³ sn Many contemporary scholars see this story as fictitious, composed by the Jews during the captivity. According to this interpretation, the spoils of war here indicate the wealth of the Jews in captivity, which was to be given to the Levites and priests for the restoration of the sanctuary in Jerusalem. The conclusion drawn from this interpretation is that returning Jews had the same problem as the earlier ones: to gain a foothold in the land. Against this interpretation of the account is a lack of hard evidence, a lack which makes this interpretation appear contrived and subjective. If this was the intent of a later writer, he surely could have stated this more clearly than by making up such a story.

¹ sn This verse may seem out of place, since the report has already been given that they came to the desert. It begins to provide details of the event that the previous verse summarizes. The announcement in verse 6 may have come in advance by means of a messenger or at the time of arrival, either of which would fit with the attention to formal greetings in verse 7. This would suit a meeting between two important men; the status of Moses has changed. The LXX solves the problem by taking the pronoun “I” as the particle “behold” and reads it this way: “one said to Moses, ‘Behold, your father-in-law has come….’”

³ tn The word עָרֹב (ʿarov) means “a mix” or “swarm.” It seems that some irritating kind of flying insect is involved. Ps 78:45 says that the Egyptians were eaten or devoured by them. Various suggestions have been made over the years: (1) it could refer to beasts or reptiles; (2) the Greek took it as the dog-fly, a vicious blood-sucking gadfly, more common in the spring than in the fall; (3) the ordinary house fly, which is a symbol of Egypt in Isa 7:18 (Hebrew זְבוּב, z^evuv); and (4) the beetle, which gnaws and bites plants, animals, and materials. The fly probably fits the details of this passage best; the plague would have greatly intensified a problem with flies that already existed.

³ tn The word rendered “fault” is the basic OT verb for “sin”—וְחָטָאת (v^ekhataʾt). The problem is that it is pointed as a perfect tense, feminine singular verb. Some other form of the verb would be expected, or a noun. But the basic word-group means “to err, sin, miss the mark, way, goal.” The word in this context seems to indicate that the people of Pharaoh—the slave masters—have failed to provide the straw. Hence: “fault” or “they failed.” But, as indicated, the line has difficult grammar, for it would literally translate: “and you [fem.] sin your people.” Many commentators (so GKC 206 §74.g) wish to emend the text to read with the Greek and the Syriac, thus: “you sin against your own people” (meaning the Israelites are his loyal subjects).

³ sn S. R. Driver (Exodus, 29) suggests that the term “Levite” may refer to a profession rather than ancestry here because both Moses and Aaron were from the tribe of Levi and there would be little point in noting that ancestry for Aaron. In thinking through the difficult problem of the identity of Levites, he cites McNeile as saying “the Levite” referred to one who had had official training as a priest (cf. Judg 17:7, where a member of the tribe of Judah was a Levite). If it was the duty of the priest to give “torah”—to teach—then some training in the power of language would have been in order.

⁶ tn The two expressions are כְבַד־פֶּה (kh^evad peh, “heavy of mouth”), and then כְבַד לָשׁוֹן (kh^evad lashon, “heavy of tongue”). Both use genitives of specification, the mouth and the tongue being what are heavy—slow. “Mouth” and “tongue” are metonymies of cause. Moses is saying that he has a problem speaking well. Perhaps he had been too long at the other side of the desert, or perhaps he was being a little dishonest. At any rate, he has still not captured the meaning of God’s presence. See among other works, J. H. Tigay, “‘Heavy of Mouth’ and ‘Heavy of Tongue’: On Moses’ Speech Difficulty,” BASOR 231 (1978): 57-67.

⁴ sn During this period of Egyptian history the royal palaces were in the northern or Delta area of Egypt, rather than up the Nile as in later periods. The proximity of the royal residences to the Israelites makes this and the plague narratives all the more realistic. Such direct contact would have been unlikely if Moses had had to travel up the Nile to meet with Pharaoh. In the Delta area things were closer. Here all the people would have had access to the tributaries of the Nile near where the royal family came, but the royal family probably had pavilions and hunting lodges in the area. See also N. Osborn, “Where on Earth Are We? Problems of Position and Movement in Space,” BT 31 (1980): 239-42.

² tn The Hebrew form שִׁילֹה (shiloh) is a major interpretive problem. There are at least four major options (with many variations and less likely alternatives): (1) Some prefer to leave the text as it is, reading “Shiloh” and understanding it as the place where the ark rested for a while in the time of the Judges. (2) By repointing the text others arrive at the translation “until the [or “his”] ruler comes,” a reference to a Davidic ruler or the Messiah. (3) Another possibility that does not require emendation of the consonantal text, but only repointing, is “until tribute is brought to him” (so NEB, JPS, NRSV), which has the advantage of providing good parallelism with the following line, “the nations will obey him.” (4) The interpretation followed in the present translation, “to whom it [belongs]” (so RSV, NIV, REB), is based on the ancient versions. Again, this would refer to the Davidic dynasty or, ultimately, to the Messiah.

² sn This account is parallel to two similar stories about Abraham (see Gen 12:10-20; 20:1-18). Many scholars do not believe there were three similar incidents, only one that got borrowed and duplicated. Many regard the account about Isaac as the original, which then was attached to the more important person, Abraham, with supernatural elements being added. For a critique of such an approach, see R. Alter, The Art of Biblical Narrative, 47-62. It is more likely that the story illustrates the proverb “like father, like son” (see T. W. Mann, The Book of the Torah, 53). In typical human fashion the son follows his father’s example of lying to avoid problems. The appearance of similar events reported in a similar way underscores the fact that the blessing has now passed to Isaac, even if he fails as his father did.

¹ tn The Greek verb μένω (menō) is commonly translated into contemporary English as “remain” or “abide,” but both of these translations have some problems: (1) “Abide” has become in some circles almost a “technical term” for some sort of special intimate fellowship or close relationship between the Christian and God, so that one may speak of Christians who are “abiding” and Christians who are not. It is accurate to say the word indicates a close, intimate (and permanent) relationship between the believer and God. However, it is very important to note that for the author of the Gospel of John and the Johannine Epistles every genuine Christian has this type of relationship with God, and the person who does not have this type of relationship (cf. 2 John 9) is not a believer at all (in spite of what he or she may claim). (2) On the other hand, to translate μένω as “remain” removes some of these problems, but creates others: In certain contexts, such a translation can give the impression that those who currently “remain” in this relationship with God can at some point choose not to “remain”, that is, to abandon their faith and return to an unsaved condition. While one may easily think in terms of the author’s opponents in 1 John as not “remaining,” the author makes it inescapably clear in 2:19 that these people, in spite of their claims to know God and be in fellowship with God, never really were genuine believers. (3) In an attempt to avoid both these misconceptions, this translation renders μένω as “reside” except in cases where the context indicates that “remain” is a more accurate nuance, that is, in contexts where a specific change of status or movement from one position to another is in view.

² tc Several significant Greek mss (P⁷⁵ א¹ A B N T W 579 1071*) along with diverse and widespread versional witnesses lack 22:43-44. In addition, the verses are placed after Matt 26:39 by ƒ¹³. Floating texts typically suggest both spuriousness and early scribal impulses to regard the verses as historically authentic. These verses are included in א*^,2 D L Θ Ψ 0171 ƒ¹ M lat Ju Ir Hipp Eus. However, a number of mss mark the text with an asterisk or obelisk, indicating the scribe’s assessment of the verses as inauthentic. At the same time, these verses generally fit Luke’s style. Arguments can be given on both sides about whether scribes would tend to include or omit such comments about Jesus’ humanity and an angel’s help. But even if the verses are not literarily authentic, they are probably historically authentic. This is due to the fact that this text was well known in several different locales from a very early period. Since there are no synoptic parallels to this account and since there is no obvious reason for adding these words here, it is very likely that such verses recount a part of the actual suffering of our Lord. Nevertheless, because of the serious doubts as to these verses’ authenticity, they have been put in brackets. For an important discussion of this problem, see B. D. Ehrman and M. A. Plunkett, “The Angel and the Agony: The Textual Problem of Luke 22:43-44, ” CBQ 45 (1983): 401-16.

¹ sn In the English Bible Esther appears adjacent to Ezra-Nehemiah and with the historical books, but in the Hebrew Bible it is one of five short books (the “Scrolls,” Heb Megillot) that appear toward the end of the biblical writings. The canonicity of the book was questioned by some in ancient Judaism and early Christianity. It is one of five OT books that were at one time regarded as antilegomena (i.e., books “spoken against”). The problem with Esther was the absence of any direct mention of God. Some questioned whether a book that did not mention God could be considered sacred scripture. Attempts to resolve this by discovering the tetragrammaton (YHWH) encoded in the Hebrew text (e.g., in the initial letters of four consecutive words in the Hebrew text of Esth 5:4) are unconvincing, although they do illustrate how keenly the problem was felt by some. Martin Luther also questioned the canonicity of this book, objecting to certain parts of its content. Although no copy of Esther was found among the Dead Sea Scrolls, this does not necessarily mean that the Qumran community did not regard it as canonical. It is possible that the absence of Esther from what has survived at Qumran is merely a coincidence. Although the book does not directly mention God, it would be difficult to read it without sensing the providence of God working in powerful, though at times subtle, ways to rescue his people from danger and possible extermination. The absence of mention of the name of God may be a deliberate part of the literary strategy of the writer.

² sn There has been much discussion about the numbers in the Israelite wilderness experience. The immediate difficulty for even the casual reader is the enormous number of the population. If indeed there were 603,550 men twenty years of age and older who could fight, the total population of the exodus community counting women and children would have been well over a million, or even two million as calculated by some. This is not a figure that the Bible ever gives, but given the sizes of families the estimate would not be far off. This is a staggering number to have cross the Sea, drink from the oases, or assemble in the plain by Sinai. It is not a question of whether or not God could provide for such a number; it is rather a problem of logistics for a population of that size in that period of time. The problem is not with the text itself, but with the interpretation of the word אֶלֶף (’elef), traditionally translated “thousand.” The word certainly can be taken as “thousand,” and most often is. But in view of the problem of the large number here, some scholars have chosen one of the other meanings attested in literature for this word, perhaps “troop,” or “family,” or “tent group,” even though a word for “family” has already been used (see A. H. McNeile, Numbers, 7; J. Garstang, Joshua-Judges, 120; J. Bright, History of Israel, 144). Another suggestion is to take the word as a “chief” or “captain” based on Ugaritic usage (see R. E. D. Clarke, “The Large Numbers of the Old Testament,” JTVI 87 [1955]: 82-92; and J. W. Wenham, “Large Numbers in the Old Testament,” TynBul 18 [1967]: 19-53). This interpretation would reduce the size of the Israelite army to about 18,000 men from a population of about 72,000 people. That is a radical change from the traditional reading and may be too arbitrary an estimate. A more unlikely calculation following the idea of a new meaning would attempt to divide the numbers and use the first part to refer to the units and the second the measurement (e.g., 65 thousand and four hundred would become 65 units of four hundred). Another approach has been to study the numbers rhetorically, analyzing the numerical values of letters and words. But this method, known as gematria, came in much later than the biblical period (see for it G. Fohrer, Introduction to the Old Testament, 184; and A. Noordtzij, Numbers [BSC], 24). On this system the numbers for “the sons of Israel” would be 603. But the number of the people in the MT is 603,550. Another rhetorical approach is that which says the text used exaggerations in the numbers on an epic scale to make the point of God’s blessing. R. B. Allen’s view that the numbers have been magnified by a factor of ten (“Numbers,” The Expositor’s Bible Commentary, 2:688-91), which would mean the army was only 60,000 men, seems every bit as arbitrary as Wenham’s view to get down to 18,000. Moreover, such views cannot be harmonized with the instructions in the chapter for them to count every individual skull – that seems very clear. This is not the same kind of general expression one finds in “Saul has killed his thousands, David his ten thousands” (1 Sam 18:7). There one expects the bragging and the exaggerations. But in a text of numbering each male, to argue that the numbers have been inflated ten-fold to form the rhetoric of praise for the way God has blessed the nation demands a much more convincing argument than has typically been given. On the surface it seems satisfactory, but it raises a lot of questions. Everything in Exodus and Numbers attests to the fact that the Israelites were in a population explosion, that their numbers were greater than their Egyptian overlords. Pharaoh had attempted to counter their growth by killing males from the ranks. That only two midwives are named must be taken to mean that they were heads of the guilds, for two could not service a population – even of the smaller estimate given above. But even though the size had to have been great and seen as a threat, we are at a loss to know exactly how to determine it. There is clearly a problem with the word “thousand” here and in many places in the OT, as the literature will show, but the problem cannot really be solved without additional information. The suggestions proposed so far seem to be rather arbitrary attempts to reduce the number to a less-embarrassing total, one that would seem more workable in the light of contemporary populations and armies, as well as space and time for the people’s movement in the wilderness. An army of 10,000 or 20,000 men in those days would have been a large army; an army of 600,000 (albeit a people’s army, which may mean that only a portion of the males would actually fight at any time – as was true at Ai) is large even by today’s standards. But the count appears to have been literal, and the totals calculated accordingly, totals which match other passages in the text. If some formula is used to reduce the thousands in this army, then there is the problem of knowing what to do when a battle has only five thousand, or three thousand men. One can only conclude that on the basis of what we know the word should be left with the translation “thousand,” no matter what difficulties this might suggest to the reader. One should be cautious, though, in speaking of a population of two million, knowing that there are serious problems with the calculation of that number, if not with the word “thousand” itself. It is very doubtful that the population of the wilderness community was in the neighborhood of two million. Nevertheless, until a more convincing explanation of the word “thousand” or the calculation of the numbers is provided, one should retain the reading of the MT but note the difficulty with the large numbers.

³ tn The second ὅτι (hoti) in 5:9 may be understood in three different ways. (1) It may be causal, in which case it gives the reason why the testimony just mentioned is God’s testimony: “because he has testified concerning his Son.” This is extremely awkward because of the preceding ὅτι clause which is almost certainly causal (although the second ὅτι could perhaps be resumptive in force, continuing the first). (2) The second ὅτι could be understood as epexegetical (explanatory), in which case it explains what the testimony of God mentioned in the preceding clause consists of: “because this is the testimony of God, [namely,] that he has testified concerning his Son.” This is much smoother grammatically, but encounters the logical problem that “the testimony of God” is defined in 5:11 (“And this is the testimony: that God has given us eternal life”) and the two definitions of what the testimony of God consists of are not identical (some would say that they are not even close). Thus (3) the smoothest way to understand the second ὅτι logically is to read it as a relative pronoun: “because this is the testimony of God that he has testified concerning his Son.” In this case it is exactly parallel to the relative clause which occurs in 5:10b: “because he has not believed the testimony that (ἣν, hēn) God has testified concerning his Son.” (There is in fact a textual problem with the second ὅτι in 5:9: The Byzantine tradition, along with ms P, reads a relative pronoun [ἣν] in place of the second ὅτι in 5:9 identical to the relative pronoun in 5:10b. This represents an obvious effort on the part of scribes to smooth out the reading of the text.) In an effort to derive a similar sense from the second ὅτι in 5:9 it has been suggested that the conjunction ὅτι should be read as an indefinite relative pronoun ὅτι (sometimes written ὅ τι). The problem with this suggestion is the use of the neuter relative pronoun to refer to a feminine antecedent (ἡ μαρτυρία, hē marturia). It is not without precedent for a neuter relative pronoun to refer to an antecedent of differing gender, especially as some forms tended to become fixed in usage and were used without regard to agreement. But in this particular context it is difficult to see why the author would use a neuter indefinite relative pronoun here in 5:9b and then use the normal feminine relative pronoun (ἣν) in the next verse. (Perhaps this strains at the limits of even the notorious Johannine preference for stylistic variation, although it is impossible to say what the author might or might not have been capable of doing.) Because of the simplicity and logical smoothness which results from reading ὅτι as equivalent to a relative pronoun, the third option is preferred, although it is not without its difficulties (as are all three options).

⁸ tc ‡ One of the most difficult textual problems in the NT is found in v. 10. The reading εὑρεθήσεται (heurethēsetai), which enjoys by far the best support (א B K P 0156^vid 323 1175 1241 1739^txt 1852) is nevertheless so difficult a reading that many scholars regard it as nonsensical. (NA²⁷ lists five conjectures by scholars, from Hort to Mayor, in this text. All conjectures were eliminated in the NA²⁸ apparatus) As R. Bauckham has pointed out, solutions to the problem are of three sorts: (1) conjectural emendation (which normally speaks more of the ingenuity of the scholar who makes the proposal than of the truth of the conjecture, e.g., changing one letter in the previous word, ἔργα [erga] becomes ἄργα [arga] with the meaning, “the earth and the things in it will be found useless”); (2) adoption of one of several variant readings (all of which, however, are easier than this one and simply cannot explain how this reading arose, e.g., the reading of P⁷² which adds λυόμενα [luomena] to the verb—a reading suggested no doubt by the threefold occurrence of this verb in the surrounding verses: “the earth and its works will be found dissolved”; or the simplest variant, the reading of the Sahidic mss and a couple of other ancient versions, οὐχ [ouch] preceding εὑρεθήσεται—“will not be found”); or (3) interpretive gymnastics which regards the text as settled but has to do some manipulation to its normal meaning. Bauckham puts forth an excellent case that the third option is to be preferred and that the meaning of the term is virtually the equivalent of “will be disclosed,” “will be manifested.” (That this meaning is not readily apparent may in fact have been the reason for so many variants and conjectures.) Thus, the force of the clause is that “the earth and the works [done by men] in it will be stripped bare [before God].” In addition, the unusualness of the expression is certainly in keeping with the author’s style throughout this little book. Hence, what looks to be suspect because of its abnormalities, upon closer inspection is actually in keeping with the author’s stylistic idiosyncrasies. The meaning of the text then is that all but the earth and mankind’s works will be destroyed. Everything will be removed so that humanity will stand naked before God. Textually, then, on both external and internal grounds, εὑρεθήσεται commends itself as the preferred reading.

³ tn Or “gave us when we believed”; or “gave us after we believed”; or “gave us who believed”; or “gave them when they believed the same gift as he also gave us.” The aorist dative plural participle πιστεύσασιν (pisteusasin) can be understood in several different ways: (1) It could modify ἡμῖν (hēmin, “us”) or αὐτοῖς (autois, “them”). Proximity (it immediately follows ἡμῖν) would suggest that it belongs with ἡμῖν, so the last option (“gave them when they believed the same gift he also gave us”) is less likely. (2) The participle could be either adverbial or adjectival, modifying ἡμῖν. This decision is primarily a contextual one. The point Peter made is not whether or not the Gentiles believed, since both groups (“us” and “they”) had believed in the Lord Jesus Christ. The point was whether or not the Gentiles received the Spirit when they believed, just as Jewish Christians had received the Spirit on the day of Pentecost when they believed. Translated as an adjectival participle, πιστεύσασιν only affirms the fact of belief, however, and raises somewhat of a theological problem if one realizes, “Would God have given the Gentiles the Spirit if they had not believed?” (In other words, belief in itself is a theological prerequisite for receiving the Spirit. As such, in the case of the Gentiles, it is assumed.) Thus in context it makes more sense to understand the participle πιστεύσασιν as adverbial, related to the time of belief in connection with the giving of the Spirit. (3) The participle πιστεύσασιν as a temporal participle can refer to action antecedent to the action of the main verb ἔδωκεν (edōken) or contemporaneous with it. Logically, at least, the gift of the Spirit followed belief in the case of the original Christians, who had believed before the day of Pentecost. In the case of Cornelius and his household, belief and the reception of the Spirit were virtually simultaneous. One can argue that Peter is “summarizing” the experience of Jewish Christians, and therefore the actions of belief and reception of the Spirit, while historically separate, have been “telescoped” into one (“gave them the same gift as he gave us when we believed”), but to be technically accurate the participle πιστεύσασιν should be translated “gave them the same gift as he also gave us after we believed.” A number of these problems can be avoided, however, by using a translation in English that maintains some of the ambiguity of the Greek original. Thus “if God gave them the same gift as he also gave us after believing” is used, where the phrase “after believing” can refer either to “them” or to “us,” or both.

³ tc Some mss (א [L] 33 it) read Bethzatha, while others read Bethsaida (P^[66],75 B T W^s [Ψ] vg); codex D has Belzetha. A lot of controversy has surrounded the name of the pool itself: The reading of the Byzantine (or majority) text (A C Θ 078 ƒ^1,13 M), Bethesda, has been virtually discarded by scholars in favor of what is thought to be the more primitive Bethzatha, even though many recent translations continue to employ Bethesda, the traditional reading. The latter is attested by Josephus as the name of a quarter of the city near the northeast corner of the temple area. He reports that the Syrian Legate Cestius burned this suburb in his attack on Jerusalem in October a.d. 68 (J. W. 2.19.4 [2.530]). However, there is some new archaeological evidence for this problem. 3Q15 (Copper Scroll) from Qumran seems to indicate that in the general area of the temple, on the eastern hill of Jerusalem, a treasure was buried in Bet ’Esdatayin, in the pool at the entrance to the smaller basin. The name of the region or pool itself seems then to have been Bet ’Esda, “house of the flowing.” It appears with the dual ending in the scroll because there were two basins. Bethesda seems to be an accurate Greek rendition of the name, while J. T. Milik suggests Bethzatha is a rendition of the Aramaic intensive plural Bet ’Esdata (DJDJ 3, 271). As for the text of John 5:2, a fundamental problem with the Bethesda reading is that it looks motivated (with an edifying Semitic etymology, meaning “House of Mercy” [TCGNT 178]). Also, apart from the Copper Scroll, the evidence for Bethesda is almost entirely shut up to the Byzantine text (C being the most notable exception, but it often has Byzantine encroachments). On the one hand, this argues the Byzantine reading here had ancient, semitic roots; on the other hand, since both readings are attested as historically accurate, a decision has to be based on the better witnesses. The fact that there are multiple readings here suggests that the original was not well understood. Which reading best explains the rise of the others? It seems that Bethzatha is the best choice.

³ tc There is a major punctuation problem here: Should this relative clause go with v. 3 or v. 4? The earliest mss have no punctuation (P^66,75* א* A B Δ al). Many of the later mss which do have punctuation place it before the phrase, thus putting it with v. 4 (P^75c C D L W^s 050* pc). NA²⁵ placed the phrase in v. 3; NA²⁶ moved the words to the beginning of v. 4. In a detailed article K. Aland defended the change (“Eine Untersuchung zu Johannes 1, 3-4. Über die Bedeutung eines Punktes,” ZNW 59 [1968]: 174-209). He sought to prove that the attribution of ὃ γέγονεν (ho gegonen) to v. 3 began to be carried out in the 4th century in the Greek church. This came out of the Arian controversy, and was intended as a safeguard for doctrine. The change was unknown in the West. Aland is probably correct in affirming that the phrase was attached to v. 4 by the Gnostics and the Eastern Church; only when the Arians began to use the phrase was it attached to v. 3. But this does not rule out the possibility that, by moving the words from v. 4 to v. 3, one is restoring the original reading. Understanding the words as part of v. 3 is natural and adds to the emphasis which is built up there, while it also gives a terse, forceful statement in v. 4. On the other hand, taking the phrase ὃ γέγονεν with v. 4 gives a complicated expression: C. K. Barrett says that both ways of understanding v. 4 with ὃ γέγονεν included “are almost impossibly clumsy” (St. John, 157): “That which came into being—in it the Word was life”; “That which came into being—in the Word was its life.” The following stylistic points should be noted in the solution of this problem: (1) John frequently starts sentences with ἐν (en); (2) he repeats frequently (“nothing was created that has been created”); (3) 5:26 and 6:53 both give a sense similar to v. 4 if it is understood without the phrase; (4) it makes far better Johannine sense to say that in the Word was life than to say that the created universe (what was made, ὃ γέγονεν) was life in him. In conclusion, the phrase is best taken with v. 3. Schnackenburg, Barrett, Carson, Haenchen, Morris, KJV, and NIV concur (against Brown, Beasley-Murray, and NEB). The arguments of R. Schnackenburg, St. John, 1:239-40, are particularly persuasive.

¹ tc The Textus Receptus, on which the KJV rests, reads “the book” of life (ἀπὸ βίβλου, apo biblou) instead of “the tree” of life. When the Dutch humanist Desiderius Erasmus translated the NT he had access to no Greek mss for the last six verses of Revelation. So he translated the Latin Vulgate back into Greek at this point. As a result he created seventeen textual variants which were not in any Greek mss. The most notorious of these is this reading. It is thus decidedly inauthentic, while “the tree” of life, found in the best and virtually all Greek mss, is clearly authentic. The confusion was most likely due to an intra-Latin switch: The form of the word for “tree” in Latin in this passage is ligno; the word for “book” is libro. The two-letter difference accounts for an accidental alteration in some Latin mss; that “book of life” as well as “tree of life” is a common expression in the Apocalypse probably accounts for why this was not noticed by Erasmus or the KJV translators. (This textual problem is not discussed in NA²⁷.)

⁴ tc There are several problems in this verse. It seems that according to the ms evidence the first two phrases (i.e., “and a haunt for every unclean spirit, and a haunt for every unclean bird” [καὶ φυλακὴ παντὸς πνεύματος ἀκαθάρτου καὶ φυλακὴ παντὸς ὀρνέου ἀκαθάρτου, kai phulakē pantos pneumatos akathartou kai phulakē pantos orneou akathartou]) are to be regarded as authentic, though there are some ms discrepancies. The similar beginnings (καὶ φυλακὴ παντός) and endings (ἀκαθάρτου) of each phrase would easily account for some mss omitting one or the other phrase. The third phrase (“a haunt for every unclean animal” [καὶ φυλακὴ παντὸς θηρίου ἀκαθάρτου, kai phulakē pantos thēriou akathartou]), however, is more problematic since it is missing in several significant mss (א C 051 M). The passage as a whole, including the third phrase, seems to be an allusion to Isa 13:21 and 34:11. It seems reasonable, in such a case, to assume that since there is good ms evidence to support the third phrase (A 1611 2329 al), it probably dropped out of certain mss because of its similarity to the two preceding clauses. It is the presence of all three phrases in the initial text that most likely gave rise to the divergent ms evidence extant today.