Returning to answer more questions in his interview on assorted text-critical issues, here is Nelson Hsieh, Assistant Professor of New Testament at New Orleans Baptist Theological Seminary:
Nelson Hsieh
Q: Nelson, when did the archetype of family 35 first appear, more or less?
Hsieh: Wilbur Pickering’s argued that Family 35 (von Soden’s Kr group) is “ancient” (going back to the first few centuries after the New Testament), “independent” of von Soden’s Kx group (the dominant Byzantine tradition), and “non-recensional,” but I remain unconvinced in light of the (fairly) objective evidence: among the 231 manuscripts identified as Kr, these manuscripts:
exhibit over 98% textual agreement
are consistent in having an elaborate lectionary apparatus (including an explainer in 28 manuscripts entitled, Δήλωσις πῶς δεῖ εὑρίσκειν τὰ καθημέραν εὐαγγέλια: “explanation of how one must find the daily gospel readings”)
omit the Eusebian canon tables (except in five manuscripts)
place obelii next to the Pericope of the Adulteress
generally show late-Byzantine calligraphic hands
have few itacistic spelling errors
most are written on parchment during a time when paper was in common use [See Georgi Parpulov, “Kr in the Gospels,” in The New Testament in Antiquity and Byzantium, ed. David Parker and Holger Strutwolf, ANTF 52 (de Gruyter, 2019), 203–13.]
among manuscripts that can be securely dated, no Kr manuscripts exist from prior to the fourteenth century (prior to 1304 to be exact). [Parpulov also redates minuscule 35 to fourteenth century (as opposed to the eleventh century); see Parpulov, “Kr in the Gospels,” 212-13.]
Georgi Parpulov favors a late-thirteenth / early-fourteenth century origin for Kr, in contrast to von Soden, who put the origin of Kr in the eleventh century, although he expressed this with uncertainty. [See Hermann von Soden, Die Schriften des Neuen Testaments, I. Teil: Untersuchungen, II: Abteilung: die Textformen, A. die Evangelien, 2nd ed. (Vandenhoeck & Ruprecht, 1911), 763.] The uniformity of these palaeographic, paratextual, and textual features suggest a recension (i.e., a deliberate revision).
However Pickering’s protest is fair: “If anyone wishes to claim that [Family 35] is [recensional], I request that they state who did it, when and where, and that they furnish evidence in support of the claim.” [Wilbur Pickering, God Has Preserved His Text! The Divine Preservation of the New Testament, 6th ed., 2024, https://www.family35nt.org/en/God-Has-Preserved-His-Text-v6.pdf. Page 338. Emphasis added.] But Pickering is asking for a level of evidence that doesn’t exist for any textual grouping of New Testament manuscripts. The argument that Family 35 is recensional is based on palaeography, paratext, and extremely high textual agreement.
External evidence is slim, but there are some hints towards possible origins of Family 35:
(1) the provenance of Kr manuscripts tends to be in and around Constantinople and Greece (especially Mount Athos). Thus, Parpulov rejects von Soden’s origin of Kr exclusively in Constantinople and suggests that Kr was circulated through a network of late Byzantine scribes and their patrons in and around Constantinople and Greece.
(2) Kr manuscripts have a good number of named scribes, who added various poems and metrical epigrams that show up in a few other Kr manuscripts, but not the whole Kr tradition, which suggests private and personal contact with some other scribes. This, then, also goes against von Soden’s suggestion that ecclesiastical authorities produced and disseminated Kr . [ Von Soden, Die Schriften des Neuen Testaments, Vol. I.2, 805.] All of these named scribes contributed to the growth and distribution of Kr manuscripts, although none of them alone would have had the authority to unilaterally impose such a revision upon others.
(3) Parpulov suggests that Kr probably was not based on a single exemplar/archetype; if such were the case, we would expect at least one scribal colophon (among 231 Kr manuscripts) to make such a claim. We would expect something similar to the so-called “Jerusalem colophon” found in some middle-Byzantine manuscripts; the colophon claims that its manuscript was “copied and corrected from ancient copies from Jerusalem, which were preserved on the holy mountain.” [See Tommy Wasserman, “The Greek New Testament Manuscripts in Sweden with an Excursus on the Jerusalem Colophon,” Svensk Exegetisk Årsbok 75 (2010): 77–107.]
(4) Probably the best guess (made by both Parpulov and von Soden) is that there was some sort of lectionary influence upon Kr manuscripts since the extensive lectionary material is the most distinctive paratextual feature of Kr manuscripts (even having an explainer of the lectionary system entitled, Δήλωσις πῶς δεῖ εὑρίσκειν τὰ καθημέραν εὐαγγέλια: “explanation on how one must find the daily gospel readings”). However, since we do not have a critical edition of the Byzantine Gospel lectionaries, it is not yet possible to understand the nature and extent of the influence of the lectionaries upon Kr manuscripts.
Thus, while we cannot satisfy Pickering with a specific person, place, and date for the creation of Kr – yet there are enough hints to say the following: Kr arose among a private network of several named scribes operating in and around Constantinople and Greece during the late-thirteenth and early-fourteenth century. Put negatively: Kr probably does not derive from a single/authoritative archetype, thus probably no ecclesiastical authority created or disseminated the Kr revision. Unless further external evidence is found (such as a new scribal colophon), our best guess is that the Byzantine Gospel lectionary influenced the text of Kr manuscripts.
Q: What are your thoughts about Snapp’s canon “Prefer the less specific reading”?
Hsieh: According to Snapp, one of his principles of equitable eclecticism is: “A variant which involves a rare, obscure, or ambiguous term or expression is more likely to be original than a rival variant which involves an ordinary or specific term or expression..” [James Snapp, New Testament Textual Analysis (Independently Published, 2023), 403-4.] Elsewhere in the same work, Snapp discusses “thoughtful/intentional changes,” which might be motivated by “a desire to augment/clarify the meaning of the original text … [therefore], a non-specific reference might be exchanged for a more specific reference.” [Snapp, New Testament Textual Analysis, 15-16.] Elsewhere, Snapp says that there was “a tendency to augment the specificity of Old Testament references. … copyists did not try to make a text more difficult – but sometimes they did, when attempting to make it clearer or more specific.” [Snapp, New Testament Textual Analysis, 323-24.]
Scattered throughout his work, Snapp provides examples of this tendency:
Matthew 1:22-23 describes Jesus’s birth as happening, “so that the word from the Lord through the prophet might be fulfilled,” then quotes from Isaiah 7:14. NA28’s text has no mention of Isaiah; Codex Bezae, several versions, and Irenaeus (in part) add “Isaiah.” According to Snapp, “this makes the text more specific.” [According to NA28, the following added “Isaiah”: 05, it (= all or a majority of Old Latin witnesses), Syriac (Sinaitic, Harklean, and Curetonian with minor variation), one Sahidic manuscript, and Irenaeus in Latin (in part). Irenaeus quotes from Matthew 1:22 three times in the Latin edition of his Against Heresies (3.16.2; 3.21.4; 4.23.1); Isaiah is added only in the second instance (Adv. haer. 3.21.4). This is not absolutely clear evidence that Irenaeus’s text of Matthew had “Isaiah”; perhaps Irenaeus himself accidentally added it. See Adelin Rousseau and Louis Doutreleau, eds., Irénée de Lyon: Contre les Hérésies, Livre III, SC 211 (Les Éditions du Cerf, 1974), 292, 410; Adelin Rousseau, ed., Irénée de Lyon: Contre les Hérésies, Livre IV, SC 100 (Les Éditions du Cerf, 1965), 692.]
Matthew 2:5-6 quotes from Micah 5:2 and in NA28, the quotation is introduced with: “as it is written through the prophet.” Codex Sinaiticus has a marginal notation incorrectly identifying this quotation as from Isaiah. The Old Latin Codex Vercellensis (VL 3), dated to the second half of the fourth century, also incorrectly identifies the quotation from Micah 5:2 by adding “Isaiah” into the main text.
Matthew 2:15 quotes from Hosea 11:1 and in NA28, the quotation is introduced with: “so that the word from the Lord through the prophet might be fulfilled.” In Codex Sinaticus, there is a marginal notation which reads: εν αριθμοις (“in Numbers”), referring to the Old Testament book of Numbers. Matthew’s quotation matches the Hebrew Masoretic text of Hosea 11:1 and differs from the Septuagint (Matt 2:15 reads ἐκάλεσα τὸν υἱόν μου, while LXX Göttingen reads μετεκάλεσα τὰ τέκνα αὐτοῦ). The scribe of Codex Sinaiticus is not entirely wrong; NA28 notes that Matt 2:15 could also be an allusion to Numbers 23:22 (“God brings them out of Egypt”) or Numbers 24:8 (“God brings him out of Egypt”). [There are many other marginal notes in Codex Sinaiticus which mark OT quotations: Matt 1:23 quoting Isaiah 7:14 has the marginal note ησαιου (“from Isaiah”); Matt 2:6 quoting Micah 5:1 incorrectly has ησαιου (“from Isaiah”); Matt 2:15 quoting Hos 11:1 has εν αριθμοις (“in Numbers”); Rom 3:4 quoting Ps 50 has εν ψαλμω Ν (“in Psalm 50”); Rom 3:10-18 quoting Psalms 13 and 52 has ψαλμω IΓ κ(αι) NB (“in Psalm 13 and 52); Rom 4:3 quoting Gen 15:6 has γενεσεως (“from Genesis”); Rom 4:7-8 quoting Ps 31 has ψαλμω ΛΑ (“in Psalm 31”); Rom 4:17 quoting Gen 17:5 has γενεσεως (“from Genesis”); Rom 4:18 quoting Gen 15:5 γενεσεως (“from Genesis”); Acts 2:34-35 quoting Ps 109 has ψαλμω ΡΘ (“in Psalm 109”); Acts 3:22 quoting Deut 18 has δευτ (“(from) Deuteronomy”); Acts 3:25 quoting Gen 22 incorrectly has δευτερ (“(from) Deuteronomy”); Acts 4:25-26 quoting Ps 2 has ψαλμ B (“(in) Psalm 2”); Acts 7:49-50 quoting Isaiah 66 has ησαιας (“Isaiah”); Acts 13:33 quoting Ps 2 has ψαλμω B (“in Psalm 2”); Acts 13:35 quoting Ps 15 has ψαλμω (“in a psalm”); Acts 13:41 quoting Hab 1:5 incorrectly has ιωηλ (“Joel”); Acts 15:16-18 quoting Amos 9 has αμως (“(from) Amos”); 1 Pet 2:22 quoting Isa 53 has ησαιας (“Isaiah”); and 1 Pet 3:10 quoting Ps 33 has ψαλ ΛΓ (“(in) Psalm 33”). The scribe errs four times in Matt 2:5; 2:15; Acts 3:25; 13:41 (or three times if we allow Matt 2:15 to be an allusion to Numbers). For further discussion of these, see Ulrich Schmid, “Diplés und Quellenangaben im Codex Sinaiticus,” in Von der Septuaginta zum Neuen Testament: Textgeschichtliche Erörterungen, ed. Martin Karrer, Siegfried Kreuzer, and Marcus Sigismund, ANTF 43 (de Gruyter, 2010), 85-87.]
Matthew 13:35 quotes from Psalm 78; in NA28, the quotation is introduced with: “so that the word through the prophet might be fulfilled.” However, the original scribe of Sinaiticus, 038, families 1 and 13, minuscule 33, and manuscripts known to Jerome add “Isaiah” (see discussion of this variant in question #2 above). According to Snapp, once again, scribes “make the text more specific, identifying the prophet incorrectly as Isaiah.” [Snapp, New Testament Textual Analysis, 192. Bold text original.]
Snapp also says: “Isaiah’s name was probably added in Matthew 13:35 in some early copies by Western copyists who liked their quotations to be specific.” [Snapp, New Testament Textual Analysis, 324.]
Matthew 21:4-5 quotes from Zechariah 9:9; in NA28, the quotation is introduced with: “so that the word through the prophet might be fulfilled.” Bruce Metzger identifies a marginal note in 021, minuscule 42, several versional witnesses (Old Latin a, c, h; some Boharic manuscripts), and Hilary as adding “Zechariah.” Metzger also identifies versional witnesses (four Vulgate manuscripts, some Boharic manuscripts, and the Ethiopic) as incorrectly adding “Isaiah.” [Bruce M. Metzger, A Textual Commentary on the Greek New Testament, 2nd ed. (Deutsche Bibelgesellschaft, 1994), 44.]
Mark 1:2-3 quotes from Malachi and Isaiah; the Byzantine text begins Mark 1:2 with: “as it is written in the prophets,” whereas NA28 begins Mark 1:2 with: “as it is written in Isaiah the prophet.” The difference is: “the prophets” vs. “Isaiah the prophet.” According to Snapp, “the reading ‘in Isaiah the prophet’ was created by an early copyist to make the statement more specific, naming the more prominent prophet who was being quoted.” [Snapp, New Testament Textual Analysis, 191.]
From these examples, Snapp extrapolates a general tendency: “in the Gospels, when a prophet is quoted but not specifically named, it is not rare to see that his name is provided.” [Snapp, New Testament Textual Analysis, 191.] While all this information is interesting and revealing of scribal habits, I cannot agree with Snapp’s rather broad observation and alleged scribal habit. In four out of the six examples, the addition of a specific name is a singular reading in the Greek tradition, only has versional support, or is only found in the margin and not the main text:
The main text of Codex Bezae, several versions, and Irenaeus (in part) in Matt 1:22-23.
The margin of Codex Sinaiticus and the main text of Codex Vercellensis in Matt 2:5-6.
The margin of Codex Sinaiticus in Matt 2:15.
The margin of 021, the text of 42, several versional witnesses, and Hilary in Matt 21:4-5.
Only in Matthew 13:35 and Mark 1:2 does the more specific reading have substantial support. In other words, four of the six examples are eccentric outliers. Thus, from two examples and four eccentric examples, I do not think we can extrapolate a broad scribal tendency to take unspecified quotations and make them more specific by providing the name of a prophet (sometimes even an incorrect name). If anything, this tendency to supply specific names is seen more in the versions than in Greek manuscripts.
But these observations are drawing attention to a phenomena that does deserve more attention. There are probably more than just these six examples, although additional examples are likely to have little Greek manuscript support (or only versional support) since they are not already known to us at this point.
Q: In Matthew 13:35, 01 includes the name “Isaiah” even though Psalm 72 is the material being cited. Some influential textual critics of past generations have supported this reading as original, but others consider it so difficult as to be nonsensical. If “Isaiah” falls off the nonsense cliff here, so to speak, why not reject “in Isaiah the prophet” in Mark 1:2?
Hsieh: What happened with Codex Sinaiticus at Matthew 13:35 is interesting and revealing of scribal habits. But first we need to look at the full data: yes, the original scribe of 01 reads “Isaiah,” despite the quotation that follows coming from Psalm 72. But the reading is also corrected by deleting “Isaiah”: NA28 and UBS6 have the deletion as coming from Sinaiticus’s first corrector, while codexsinaiticus.org has “S1”: a correction during the original production of the manuscript, either while copying (in scribendo), or after copying. A look at the manuscript reveals the use of cancellation dots and small diagonal strike-through lines:
Codex Sinaiticus at the deletion in Matthew 13:35 (fol. 207r)
This style of deletion matches the earliest corrector(s) in having only one dot per letter. [Dirk Jongkind, Scribal Habits of Codex Sinaiticus (Gorgias, 2007), 252; Peter Malik, “The Earliest Corrections in Codex Sinaiticus: A Test Case from the Gospel of Mark,” The Bulletin of the American Society of Papyrologists 50 (2013): 229.]Thus, in deciding on the original text of Matthew 13:35, Sinaiticus is a witness to both the addition of “Isaiah,” as well as its omission – both during the original production of the manuscript. But the deletion of “Isaiah” is the final reading provided by Sinaiticus during its original production, so the mistake was recognized and corrected.
According to NA28, the reading adding “Isaiah” is also attested by families 1 and 13, as well as the “queen of the cursives” (minuscule 33); S. C. E. Legg’s edition also has 038, minuscules 174, 230, 543, 713, 788, 826, 828, 983, 1573, 1689, one Vulgate manuscript, and some Ethiopic codices in support of the “Isaiah” reading. There are some good witnesses in support of the “Isaiah” reading (families 1 and 13 and minuscule 33), but not enough to overturn the overwhelming evidence for the omission of “Isaiah” (01 first corrector, 03, 04, 05, 032, 579, 700, 892, 1241, and the Byzantine tradition). [Among minuscule manuscripts, some of the “best” in the Gospels include 33, 69, 579, 700, 892, and 1241. See Gregory R. Lanier, “Dating Myths, Part Two: How Later Manuscripts Can Be Better Manuscripts,” in Myths and Mistakes in New Testament Textual Criticism, ed. Elijah Hixson and Peter J. Gurry (IVP Academic, 2019), 130. Minuscule 69 is not extant for Matthew 13:35; minuscule 33 supports the “Isaiah” reading, while minuscules 579, 700, 892, and 1241 support the omission of “Isaiah.”]
In addition, we can explain on transcriptional grounds how the reading adding “Isaiah” arose in Matthew 13:35. The text being copied is ὅπως πληρωθῇ τὸ ῥηθὲν διὰ τοῦ προφήτου (“so that the word of the prophet might be fulfilled”). The mention of “the prophet” alongside “fulfill” could easily trigger “Isaiah” in the mind of a scribe, especially since similar phrasing is used elsewhere in Matthew:
1:22 fulfilling Isaiah 7:14 (τὸ ῥηθὲν ὑπὸ τοῦ κυρίου διὰ τοῦ προφήτου = “the word from the Lord through the prophet”). 05 and a few others add “Isaiah.”
2:15 fulfilling Hosea 11:1 (τὸ ῥηθὲν ὑπὸ τοῦ κυρίου διὰ τοῦ προφήτου = “the word from the Lord through the prophet”)
2:17 fulfilling Jeremiah 31:15 (τὸ ῥηθὲν διὰ Ἱερεμίου τοῦ προφήτου = “the word through the prophet Jeremiah”)
2:23 fulfilling something in “the prophets” (τὸ ῥηθὲν διὰ τῶν προφητῶν = “the word through the prophets”)
4:14 fulfilling Isaiah 9:1-2 (τὸ ῥηθὲν διὰ Ἡσαΐου τοῦ προφήτου = “the word through Isaiah the prophet”)
8:17 fulfilling Isaiah 53:4 (τὸ ῥηθὲν διὰ Ἡσαΐου τοῦ προφήτου = “the word through Isaiah the prophet”)
12:17 fulfilling Isaiah 42:1-3 (τὸ ῥηθὲν διὰ Ἡσαΐου τοῦ προφήτου = “the word through Isaiah the prophet”)
21:4-5 fulfilling Zechariah 9:9 (τὸ ῥηθὲν διὰ τοῦ προφήτου = “the word through the prophet”)
27:9 fulfilling Jeremiah – although the quotation is actually from Zechariah 11:13 (τὸ ῥηθὲν διὰ Ἱερεμίου τοῦ προφήτου = = “the word through Jeremiah the prophet”)
Different but still relevant is Matthew 2:5-6, where a quotation from Micah 2:5 is introduced with γέγραπται διὰ τοῦ προφήτου (“as it is written through the prophet”).
Alongside these examples is our text under consideration, Matthew 13:35, which reads πληρωθῇ τὸ ῥηθὲν διὰ τοῦ προφήτου (“so that the word through the prophet might be fulfilled”). It is completely understandable how a scribe may have accidentally and unconsciously added “Isaiah,” given that the three preceding occurrences of πληρωθῇ τὸ ῥηθὲν διὰ τοῦ προφήτου have the exact same wording as Matt 13:35 and Isaiah was explicitly identified (Matt 4:14; 8:17; 12:17). These three preceding passages all explicitly mention Isaiah and are without textual variation (according to Swanson’s volume on Matthew).
Matthew 1:22-23 could also be mentioned since Isaiah 7:14 is fulfilled, although the text of Matthew does not explicitly name “Isaiah” except in 05 and a few other witnesses. Perhaps it is a testament to the quality of Sinaiticus’s original scribe(s) that he alone corrected his mistaken addition of “Isaiah” in Matthew 13:35 (whereas families 1 and 13, and minuscule 33 have no corrections).
We can add the testimony of Eusebius and Jerome. [For Eusebius’s and Jerome’s text-critical comments on Matthew 13:35, see Amy M. Donaldson, “Explicit References to New Testament Variant Readings Among Greek and Latin Church Fathers” (PhD diss, University of Notre Dame, 2009), vol. 2, 368-72. Available here: https://curate.nd.edu/articles/thesis/Explicit_References_to_New_Testament_Variant_Readings_among_Greek_and_Latin_Church_Fathers/24732099] Eusebius, in his commentary on Psalm 77 says that “in the accurate copies” (τοῖς ἀκριβέσιν ἀντιγράφοις), Isaiah is omitted. Jerome claims to have found the reading “Asaph” (the author in the subscription of Psalm 78) “in all the ancient copies, but people in their ignorance changed it.” Jerome conjectures that scribes didn’t know who Asaph was, so they changed the reading to the more familiar name Isaiah. According to NA28, the Asaph reading is only attested in manuscripts known to Jerome and is thus found in no extant manuscripts; the only evidence we have is that Jerome claimed to have found the Asaph reading “in all the ancient copies.” But the Asaph reading clearly seems like an improvement to the text. However, we see Jerome attempting to deal with the problem based on transcriptional probability (a scribe being unfamiliar with Asaph changed the text to Isaiah) rather than tossing out the “Isaiah” reading as nonsensical.
Now regarding “Isaiah” in Mark 1:2, the situation in Matthew 13:35 is not comparable. They are different in the three ways:
(1) In Mark 1:2, the “mistaken” reading in NA28 has γέγραπται ἐν τῷ Ἡσαΐᾳ τῷ προφήτῃ (“as it is written in Isaiah the prophet”), which feels like an “error” because Mark first quotes from Malachi 3:1, then from Isaiah 40:3. Thus the NA28 reading with “Isaiah” in Mark 1:2 is partially true to the context. The Byzantine reading γέγραπται ἐν τοῖς προφήταις (“as it is written in the prophets”) removes the “error” of NA28’s reading. In contrast, Matthew 13:35 only quotes from Psalms and has no quotation from Isaiah. Thus the “Isaiah” reading in Matthew 13:35 is not even partially true but completely wrong. In summary: while the readings with “Isaiah” in Mark 1:2 and Matthew 13:35 are both “harder” readings because they appear to be a mistake, the context of Mark 1:2 at least has a quotation from Isaiah, whereas Matthew 13:35 does not.
(2) The two readings in Mark 1:2 are different syntactically, so that an accidental change is unlikely (plural: ἐν τοῖς προφήταις, “in the prophets” vs. singular: ἐν τῷ Ἡσαΐᾳ τῷ προφήτῃ, “in Isaiah the prophet”). Thus, Mark 1:2’s variants almost certainly must have been an intentional change (in either direction). In contrast, the two readings in Matthew 13:35 both contain the singular “prophet” and the difference is the addition or omission of one word; the longer reading adding “Isaiah” could have been an unintentional addition, influenced by and harmonized towards earlier passages, which contain the same exact phrasing (πληρωθῇ τὸ ῥηθὲν διὰ τοῦ προφήτου) and add “Isaiah.”
(3) The better witnesses attesting to the “Isaiah” reading in Mark 1:2 and Matt 13:35 are different: In Matthew 13:35, the “more specific reading” (including “Isaiah”) does not have the support of 01 (since 01 first corrector removes “Isaiah”), 03, and several “good” minuscules (700, 892, 1241). Whereas these five manuscripts switch sides to the more specific reading including “Isaiah” in Mark 1:2. The “better” manuscripts which stay consistent in supporting the more specific reading in both Matt 13:35 and Mark 1:2 are family 1 and minuscule 33.
Perhaps this reveals a “cult of best manuscripts” in favoring 01 and 03, but other evidence (mentioned above) factors into the decision: on transcriptional probability, the accidental addition of “Isaiah” in Matthew 13:35 is understandable given how often Isaiah occurs in Matthew’s fulfillment formulas. In contrast, accidental change in Mark 1:2 is unlikely, given that a change from the plural (ἐν τοῖς προφήταις) to the singular (ἐν τῷ Ἡσαΐᾳ τῷ προφήτῃ) seems more deliberate (in either direction). And the literary contexts also differ substantially: Isaiah is quoted in the near context of Mark 1:3, so that an initial “apparent” difficulty with the “Isaiah” reading in Mark 1:2 is not really problematic on further reflection. In contrast, Isaiah is not quoted at all in the near context of Matthew 13:35 (only in the far context of Matthew 13:14-15 is Isaiah mentioned); “Isaiah” is a truly difficult reading in Matthew 13:35, but for reasons of external evidence (lacking the support of 01, 03, 700, 892, 1241) and likely accidental error in adding “Isaiah,” the more difficult “Isaiah” reading should be rejected in Matthew 13:35.
Q: The relatively new Evangelical Heritage Version (2017) has an interesting New Testament base-text, incorporating significantly more Byzantine readings into the text that the ESV, NIV, NLT, etc. Is this going to start a trend?
Hsieh: Bible translations (like Greek critical editions) are highly monopolistic, so that a few translations and editions dominate the market. While it is interesting that a new translation adopts more Byzantine readings, I find it hard to imagine more mainstream translations doing the same, unless they follow the new trends in the Tyndale House Greek New Testament (THGNT) and Editio Critica Maior (ECM), both of which adopt more Byzantine readings. [See my forthcoming chapter, “Two Ships Sailing Toward Byzantium,” in the 2026 conference proceedings of the Center for the Study of New Testament Manuscripts (CSNTM) Text & Manuscript conference (forthcoming with Hendrickson Publishers).] And even if they consult these new editions, the translation committees may not agree with all the changes.
But even if revised versions of the ESV, NIV, etc. follow the new critical texts, changes happen slowly as committees meet annually (if even that often), so it could be many, many years before these new Byzantine readings show up in translations. But the new trend will probably eventually make its way into mainstream translations. It will take time and the ECM will not be complete until the mid-2030s, with Nestle-Aland 30 probably published in the late-2030s. [This estimate is based on Martin Karrer’s editing of ECM Hebrews, which has a ten-year grant from the German government, running from 2025 to 2035.]
Q: Will the Nestle-Aland compilation ever adopt the reading of 01 and 03 at Matthew 27:49 and thus put the last nail in the coffin of the doctrine of Biblical inerrancy?
Hsieh: This famous “Western non-interpolation” tortured Westcott & Hort into printing the longer reading in their main text in double brackets rather than relegating it into the margin, despite the longer reading being a clear harmonization to John 19:34. The longer reading is found in the reasoned eclectic’s “favorite” manuscripts: Codices Sinaiticus and Vaticanus (also 04, 019, 036, 1010, some Vulgate manuscripts, Christian Palestinian Aramaic in part, one Ethiopic manuscript, Middle Egyptian Coptic, and 030 with some modifications to match John 19:34 more closely). Sinaiticus and Vaticanus supporting the harmonization should not be alarming: all manuscripts have errors and even Sinaiticus and Vaticanus err with harmonizations. Willem Wisselink has shown that Codex Vaticanus harmonized in 31 percent of the 1,489 variations he investigated; this was the lowest number among manuscripts he surveyed (P45, P75, 01*, 01 corrector, 02, 05, 032, 038, 045, 33), but it was definitely not zero harmonizations! [
Willem Wisselink, Assimilation as a Criterion for the Establishment of the Text (Uitgeversmaatschappij J. H. Kok, 1989), 78.]
An intentional change in Matthew 27:49 could go either way: (1) a scribe could have removed the longer reading because of the problem of chronology and a contradiction between Matthew and John. [Brooke Foss Westcott and John Anthony Hort, “Appendix I: Notes on Selected Readings,” in The New Testament in the Original Greek, Volume 1: Introduction and Appendix (Macmillan, 1881), 22; Philip Wesley Comfort, A Commentary on the Manuscripts and Text of the New Testament (Kregel Academic, 2015), 176.] John has the spear piercing Jesus’s side after his death, whereas the longer reading in Matthew 27:49 would have the spear piercing Jesus’s side before his death. Or, (2) Bart Ehrman suggests the longer reading was added as an anti-docetic “orthodox corruption,” proving that Jesus had real flesh and blood. [Bart D. Ehrman, The Orthodox Corruption of Scripture: The Effect of Early Christological Controversies on the Text of the New Testament, 2nd ed. (Oxford University Press, 2011), 228-29.]
Another suggestion from Ulrich Schmid is that the longer reading was an early reader’s note in the margin of manuscript, which eventually made its way into the main text of a manuscript. [Ulrich Schmid, “Conceptualizing ‘Scribal’ Performances: Reader’s Notes,” in The Textual History of the Greek New Testament, ed. Klaus Wachtel and Michael Holmes (Society of Biblical Literature, 2011), 60-61.] Schmid’s argument is based partly on how the version of the spear incident in Matthew 27:49 is not an exact replica of John 19:34 – the beginning is substantially different (ἄλλος δὲ λαβὼν λόγκην in the long reading of Matthew 27:49 vs. ἄλλ᾽ εἷς τῶν στρατιωτῶν λόγχῃ in John) and in the long reading of Matthew 27:49, the word ἔνυξεν is placed earlier, ἐυθύς is omitted, and the word order for blood and water is reversed (030 “corrects” the last two of these differences).
Matthew 27:49 would not be the only place in Matthew where material from John’s Gospel has intruded into Matthew:
At Matt 27:35, a good number of manuscripts (037, 038, families 1 and 13, 1243, 1424, Harklean Syriac) have: ινα πληρωθη το ρηθεν δια του προφητου διεμερισαντο τα ιματια μου εαυτοις και επι τον ιματισμον μου εβαλον κληρον (“so that the word through the prophet might be fulfilled: ‘They divided my garments among themselves and cast lots for my clothing’”). This comes from John 19:24.
At Matt 28:18, a few manuscripts (038 and Syriac Peshitta; 1604 with minor variation) have: καθως απεστηλεν (απεσταλκεν 1604) με ο πατηρ· καγω αποστελω (αποστελλω 1604) υμας (“just as the father sent me, so also I send you”). This comes from John 20:21.
Matthew 27:49 would be the only instance where 01, 03, and 04 have a Johannine intrusion, but none of these harmonizations should be accepted as original. Even early manuscripts have harmonizations (as Wisselink demonstrated).
Therefore, I find it unlikely that Nestle-Aland will adopt the longer reading in Matthew 27:49, especially given the small number of witnesses in favor of the longer reading and the aversion to harmonizations by the Nestle-Aland editors. We await ECM Matthew and its results, but the Coherence Based Genealogical Method (CBGM) will probably find the shorter reading to have better coherence given its wide attestation.
Biblical inerrancy rests on a stronger foundation than one textual variant. Among the large amount of literature on inerrancy, I find Paul Feinberg’s chapter to be carefully argued. [Paul D. Feinberg, “The Meaning of Inerrancy,” in Inerrancy, ed. Norman L. Geisler (Zondervan, 1980), 265–304, 468–72.]
] And there is not even a strong text-critical argument to accept the longer reading of Matthew 27:49, which then introduces a potential contradiction. The support of Codex Sinaiticus and Codex Vaticanus for the longer harmonizing reading is out of the ordinary, but only shows that all manuscripts are flawed.
**********
Thanks, Nelson, for sharing your thoughts with us!
