Video Lecture: Testing the Tests

Testing the Tests

Now at YouTube: Lecture 14 - Testing the Tests In this 32-minute video, I review some shortcomings of earlier canons, or guidelines, of textual criticism (especially "prefer the shorter reading"), and propose some new ones. (This includes a brief look at twelve textual contests.)

https://www.youtube.com/watch?v=JgCEc3bjNc0

Book Review: Textual Criticism of the Bible (Anderson & Widder)

            Seminary students unfamiliar with New Testament textual criticism may learn a lot from Textual Criticism of the Bible, by Amy Anderson and Wendy Widder, recently published by Lexham Press.  (The same volume includes an introduction to the textual criticism of the Old Testament – but since my research-specialty is in New Testament textual criticism, this review is not intended to address the portions about the Old Testament; this review does not cover the third chapter, which concentrates on the textual criticism of the Old Testament.)  Novices should be cautioned, however, that they would learn just enough to be dangerous if they were to stop with this book.

            In chapter 1, a scant eight pages are sufficient to define textual criticism, to describe the distinct goals of textual criticism, and to illustrate the need for its existence.  Readers will notice here the authors’ use of the novel term “Ausgangstext,” defined as “the ancient form of the text that is the ancestor of all extant copies, the beginning of the manuscript tradition.”         

Chapter 2, An Overview of Textual Criticism, includes a helpful tour of examples of different kinds of scribal corruptions, from the Old Testament and the New Testament.  Haplography, parablepsis, conflation, glosses, metathesis, and harmonizations are among the error-inducing mechanisms that are covered.  (Conflations are incorrectly described as “common” in the Byzantine Text on p. 117.)  

Readers who approach this chapter expecting to learn about the excellence of Codex Vaticanus may be surprised at some of the examples given:  although Alexandrian readings are favored in Mark 1:2, Luke 23:17, Luke 24:53 and I Tim. 3:16, readings found in Vaticanus in John 17:5, Mark 3:16, Acts 19:34, and First Thessalonians 2:7 are presented as examples of scribal errors.  (I think.  The authors seem undecided in some cases; in the discussion about Mark 3:16 on page 24, it is not easy to tell which reading is preferred by the authors.)  In most cases, external evidence is only frugally described; a paragraph on John 5:3-4, for example, fails to mention Tertullian’s support for the passage. 

Although several mistakes in the text of Codex Vaticanus are used as examples of scribal errors, the appeal to “the oldest and best manuscripts” is made repeatedly in this chapter, and “the oldest and best manuscripts’ can safely be understood to mean “Codex Vaticanus and whatever happens to agree with it.”  Readers might wonder how it is that the best manuscripts contain some of the worst errors.

Similarly some readers may feel misled by the following statement on pages 12-13:  “Those variation units that affect the meaning of a biblical text are found in the footnotes of any good English Bible.  Even these variants do not affect doctrine or theology.”  This is misleading on two levels.  First, it is misleading because the textual footnotes in most English translations are minimal:  consider that the NASB has only one textual footnote in Jude; the ESV has one; the CSB has five; meanwhile, the number of meaning-affecting textual differences between the Byzantine Text and the Nestle-Aland compilation in Jude is twelve.  Do the authors think that the NASB and ESV and CSB are not good English Bibles?  Second, it is misleading because textual variations that affect doctrine exist.  For example, the doctrine of inerrancy would be wrecked if the reading found in Matthew 27:49 in “the earliest and best manuscripts” were to be placed in the text.

The chapter concludes with a brief review of basic principles of textual criticism, and fortunately this includes the key principle that that the reading that best explains the origin of its rivals is to be preferred.  Unfortunately it also includes the idea that “The reading found in the oldest manuscripts” should be preferred; this canon should be reworded to include not only manuscripts, but all forms of external evidence, and it should be qualified by mentioning that this canon gives an edge to evidence found in Egypt, inasmuch as the low-humidity climate there allowed papyrus to survive longer than elsewhere. 

Worse than this, however, is the authors’ endorsement of lectio brevior (prefer the shorter reading).  Only readers careful enough to consult the footnotes will notice a heavy qualification that this principle has been challenged and empirically reversed in research on some early manuscripts by James Royse.  No means are offered by which to verify the authors’ assertion that lectio brevior is nevertheless a valid criterion for later documents; recent research by Alan Taylor Farnes indicates that the authors are wrong, and that even in later manuscripts neither the length nor the brevity of a reading should be used as an indicator of genuineness.

Chapter 4 – Introduction to New Testament Textual Criticism – serves the same purpose as similar sections of Bruce Metzger’s The Text of the New Testament, describing early scholarship in the field (beginning with patristic writers, and continuing to the present day) and manuscripts, versions, and patristic evidence.  Some significant researchers, and some significant witnesses, have been left unmentioned.  Bengel and Griesbach are mentioned, but not Bentley and Scrivener.  Codices Π, Κ, and Σ receive no special attention. 

Readers will learn on page 116 that the early papyri demonstrate a “careful scribal tradition” but when they read page 127 they will also learn that in the early papyri, “errors and changes” may be “fairly common.”  This might confuse beginners who may not understand how the best manuscripts can be riddled with mistakes. 

It was a little disappointing that Amy Anderson, who has done some detailed research about family-1, did not say more about this cluster of manuscripts; there is not even a list of its core members.  Its text has received a slight chronological upgrade, although the confidence with which it is made has not:  whereas Metzger described it as a text which appears to go back to the third and fourth centuries, here the agreements of minuscules 1 and 1582 are considered likely to be the text of the archetype of the group, which is thought to represent a text from the third century.

Very little is said about lectionaries, and nothing is said about amulets, and nothing is said about conjectural emendations – just the opposite of what should have happened, considering the significant research on these subjects that has occurred since Metzger wrote his introduction. 

Versions are given surprisingly little attention – so little that one can only imagine that the authors foresee that their readers are likely to have Metzger’s book about versions on the shelf; otherwise they will not learn anything about the Gothic, Armenian and Georgian versions, or the Peshitta, in this chapter, except that they exist.  The description of the Sinaitic Syriac and Curetonian Syriac should be changed, lest readers conclude that these manuscripts, rather than the texts they contain, are “possibly dating back to the early third century.”

In a brief review of modern editions of the Greek text of the New Testament, a bias in favor of the Nestle-Aland compilation is not entirely concealed; the Nestle-Aland compilation is called “eclectic” even though it is very heavily Alexandrian.  Metzger’s chapter From Griesbach to the Present is superior to the too-concise review here in all respects except one:  Anderson is up-to-date, and includes sections on the Byzantine Textform compiled by Robinson and Pierpont, the SBLGNT compiled by Michael Holmes, and the Tyndale House edition of the Greek New Testament.  (It is only via a quotation from the preface to the Tyndale House compilation that Samuel Tregelles is mentioned at all in this book; he (like Scrivener and Burgon) was completely passed over in the section on New Testament textual criticism in the late seventeenth century to the present.) 

The Robinson-Pierpont Byzantine Textform is mistreated in this book.  First, the authors associate it with KJV-Onlyism, even though Dr. Maurice Robinson, co-editor, has plainly dismissed KJV-Onlyism as a flawed position.  Second, rather than utilize statements from advocates of Byzantine Priority (except for a 24-word quotation from Hodges and Farstad), the authors have turned to Philip Comfort and Daniel Wallace to describe it.  This is patently unfair (especially because the essays by Wallace on the subject – mentioned in a footnote – contain more than their fair share of specious reasoning and mistakes). 

It did not go unnoticed that the authors mention, on page 148, that in the Tyndale House edition (2017), the books of the New Testament are arranged in the order Gospels-Acts-Catholic Epistles-Pauline Epistles – Revelation.  Why was this not also said about the 2005 Byzantine Textform, which also arranges the text this way?  I cannot help wondering, Did the authors not bother to read the Byzantine Textform enough to notice? Similarly, the authors mention that the SBLGNT is online – but so is the Byzantine Text; why was this not mentioned?

Continuing:  readers are told (p. 121) that the discovery of the papyri “provided scholars with new evidence to confirm or refine Westcott and Hort’s findings.”  Perhaps a more candid description would add, “and refute,” inasmuch as Hort’s assumptions that distinctly Byzantine readings are all late, and that conflations imply lateness, and that some shorter readings in Codex D (Western Non-interpolations) are original, were opposed by the data from the early papyri (see Sturz, The Byzantine Text-type and New Testament Textual Criticism, for details).    

 As chapter 4 concludes, the authors take readers on a tour of selected variant units, and this section serves (as its sub-title indicates) as a lesson in How To Do New Testament Textual Criticism.  Along the way, readers are introduced to the various symbols that can be found in the Nestle-Aland compilation (such as º and ^□ and ^┬).  The main variant-units covered are in Ephesians 1:1, John 3:32, Mark 1:2, First Thessalonians 2:7, Luke 4:4, Revelation 1:8, and Romans 5:2. 

This section is unsatisfactory in several ways:

● First, lectio brevior continues to be used as if it is a valid premise. 

● Second, parts of this section just restate what has already been stated:  compare for example page 45 – “The reading that best explains the origin of the other readings is probably original” – with page 158 in this section:  “If you can explain how one variant led to the occurrence of the other variants, that variant is the most probable Augsgangstext.” 

● Third, the section is not really a lesson in how to do textual criticism; it is a lesson in how to let Bruce Metzger posthumously do your textual criticism.  For example, in the portion about Mark 1:2, witnesses that are not mentioned in the Nestle-Aland or UBS apparatuses are treated as if they do not exist.  Variants that are not mentioned in the Nestle-Aland or UBS apparatuses are treated as if they do not exist.  An appeal is made (on behalf of the Alexandrian reading) that “the citing habit of ancient times” account for the reading “in Isaiah the prophet,” without anything to verify that ancient citation-habits actually do this.  The authors attempt to excuse the reading “in Isaiah the prophet” by theorizing that Mark “may not have had any access at all to copies of OT books.”  Seriously?  Finally, the authors propose that “It is much more difficult to imagine a later scribe changing τοῖς προφήταις to τῷ Ἠσαΐᾳ τῷ προφήτῃ” – but there is an abundance of evidence that copyists tended to make citation-sources more specific, not more generalized – supplying a prophet’s name (and in some instances the wrong name) when the reference in the original text was non-specific.

● Fourth, the authors misrepresent some witnesses as if they are members of textual groups to which they do not really belong.  For example, as the authors discuss the variant-unit in Luke 4:4 they state that the reading chosen by the Nestle-Aland editors is “supported by the Syriac Sinaiticus, Sahidic, and part of the Bohairic.  This group is mainly related to the Alexandrian textual stream.”  The Sinaitic Syriac (I do not know why they described it as the “Syriac Sinaiticus,” which is correct but old-fashioned) is, however, not Alexandrian; as Metzger states in The Text of the New Testament, “In general the Old Syriac version is a representative of the Western type of text.”  This error is compounded when the authors proceed to argue that “only one textual group is represented” in favor of the shorter reading.

● Fifth, although the authors recommend consulting Metzger’s Textual Commentary and commentaries in which the commentators give special attention to text-critical issues, there is no special mention of Wieland Willker’s superior (and free) Textual Commentary on the Greek Gospels, nor of Reuben Swanson’s multi-volume New Testament Greek Manuscripts, in which the exact words of important Greek manuscripts are lined up horizontally and compared line by line, giving readers access to much more data than what is presented in the Nestle-Aland and UBS apparatuses.  Robert Waltz’s online Encyclopedia of New Testament Textual Criticism and Alan Bunning’s Center for New Testament Restoration have also gone unmentioned.   

           

            Readers may find useful the inclusion of a list of Resources for Further Study at the end of each chapter.   Strangely, David Alan Black’s New Testament Textual Criticism:  A Concise Guide and Philip Comfort’s Encountering the Manuscripts are in the list in chapter two and in chapter four – with two different descriptions.  J. Scott Porter’s 1848 Principles of Textual Criticism is in the list for chapter 2; readers may find it interesting that J. Scott Porter was also the author of Twelve Lectures on Unitarianism, in which he attempted to use textual criticism as a weapon in his battle against the doctrine of the deity of Christ; meanwhile Scrivener’s A Plain Introduction to the Criticism of the New Testament has gone unmentioned.    

The glossary at the end of the book is helpful, but it would be more helpful if it included a few more entries, and did not display a slight bias in favor of the Alexandrian Text.

In closing:  while Anderson and Widder’s book is ten times better than the recently escaped Fundamentals of New Testament Textual Criticism by Porter and Pitts, it may serve readers best as an update to Metzger’s Text of the New Testament than as a stand-alone introduction, generally echoing the same pro-Alexandrian approach and assumptions.  New Testament researchers with access to Metzger’s Text of the New Testament and the online materials by Wieland Willker and Robert Waltz may consider it superfluous.       

P.S.  The word “critic” is missing in line 10 on page 43.  Also, contrary to the list on page 140, minuscule 1424 is no longer housed in Chicago; it was returned to Greece in 2016.

The Text of Reasoned Eclecticism: Is It Reasonable and Eclectic? Part Three of a Four-Part Response to Dan Wallace

Internal Evidence

            I now turn to Wallace’s comments sub-titled Internal Evidence.  In this section, Wallace tends to misrepresent the Byzantine Priority position as it is currently framed by its chief advocate, Maurice Robinson.  Wallace cited Michael Holmes to support the idea that majority-text advocates “object quite strenuously to the use of the canons of internal evidence.”  Meanwhile in the real world, Robinson candidly affirms in The Case for Byzantine Priority:  “The basic principles of internal and external evidence utilized by Byzantine-priority advocates are quite familiar to those who practice either rigorous or reasoned eclecticism.”  Robinson proceeds to list and describe eight principles of internal evidence!²³ 

            As much as Wallace might wish that Byzantine-priority advocates disregard internal evidence, that is simply not true.  Rather, Byzantine-priority advocates (and other textual critics with a dislike of propaganda disguised as textual commentary) object to inconsistent applications of internal evidence which are primarily steered, not by the internal evidence as it exists for specific variant-units, but by a model of transmission-history that precludes, or sets ridiculously high hurdles against, the authenticity of Byzantine readings.  In other words, textual critics who believe, as Wallace appears to believe, that the Byzantine Text did not exist until after 300, will approach the internal evidence accordingly, and will consider the most crucial internal aspects of the evidence to be whatever results in the rejection of the Byzantine reading.

            Wallace invited readers to consult Metzger’s A Textual Commentary on the Greek New Testament to see how internal evidence was used in the United Bible Societies’ compilation-committee decisions to accept one reading over another – especially when the adopted reading was given an “A” rank, which means that the compilers regarded the adopted text as “virtually certain.”²⁴   With apologies for lengthening this composition – Invitation accepted.  Let’s consider four “A” readings from the Gospel of Mark.  (These are not the best examples of the extreme lengths to which the UBS compilers have taken internal evidence as a means to reject Byzantine readings, but as a response to the invitation to use only A-ranked readings, they will have to do.)

Mk 1:14 – Against evidence from the Byzantine Text, the Vulgate, the Gothic version, the Peshitta, and codices D and W (among others – see the textual apparatus for further details) that supports the inclusion of της βασιλείας, Metzger argued for the Alexandrian non-inclusion of the phrase.  The longer reading, he claimed, “was obviously made by copyists in order to bring the unusual Markan phrase into conformity with the much more frequently used expression “the kingdom of God” (cf. ver. 15).”  On the other hand, however, it could be argued from internal evidence that the longer reading is more consistent with Markan style, because nowhere else in Mark does the phrase “gospel of God” appear.  Metzger’s idea requires that somewhere in the ancestry of diverse witnesses such as D, W, the Peshitta, and the Vulgate, they were all influenced by some copyist’s intentional insertion – whereas all that is required to produce the shorter reading is an accidental error that resulted when copyists’ line of sight drifted from the τ in της to the τ in του.

Mk 1:34 – Agreeing with external evidence from the Byzantine Text, the Vulgate, the Old Latin copies, the Gothic version, and the Peshitta, Metzger conceded that the longer reading in the Alexandrian text (Χριστον ειναι, read by Codex Vaticanus, L, and f¹) was an addition, by which copyists harmonized the text in Mark to the text in Luke 4:41.  There would be no reason for copyists to omit the phrase, if it had been present originally, but it would be a natural expansion.

Mk. 7:4 – Agreeing with external evidence from the Byzantine Text, the Vulgate, D, the Old Latins, the Peshitta, the Gothic version, and Origen, Metzger conceded that the Alexandrian reading ραντισωνται (supported by B, À, and the Coptic version) was an alteration introduced by Alexandrian copyists.  (This is given an “A” certainty-rank in UBS-2, but in Metzger’s Textual Commentary it has a “B” rank.)

Mk. 9:29 – Against evidence from P⁴⁵, the Byzantine Text, codices A, D, W, L, f¹, the Old Latin copies, the Vulgate, the Peshitta, and almost everything else, Metzger argued that the shorter reading (supported only by À* B, 0274, one Old Latin copy (Bobbiensis), one Old Georgian copy, and Clement) was original, and that the phrase “and fasting” was the invention of copyists.  Apparently neither Metzger nor Wallace²⁵ could perceive that copyists were troubled by a reading which could be interpreted to suggest that Jesus could not have cast out a particular kind of demon unless He had first fasted, even with the earliest (P⁴⁵) and most diverse evidence (from multiple locales) pointing the way.    

Mk. 11:26 – Against evidence from the Byzantine Text, almost all the Old Latin copies, the Vulgate, the Peshitta, and the Gothic version, Metzger rejected the inclusion of verse 26, supposing it to be an insertion based on Matthew 6:15, even though its correspondence to that verse is loose.  This is an instructive example of favoring external evidence over internal evidence, because Metzger noticed that it would have been very easy for copyists to skip verse 26 if their line of sight drifted from the words τα παραπτωματα υμων at the end of verse 25 to the exact same words at the end of verse 26, but did not consider this internal consideration to be persuasive against the Alexandrian witnesses.  Metzger argued that the shorter reading is in “early witnesses that represent all text-types,” although the same thing can be said about the longer reading, which is supported by Byzantine, Western, and Caesarean witnesses).      

These four samples indicate that internal evidence can be a very slippery thing, and that its force often depends upon the extent to which the textual critic is willing to engage his imagination to devise explanations for the origination of whatever reading is least consistent with his model of the text’s transmission-history.  Were we to explore Metzger’s Textual Commentary in detail, we would see that very frequently, the UBS compilation-committee attributed longer readings to scribal creativity where the shorter reading is attributable to scribal carelessness – especially when the shorter reading is Alexandrian.²⁶ 

Only when the Alexandrian reading is longer (as in Mark 1:34), or when the Alexandrian reading is so obviously secondary that no amount of ingenuity can plausibly salvage it (as in Mark 7:4), does internal evidence seem to have the ability to outweigh Alexandrian external evidence.  Metzger and his fellow UBS committee-members were not alone in their consideration of the length of rival variants as an internal consideration by which to judge between them:  the canon, “prefer the shorter reading” has been a major text-critical canon for a very long time, and in the late 1800’s and throughout the 1900’s, it was appealed to very frequently as a pivotal consideration, resulting in the rejection of very many longer Byzantine readings, in favor of shorter Alexandrian readings.  When the Byzantine Text has the shorter reading – as it does hundreds of times, for example in James 4:12 (“and Judge” is not in the Byzantine Text) and Jude v. 25 (“through Jesus Christ our Lord” is not in the Byzantine Text) – the compilers seem to have no problem recognizing parableptic errors, but when the Alexandrian Text has the shorter reading, the theory of first resort is that the longer reading originated with copyists.²⁷

Not only is this inconsistent, but it defies the analysis which James Royse provided of the text of some early New Testament papyri, showing that the copyists of those manuscripts tended to make more omissions than additions, at a proportion of three omissions for every two additions.²⁸  This tends to altogether invalidate the “prefer the shorter reading” canon, and indicates that every variant-unit that has been decided on the basis of that internal consideration ought to be revisited.

- Continued in Part Four - 

_______________

FOOTNOTES

23 – Of course Wallace could not anticipate, in the 1990’s, what Maurice Robinson would write in 2001.  But it is 2015 and Wallace’s article remains online, spreading false impressions about the nature of the view of the leading proponents of Byzantine Priority. 

24 – See page x of the introduction to the second edition of the United Bible Societies’ Greek New Testament:  “The letter A signifies that the text is virtually certain.”

25 - See the NET’s note on this variant-unit:  “a good reason for the omission is difficult to find.”

26 – See Dennis Kenaga’s Skeptical Trends in New Testament Textual Criticism for additional critiques of inconsistencies in Metzger’s use of internal evidence. 

27 – The treatment of Matthew 12:47, which is absent in À and B and the Sahidic version (et al), is an outstanding example of pro-Alexandrian compilers’ reluctance to recognize parableptic omissions in the Alexandrian text.  Internal evidence has prevailed, but just barely.  Although the cause of a parableptic error is obvious – verses 46 and 47 end with the same word, λαλησαι – the UBS compilers assigned it a “C” rank (meaning that they harbor a “considerable degree of doubt” about the adopted text) and, echoing Nestle’s earlier treatment, bracketed the verse (to emphasize that the enclosed words have “dubious textual validity”).  

28 – See James R. Royse’s painstakingly prepared Scribal Habits in Early Greek New Testament Papyri, released in 2008.