ 070 579 1241 pc sy^s sa) lack v. 34a. It is included in א*^,2 (A) C D² L Ψ 0250 Ë^1,(13) 33 Ï lat sy^c,p,h. It also fits a major Lukan theme of forgiving the enemies (6:27-36), and it has a parallel in Stephen’s response in Acts 7:60. The lack of parallels in the other Gospels argues also for inclusion here. On the other hand, the fact of the parallel in Acts 7:60 may well have prompted early scribes to insert the saying in Luke’s Gospel alone. Further, there is the great difficulty of explaining why early and diverse witnesses lack the saying. A decision is difficult, but even those who regard the verse as inauthentic literarily often consider it to be authentic historically. For this reason it has been placed in single brackets in the translation.

² tc ‡ Two early and important Greek mss along with two versional witnesses (Ì⁷⁵ B sy^s ac²) lack ἡ γυνή (Jh gunh, “the woman”) here; א* has ἐκείνη (ekeinh, “that one” or possibly “she”) instead of ἡ γυνή. It is possible that no explicit subject was in the original text and scribes added either ἡ γυνή or ἐκείνη to make the meaning clear. It is also possible that the archetype of Ì⁷⁵ א B expunged the subject because it was not altogether necessary, with the scribe of א later adding the pronoun. However, ἡ γυνή is not in doubt in any other introduction to the woman’s words in this chapter (cf. vv. 9, 15, 17, 19, 25), suggesting that intentional deletion was not the motive for the shorter reading in v. 11 (or else why would they delete the words only here?). Thus, the fact that virtually all witnesses (Ì⁶⁶ א² A C D L W^s Θ Ψ 050 083 086 Ë^1,13 Ï latt sy^c,p,h sa bo) have ἡ γυνή here may suggest that it is a motivated reading, conforming this verse to the rest of the pericope. Although a decision is difficult, it is probably best to regard the shorter reading as authentic. NA²⁷ has ἡ γυνή in brackets, indicating doubts as to their authenticity. For English stylistic reasons, the translation also includes “the woman” here.

³ tc D 091 a e sy^s,c lack the phrase “after the Lord had given thanks” (εὐχαριστήσαντος τοῦ κυρίου, eucaristhsanto" tou kuriou), while almost all the rest of the witnesses ({Ì⁷⁵ א A B L W Θ Ψ 0141 [˹] ˹³ 33 Ï as well as several versions and fathers}) have the words (though {l672 l950 sy^p pbo} read ᾿Ιησοῦ [Ihsou, “Jesus”] instead of κυρίου). Although the shorter reading has minimal support, it is significant that this Gospel speaks of Jesus as Lord in the evangelist’s narrative descriptions only in 11:2; 20:18, 20; 21:12; and possibly 4:1 (but see tc note on “Jesus” there). There is thus but one undisputed preresurrection text in which the narrator calls Jesus “Lord.” This fact can be utilized on behalf of either reading: The participial phrase could be seen as a scribal addition harking back to 6:11 but which does not fit Johannine style, or it could be viewed as truly authentic and in line with what John indisputably does elsewhere even if rarely. On balance, in light of the overwhelming support for these words it is probably best to retain them in the text.

² tc A few witnesses lack με (me, “me”; א A a b e q sy^s,c), while the rest of the tradition has the word (Ì^66,75vid rell). It is possible that the mss that lack the pronoun preserve the original wording here, with the rest of the witnesses adding the pronoun for clarity’s sake. This likelihood increases since the object is not required in Greek. Without it, however, ambiguity increases: The referent could be “me” or it could be “signs,” reaching back to vv. 26 and 30. However, the oblique form of ἐγώ (egw, the first person personal pronoun) occurs some two dozen times in this chapter alone, yet it vacillates between the emphatic form and the unemphatic form. Although generally the unemphatic form is used with verbs, there are several exceptions to this in John (cf. 8:12; 12:26, 45, 48; 13:20; 14:9). If the pronoun is a later addition here, one wonders why it is so consistently the unemphatic form in the mss. Further, that two unrelated Greek witnesses lack this small word could easily be due to accidental deletion. Finally, the date and diversity of the witnesses for the pronoun are so weighty that it is likely to be authentic and should thus be retained in the text.

¹ tc ‡ The earliest witnesses, as well as a few other important mss (א B Ψ 892* 2427 sy^s), lack the rest of the quotation from Gen 2:24, “and will be united with his wife.” Most mss ([A C] D [L N] W [Δ] Θ Ë^[1],13 [579] Ï lat co) have the clause. It could be argued that the shorter reading was an accidental omission, due to this clause and v. 8 both beginning with καί (kai, “and”). But if that were the case, one might expect to see corrections in א or B. This can be overstated, of course; both mss combine in their errors on several other occasions. However, the nature of the omission here (both its length and the fact that it is from the OT) argues that א and B reflect the original wording. Further, the form of the longer reading is identical with the LXX of Gen 2:24, but different from the quotation in Matt 19:5 (προσκολληθήσεται vs. κολληθήσεται [proskollhqhsetai vs. kollhqhsetai], πρὸς τὴν γυναῖκα vs. τῇ γυναικί [pro" thn gunaika vs. th gunaiki]). The significance of this is that Matthew’s quotations of the OT are often, if not usually, directly from the Hebrew – except when he is following Mark’s quotation of the OT. Matthew in fact only departs from Mark’s verbatim quotation of the LXX in 15:4 and 19:19, both texts quoting from Exod 20:12/Deut 5:6 (and in both places the only difference from Mark/LXX is the dropping of σου [sou, “your”]). This might suggest that the longer reading here was not part of what the first evangelist had in his copy of Mark. Further, the reading without this line is harder, for the wife is not explicitly mentioned in v. 7; the casual reader could read “the two” of v. 8 as referring to father and mother rather than husband and wife. (And Mark is known for having harder, shorter readings that scribes tried to soften by explanatory expansion: In this chapter alone, cf. the textual problems in v. 6 [the insertion of ὁ θεός]; in v. 13 [the replacement of αὐτοῖς with τοῖς προσφέρουσιν or τοῖς φέρουσιν]; in v. 24 [insertion of ἐστιν τοὺς πεποιθότας ἐπὶ χρήμασιν, πλούσιον, or τὰ χρήματα ἔχοντες; and perhaps in v. 2 [possible insertion of προσελθόντες Φαρισαῖοι or similar permutations].) Although a decision is difficult, the preferred reading lacks “and will be united with his wife.” NA²⁷ has the longer reading in brackets, indicating doubts as to its authenticity.

¹ tn A decision about the proper translation of this Greek phrase (ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως, ejpi Abiaqar ajrcierew") is very difficult for a number of reasons. The most natural translation of the phrase is “when Abiathar was high priest,” but this is problematic because Abiathar was not the high priest when David entered the temple and ate the sacred bread; Ahimelech is the priest mentioned in 1 Sam 21:1-7. Three main solutions have been suggested to resolve this difficulty. (1) There are alternate readings in various manuscripts, but these are not likely to be original: D W {271} it sy^s and a few others omit ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως, no doubt in conformity to the parallels in Matt 12:4 and Luke 6:4; {A C Θ Π Σ Φ 074 ˹³ and many others} add τοῦ before ἀρχιερέως, giving the meaning “in the days of Abiathar the high priest,” suggesting a more general time frame. Neither reading has significant external support and both most likely are motivated by the difficulty of the original reading. (2) Many scholars have hypothesized that one of the three individuals who would have been involved in the transmission of the statement (Jesus who uttered it originally, Mark who wrote it down in the Gospel, or Peter who served as Mark’s source) was either wrong about Abiathar or intentionally loose with the biblical data in order to make a point. (3) It is possible that what is currently understood to be the most natural reading of the text is in fact not correct. (a) There are very few biblical parallels to this grammatical construction (ἐπί + genitive proper noun, followed by an anarthrous common noun), so it is possible that an extensive search for this construction in nonbiblical literature would prove that the meaning does involve a wide time frame. If this is so, “in the days of Abiathar the high priest” would be a viable option. (b) It is also possible that this phrasing serves as a loose way to cite a scripture passage. There is a parallel to this construction in Mark 12:26: “Have you not read in the book of Moses, in the passage about the bush?” Here the final phrase is simply ἐπὶ τοῦ βάτου (ejpi tou batou), but the obvious function of the phrase is to point to a specific passage within the larger section of scripture. Deciding upon a translation here is difficult. The translation above has followed the current consensus on the most natural and probable meaning of the phrase ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως: “when Abiathar was high priest.” It should be recognized, however, that this translation is tentative because the current state of knowledge about the meaning of this grammatical construction is incomplete, and any decision about the meaning of this text is open to future revision.

¹ tc The Gospel of Mark ends at this point in some witnesses (א B 304 sy^s sa^ms arm^mss Eus Eus^mss Hier^mss), including two of the most respected mss (א B). The following shorter ending is found in some mss: “They reported briefly to those around Peter all that they had been commanded. After these things Jesus himself sent out through them, from the east to the west, the holy and imperishable preaching of eternal salvation. Amen.” This shorter ending is usually included with the longer ending (L Ψ 083 099 0112 579 al); k, however, ends at this point. Most mss include the longer ending (vv. 9-20) immediately after v. 8 (A C D W [which has a different shorter ending between vv. 14 and 15] Θ Ë¹³ 33 2427 Ï lat sy^c,p,h bo); however, Jerome and Eusebius knew of almost no Greek mss that had this ending. Several mss have marginal comments noting that earlier Greek mss lacked the verses, while others mark the text with asterisks or obeli (symbols that scribes used to indicate that the portion of text being copied was spurious). Internal evidence strongly suggests the secondary nature of both the short and the long endings. Their vocabulary and style are decidedly non-Markan (for further details, see TCGNT 102-6). All of this evidence strongly suggests that as time went on scribes added the longer ending, either for the richness of its material or because of the abruptness of the ending at v. 8. (Indeed, the strange variety of dissimilar endings attests to the probability that early copyists had a copy of Mark that ended at v. 8, and they filled out the text with what seemed to be an appropriate conclusion. All of the witnesses for alternative endings to vv. 9-20 thus indirectly confirm the Gospel as ending at v. 8.) Because of such problems regarding the authenticity of these alternative endings, 16:8 is usually regarded as the last verse of the Gospel of Mark. There are three possible explanations for Mark ending at 16:8: (1) The author intentionally ended the Gospel here in an open-ended fashion; (2) the Gospel was never finished; or (3) the last leaf of the ms was lost prior to copying. This first explanation is the most likely due to several factors, including (a) the probability that the Gospel was originally written on a scroll rather than a codex (only on a codex would the last leaf get lost prior to copying); (b) the unlikelihood of the ms not being completed; and (c) the literary power of ending the Gospel so abruptly that the readers are now drawn into the story itself. E. Best aptly states, “It is in keeping with other parts of his Gospel that Mark should not give an explicit account of a conclusion where this is already well known to his readers” (Mark, 73; note also his discussion of the ending of this Gospel on 132 and elsewhere). The readers must now ask themselves, “What will I do with Jesus? If I do not accept him in his suffering, I will not see him in his glory.”

¹ tc ‡ What did John the Baptist declare about Jesus on this occasion? Did he say, “This is the Son of God” (οὗτός ἐστιν ὁ υἱὸς τοῦ θεοῦ, |outo" estin Jo Juio" tou qeou), or “This is the Chosen One of God” (οὗτός ἐστιν ὁ ἐκλεκτὸς τοῦ θεοῦ, outo" estin Jo eklekto" tou qeou)? The majority of the witnesses, impressive because of their diversity in age and locales, read “This is the Son of God” (so {Ì^66,75 A B C L Θ Ψ 0233^vid Ë^1,13 33 1241 aur c f l g bo as well as the majority of Byzantine minuscules and many others}). Most scholars take this to be sufficient evidence to regard the issue as settled without much of a need to reflect on internal evidence. On the other hand, one of the earliest mss for this verse, {Ì⁵} (3rd century), evidently read οὗτός ἐστιν ὁ ἐκλεκτὸς τοῦ θεοῦ. (There is a gap in the ms at the point of the disputed words; it is too large for υἱός especially if written, as it surely would have been, as a nomen sacrum [uMs]. The term ἐκλεκτός was not a nomen sacrum and would have therefore taken up much more space [eklektos]. Given these two variants, there is hardly any question as to what Ì⁵ read.) This papyrus has many affinities with א*, which here also has ὁ ἐκλεκτός. In addition to their combined testimony Ì^106vid b e ff^2* sy^s,c also support this reading. ̹⁰⁶ is particularly impressive, for it is a second third-century papyrus in support of ὁ ἐκλεκτός. A third reading combines these two: “the elect Son” (electus filius in ff^2c sa and a [with slight variation]). Although the evidence for ἐκλεκτός is not as impressive as that for υἱός, the reading is found in early Alexandrian and Western witnesses. Turning to the internal evidence, “the Chosen One” clearly comes out ahead. “Son of God” is a favorite expression of the author (cf. 1:49; 3:18; 5:25; 10:36; 11:4, 27; 19:7; 20:31); further, there are several other references to “his Son,” “the Son,” etc. Scribes would be naturally motivated to change ἐκλεκτός to υἱός since the latter is both a Johannine expression and is, on the surface, richer theologically in 1:34. On the other hand, there is not a sufficient reason for scribes to change υἱός to ἐκλεκτός. The term never occurs in John; even its verbal cognate (ἐκλέγω, eklegw) is never affirmed of Jesus in this Gospel. ἐκλεκτός clearly best explains the rise of υἱός. Further, the third reading (“Chosen Son of God”) is patently a conflation of the other two. It has all the earmarks of adding υἱός to ἐκλεκτός. Thus, ὁ υἱός τοῦ θεοῦ is almost certainly a motivated reading. As R. E. Brown notes (John [AB], 1:57), “On the basis of theological tendency…it is difficult to imagine that Christian scribes would change ‘the Son of God’ to ‘God’s chosen one,’ while a change in the opposite direction would be quite plausible. Harmonization with the Synoptic accounts of the baptism (‘You are [This is] my beloved Son’) would also explain the introduction of ‘the Son of God’ into John; the same phenomenon occurs in vi 69. Despite the weaker textual evidence, therefore, it seems best – with Lagrange, Barrett, Boismard, and others – to accept ‘God’s chosen one’ as original.”

³ tc Most witnesses, including a few important ones (A^[*^] Θ Ψ 050 Ë^1,13 Ï latt sy^c,p,h), have at the end of this verse “the one who is in heaven” (ὁ ὢν ἐν τῷ οὐρανῷ, Jo wn en tw ouranw). A few others have variations on this phrase, such as “who was in heaven” (e sy^c), or “the one who is from heaven” (0141 pc sy^s). The witnesses normally considered the best, along with several others, lack the phrase in its entirety (Ì^66,75 א B L T W^s 083 086 33 1241 pc co). On the one hand, if the reading ὁ ὢν ἐν τῷ οὐρανῷ is authentic it may suggest that while Jesus was speaking to Nicodemus he spoke of himself as in heaven even while he was on earth. If that is the case, one could see why variations from this hard saying arose: “who was in heaven,” “the one who is from heaven,” and omission of the clause. At the same time, such a saying could be interpreted (though with difficulty) as part of the narrator’s comments rather than Jesus’ statement to Nicodemus, alleviating the problem. And if v. 13 was viewed in early times as the evangelist’s statement, “the one who is in heaven” could have crept into the text through a marginal note. Other internal evidence suggests that this saying may be authentic. The adjectival participle, ὁ ὤν, is used in the Fourth Gospel more than any other NT book (though the Apocalypse comes in a close second), and frequently with reference to Jesus (1:18; 6:46; 8:47). It may be looking back to the LXX of Exod 3:14 (ἐγώ εἰμι ὁ ὤν). Especially since this exact construction is not necessary to communicate the location of the Son of Man, its presence in many witnesses here may suggest authenticity. Further, John uses the singular of οὐρανός (ourano", “heaven”) in all 18 instances of the word in this Gospel, and all but twice with the article (only 1:32 and 6:58 are anarthrous, and even in the latter there is significant testimony to the article). At the same time, the witnesses that lack this clause are very weighty and must not be discounted. Generally speaking, if other factors are equal, the reading of such mss should be preferred. And internally, it could be argued that ὁ ὤν is the most concise way to speak of the Son of Man in heaven at that time (without the participle the point would be more ambiguous). Further, the articular singular οὐρανός is already used twice in this verse, thus sufficiently prompting scribes to add the same in the longer reading. This combination of factors suggests that ὁ ὢν ἐν τῷ οὐρανῷ is not a genuine Johannism. Further intrinsic evidence against the longer reading relates to the evangelist’s purposes: If he intended v. 13 to be his own comments rather than Jesus’ statement, his switch back to Jesus’ words in v. 14 (for the lifting up of the Son of Man is still seen as in the future) seems inexplicable. The reading “who is in heaven” thus seems to be too hard. All things considered, as intriguing as the longer reading is, it seems almost surely to have been a marginal gloss added inadvertently to the text in the process of transmission. For an argument in favor of the longer reading, see David Alan Black, “The Text of John 3:13,” GTJ 6 (1985): 49-66.