Friday, January 30, 2015

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


The 2005 edition of
The New Testament
in the Original Greek -
Byzantine Textform
          In the years since Daniel Wallace wrote The Majority Text and the Original Text:  Are They Identical?,1 no less than four English translations of the majority (i.e., Byzantine) text of the New Testament have been published, and another version (the Modern English Version) has recently been released which was based on the Textus Receptus.  A definitive edition of the Byzantine New Testament has been published.2  A recent poll showed that the King James Version is the most widely read Bible version in the United States, by a large margin.3  And the Center for Study and Preservation of the Majority Text has been established.

            In this article I will address some of the inaccuracies, overstatements, and poor argumentation in Dr. Wallace’s article.  My purpose here is not to endorse the Byzantine Text in its entirety – I do not subscribe to Byzantine Priority – but to show that Dr. Wallace’s reasons for rejecting it are insufficient (and to clarify some peripheral misstatements in his article). 
            Wallace’s very first paragraph echoes a common put-down of the Textus Receptus:  “In compiling the TR Erasmus simply used about a half dozen late manuscripts that were available to him.”  Although it is true that Erasmus, when preparing the first edition of his Greek text, only had immediate access to less than a dozen Greek manuscripts, during the 1500’s the Textus Receptus went through multiple editions and comparisons involving consultations of patristic references (in which patristic writers cited their manuscripts), Lorenzo Valla’s research (which mentioned other manuscripts), the Complutensian Polyglot (the editors of which claimed to have consulted their ancient manuscripts) and manuscripts that were used in the exchanges between Erasmus and his critics, and by Stephanus and by Theodore Beza. 
            A close examination of the annotations of Erasmus and the textual notes of Beza give a much different picture than Wallace's distant glance.  The editors of the texts that reached their standardized expression in the Textus Receptus tended to stick with the Byzantine readings in their Byzantine manuscripts, but to say that they did not use ancient manuscripts such as Codex Bezae and Codex Claromontanus (and, judging from the contents of the Textus Receptus in Matthew 9:36, Codex Regius4) and a list of 365 readings from Codex Vaticanus is blurry reporting.  One can use a manuscript and still reject its readings, as Wallace shows when he says that the Nestle-Aland text is based on over 5,000 manuscripts – over 4,000 of which persistently display Byzantine readings which the editors persistently rejected.               

Preservation and the Byzantine Text

            Wallace targeted Wilbur Pickering’s view of inspiration and preservation as if Pickering’s doctrinal view was the foundation of Pickering’s case for the Byzantine Text (or for a sub-group of it; he currently endorses the f35 text).  Wallace summarized Pickering’s doctrinal premise as follows:  “The doctrine of the preservation of Scripture requires that the early manuscripts cannot point to the original text better than the later manuscripts can, because these early manuscripts are in the minority.”
            When Pickering affirmed the superiority of the majority text over other text-types, he was not really saying much more than what the Westminster Confession of Faith – a major creedal statement in the history of Reformed Christendom, produced in 1646 – said in its first chapter, part eight: 

            “The Old Testament in Hebrew (which was the native language of the people of God of old), and the New Testament in Greek (which, at the time of the writing of it, was most generally known to the nations), being immediately inspired by God, and, by His singular care and providence, kept pure in all ages, are therefore authentical; so as, in all controversies of religion, the Church is finally to appeal unto them.”

            The producers of the Westminster Confession thus enunciated the doctrine that the New Testament text has been kept pure in all ages – and the text they knew was an essentially Byzantine text.  This purity pertains to the message which was conveyed by manuscript after manuscript, not to the exact form of the text, as if spelling-variations and the quirky mistakes of individual scribes rendered the text impure.  But when one text contains Mark 16:9-20, and Luke 22:43-44, and Luke 23:34a, and John 7:53-8:11, and another text does not contain any of them, but says in Mark 6:22 that Herod’s daughter danced, and says in Matthew 27:49 that Jesus was pierced before He died, what would an impure text look like, if these are both called pure?  These differences significantly shape the message that is being conveyed.  Meanwhile, the differences between the Textus Receptus and the Byzantine Text are not so immense that they significantly shape the message being conveyed.   
            The doctrine of the preservation of Scripture does not preclude the idea that early manuscripts can preserve the form of the text better than the later manuscripts.  Rather, the doctrinal statement that God has kept the text pure in all ages implies that the message of the original text has been perpetuated in the Greek text used by the church.  And the Greek text used by the church, as displayed in thousands of Greek manuscripts, is the Byzantine Text or a sub-group of it.   
            Perhaps Wallace feels that the Westminster Confession’s statement that God has providentially kept the Biblical text “pure in all ages” is simply false.  (Speaking for myself, I do not subscribe to the Westminster Confession.)  But the thing to see is that Pickering is not introducing anything novel into the doctrinal equation.   
            Wallace’s article suddenly jumped to a different topic:  Pickering’s assessment of Hort’s anti-Byzantine view:  “Pickering,” wrote Wallace, “has charged Hort with being prejudiced against the Byzantine texttype from the very beginning of his research: “It appears Hort did not arrive at his theory through unprejudiced intercourse with the facts. Rather, he deliberately set out to construct a theory that would vindicate his preconceived animosity for the Received Text.””
            Wallace did not contradict Pickering on this point, and wisely so, because anyone can consult Hort’s writings and see where, in 1851, near the outset of his research, Hort denounced the Textus Receptus as “vile” and “villainous.”5  Instead, Wallace accused Pickering of doing the same thing:  “His particular view of preservation seems to have dictated for him that the majority text must be right.” 
            Such a tu quoque retort misses the point, which is that Hort clearly had some idea, in 1851, of what sort of conclusions he was going to reach by 1881 – yet his clearly expressed prejudice has not prevented very many textual critics from regarding his conclusions as correct.  So why, if Pickering was not the only researcher on earth with an entirely objective and detached mind, should this be used as a basis to reject his conclusions?  When Pickering insists that the purity of the Hebrew and Greek texts of Scripture has been providentially safeguarded for the church in all ages, he is not saying anything that the Reformers who composed the Westminster Confession did not say.  If the mere possession of a doctrinal view is sufficient to dictate one’s conclusions, then we must all be mute, including Wallace. 
            Pickering’s affirmation of a belief in the providential preservation of the purity of the Hebrew and Greek texts of the Bible is seized by Wallace as if it proves that he worked from presuppositions; i.e., that he was out to prove a premise.  This is, however, only partly true:  Pickering has affirmed that a combination of Scripture-verses may reasonably be taken to imply a promise of the preservation of the Scriptures, “but no intimation is given as to just how God proposed to do it. We must deduce the answer from what He has indeed done.”6  Does not Wallace assert a very similar premise about the New Testament text when he states that all of the original text is extant somewhere – that is, it has all been preserved – and it is the task of the textual critic to discover where?
          There is no evidence that Pickering pre-judged the Alexandrian Text to be vile and villainous in the way that Hort pre-judged the Textus Receptus.  Even if Pickering had done so, however, it would not make the Byzantine Priority view right or wrong, any more than Hort’s initial bias against the Textus Receptus rendered any of his text-critical judgments right or wrong. 

Next, Wallace expressed three “serious problems” with the doctrine of preservation as expressed by Pickering.  Please bear with me, reader, as I address these three points in some detail.

(1)  ● Wallace opposes the idea that God providentially preserved the purity of the text in all ages in the majority of Greek manuscripts because it could be that God providentially did so “in a small handful of witnesses.”  Wallace has unfortunately overlooked the phrase “in all ages” and as a result his idea does not make sense, because it is obvious that the Greek text of the small handful of manuscripts that form the primary basis for the Nestle-Aland text has not been used by the church in all ages. 

(2)  ● Wallace opposes the idea that God providentially preserved the purity of the Greek text in all ages in the majority of Greek manuscripts because, “assuming that the majority text is the original, then this pure form of text has become available only since 1982.”  But it is not as if the readings in the Hodges-Farstad Majority Text were created in 1982!  Hodges and Farstad derived them from the majority of manuscripts (which existed long before 1982) and those manuscripts display the text that was in use in Greek-reading Christendom at the times they were produced. 
            In the course of elaborating on this strained objection, Wallace points out that the Textus Receptus differs from the Hodges-Farstad Greek New Testament “in almost 2,000 places.”  This merits some explanation, so let’s pause a moment to consider some of those differences.  Let’s ask, “If a copyist were to make a manuscript that contained exactly the same text that is in the Textus Receptus, to what extent would it be a non-Byzantine manuscript?”  Here is some data to answer that question.  I will list, for each Gospel, the number of differences between the Byzantine Text (where its testimony is not divided) and the Textus Receptus, followed by the number of differences which indicate an origin in a non-Byzantine source other than parableptic errors, itacisms and orthographic variations attributable to an individual copyist.  This will provide some idea of how non-Byzantine the Textus Receptus is. 

            ■ Matthew:  out of 159 differences, 46 readings in the Textus Receptus are non-Byzantine.
            ■ Mark:  out of 142 differences, 73 readings in the Textus Receptus are non-Byzantine. 
            ■ Luke:  out of 221 differences, 140 readings in the Textus Receptus are non-Byzantine.
            ■ John:  out of 158 differences, 107 readings in the Textus Receptus are non-Byzantine. 

            Thus, in the Gospels, there are 680 differences between the Byzantine Text and the Textus Receptus.  Of those differences, 366 indicate an origin outside the Byzantine transmission-line (i.e., these 366 readings do not look like they began when a copyist was copying from an exemplar that contained nothing but Byzantine readings).  Many of those 366 variants have no effect on the meaning of the text.  (For example, out of the 107 distinctive variants in John, 22 of them are instances where the untranslatable definite article (a single Greek letter) is put before Jesus’ name.) 
            These statistics show that in the Gospels, if one considers Codex Vaticanus and Codex Sinaiticus to both be good representatives of the Alexandrian Text, then the Textus Receptus must be considered a relatively close representative of the Byzantine Text.  Furthermore, a huge chunk of those differences are found in the text of the book of Revelation.  This is partly because of the unique transmission-history of the text of Revelation, and partly because Erasmus relied heavily on a manuscript of Revelation (2814) which has a text that frequently diverges from the norm.  (I have not counted all the differences, but it looks like the average chapter of Revelation in the Textus Receptus disagrees 30 times with the Byzantine Text.) 
            The statistics may be crunched as follows:  out of 1,838 differences between the Textus Receptus and the Byzantine Text,7 680 are in the Gospels, about 660 are in Revelation, and about 500 are in Acts and the Epistles.  That’s 1,199 fewer variants between the Textus Receptus and the Byzantine Text in the entire New Testament than the 3,037 variants that exist between Codex Vaticanus and Codex Sinaiticus, the flagship manuscripts of the Nestle-Aland text.     
            Regarding disagreements between the Textus Receptus and the Byzantine Text, Wallace claims, “Many of these passages are theologically significant texts.”  However, he provides only two examples:  First John 5:7-8 and Revelation 22:19 – the first of which, albeit an interpolation, is one that expresses a teaching found elsewhere in Scripture, and the second of which conveys the same message no matter whether one refers to the book of life (as in the Textus Receptus and KJV) or to the tree of life.  Wallace scrapes the bottom of the barrel, so to speak, in his search for examples of variants within the Byzantine textual tradition that do not uniformly convey the same message. 
            Furthermore, via these comparisons, Wallace has confused the printed text that was used by the church for several centuries – the Textus Receptus – with the text that was available to the church in the manuscripts.  Perhaps Pickering, via some poor wording, left himself vulnerable to the charge that he believed that whatever printed Greek text the church used must be the preserved, original text.  For whatever reason, Wallace attempts to make Pickering’s arguments require an endorsement of the Textus Receptus – which Pickering clearly does not endorse.  This is a common tactic used by those who wish to dismiss discussion of the value of the Byzantine Text:  make it easy for those viewing the debate from a distance to think that one’s opponents are KJV-Onlyists or advocates of every iota of the Textus Receptus.       
            In the course of pointing out that the Textus Receptus includes minority readings, Wallace claimed that “Virtually no one had access to any other text from 1516 to 1881, a period of over 350 years.”  That is a rather inaccurate claim.  The Complutensian Polyglot was not based on the Textus Receptus.  Anyone who possessed a Byzantine manuscript had access to a Greek text other than the Textus Receptus.  The Orthodox churches continued to use the (essentially Byzantine) text of their lectionaries.  Griesbach’s text, in the 1770’s, was very different from the Textus Receptus; Abner Kneeland’s 1823 English version was very different; Granville Penn’s 1836 version was drastically unlike the Textus Receptus.    

(3) ● Wallace opposes the idea that God providentially preserved the purity of the text in all ages in the majority of Greek manuscripts because this would mean that such a text was available in Egypt in the first four centuries.  “But this is demonstrably not true,” he claims.  As the sole support for his claim, Wallace cites Bart Ehrman’s dissertation on the Gospels-text of Didymus the Blind (who lived in the mid-and late 300’s).  As Wallace mentioned, Ehrman concluded that his findings “indicate that no ‘proto-Byzantine’ text existed in Alexandria in Didymus’ day or, at least if it did, it made no impact on the mainstream of the textual tradition there.”  (Yet, not long after Didymus’ death, Codex W appeared, it would seem, in Egypt.)
            However, Ehrman’s data is capable of being interpreted to support a different conclusion.  I have systematically worked through Ehrman’s data about Didymus’ utilizations of the Gospels, and here are some observations:
            ► In Matthew, Didymus agrees with either B or Byz (but not both) 49 times.  Didymus agrees with B against Byz 24 times (49%).  Didymus agrees with Byz against B 25 times (51%).
            ► Ehrman concedes that the data from Mark is too sparse to justify confidence that it reflects the affinities of Didymus’ Gospels-text.  In three of the four cases where Ehrman concludes that Didymus supports a reading in B in Mark, the grounds seem especially questionable.  With these qualifications in mind, in Mark, Didymus agrees with B against Byz 4 times (80%) and with Byz against B 1 time (20%).
            ► In Luke, Didymus agrees with either B or Byz (but not both) 45 times.  Didymus agrees with B against Byz 28 times (62%).  Didymus agrees with Byz against B 17 times (38%).
            ► In John, Didymus agrees with either B or Byz (but not both) 40 times.  Didymus agrees with Byz against B 23 times (57.5%).  Didymus agrees with B against Byz 17 times (42.5%).8
            Thus, rather than concluding that the evidence from Didymus shows that there was no Byzantine or Proto-Byzantine Text available in Egypt (and keeping in mind that we are talking about the Gospels-text used by a man who was blind from his childhood – and also keeping in mind that the evidence from Didymus would never lead us to conclude that a Gospels-text resembling what is found in P45 existed in Egypt; nevertheless there it is), it shows the following:  (1)  Didymus’ text of Matthew agreed with the Robinson-Pierpont Byzantine Text as much as it agreed with Codex Vaticanus; (2)  We don’t have enough data to discern what Didymus’ text of Mark was like; (3)  Didymus’ text of Luke frequently contained readings that are found in the Robinson-Pierpont Byzantine Text but not in Codex Vaticanus, and (4)  Didymus’ text of John agreed with the Robinson-Pierpont Byzantine Text more often than it agreed with the flagship manuscript of the Alexandrian Text. 
            I conclude that the evidence from Didymus’ Gospels-utilizations does not post a problem at all for the idea that the Byzantine Text of the Gospels, or a text very similar to it, was available in Egypt in the 300’s.

            After firing blanks at the Byzantine Text, Wallace attempts to minimize the differences between the Byzantine Text and the Alexandrian Text, arguing that it does not make much doctrinal difference which text-type is used; both, he claims, are doctrinally orthodox:  “For over 250 years, New Testament scholars have argued that no textual variant affects any doctrine.” 
            He’s partly right:  many New Testament scholars (such as D. A. Carson) have made such a claim.  But others, such as George Vance Smith, a Unitarian scholar who was on the translation-committee of the 1881 Revised Version, have made different assessments.  Referring to the changes (some translational, but mostly textual) introduced in the Revised Version, Smith wrote,

          “The changes just enumerated are manifestly of great importance, and are they not wholly unfavourable to the popular theology?  Many persons will deny this, but it is hard to see on what grounds they do so.  Or, if it be true that the popular orthodoxy remains unaffected by such changes, the inference is unavoidable that popular orthodoxy must be very indifferent as to the nature of the foundation on which it stands.”9

            As a person who is much more aware of doctrinally significant textual variants and their implications, Wallace is much more careful than Carson when attempting to reassure people that pro-Alexandrian textual criticism poses no doctrinal challenges.  But he still seems willing to make it seem otherwise by holding up Carson’s fuzzy assurances when this subject comes up.  One must jump to footnote #25 in Wallace’s essay to see his actual view:  “No viable variant affects any major doctrine.” 
            But what constitutes a viable variant, and what constitutes a major doctrine?  Is the Sinaitic Syriac’s reading in Matthew 1:16 a viable variant?  Von Soden thought so.10  Is the belief that Mary was a virgin when Jesus was conceived a major doctrine?  And is the inclusion of the name “Isaiah” in Matthew 13:35 a viable variant?  Eberhard Nestle thought so, and regarded it as the original reading.11  Is the Alexandrian reading of Matthew 27:49 a viable variant?  Hort thought so.12  Is the belief that the authors of Scripture did not produce errors a major doctrine?  Or is it, as Wallace has described it, a more peripheral doctrine?  Further examples of doctrinal subjects impacted by textual variants could be listed, such as the importance of fasting, standards for divorce and remarriage, and the bodily resurrection of Christ.      
            Wallace attempted to minimize the differences between the Byzantine Text and the Nestle-Aland text in terms of quantity:  they disagree, he states, “in only about 6,500 places,” most of which do not affect translation or interpretation.  “The majority text and modern critical texts,” he continues, “are very much alike, in both quality and quantity.”  But if the differences are clearly inconsequential, then why are hundreds of thousands of dollars and countless hours of labor being spent to collect and (someday) study digital photographs of New Testament manuscripts by the Center for the Study of New Testament Manuscripts, of which Wallace is the Executive Director? 
            He does not seem to act as if all the differences are a matter of minutiae.  And rightly so, because well over 2,000 variant-units in the Gospels alone affect translation.  Their effects range from the inclusion/exclusion of two large 12-verse passages (Mark 16:9-20 and John 7:53-8:11) to the inclusion/exclusion of entire verses, to the inclusion/exclusion of phrases, to the inclusion/exclusion of words that are pivotal to the sense of the sentences in which they occur (or do not occur).   To speak of the importance of textual variants in terms of quantity, as if to say that no one should worry about a low percentage of variation, is a poor way to state the problem.  It’s like telling people that there are only 6,500 stray cats in the city, and the vast majority of them are harmless.  If 65 of them are rabid it is still a concern.
- Continued in Part Two - 



1 – Wallace’s essay seems to have been written sometime in the 1990’s.  All quotations of Wallace in this essay are from that article, unless otherwise noted.
2 – The recently-produced English translations of the New Testament based on the Byzantine Text are:
            ● World English Version (at ),
            ● English Majority Text Version (at  ),
            ● Analytical-Literal Translation of the New Testament (at ), and
            ● Eastern/Greek Orthodox New Testament (at ),
The second edition of The New Testament in the Original Greek – Byzantine Textform, prepared by Maurice A. Robinson and William G. Pierpont, was released in 2005, with a generous copyright policy:  “Anyone is permitted to copy and distribute this text or any portion of this text,” effectively placing it in the public domain.  RP-2005 is online, formatted as a PDF, with a collection of other free resources, at
            Also worthy of mention are The New Testament – The Original Greek (Koine) Text, at and Wilbur Pickering’s compilation of the f35 text, at and his English translation of it, at , and the Byzantine Greek New Testament, at .
            The official website of the Modern English Version (with a Textus Receptus-based New Testament) is at .
3 – As reported on March 13, 2014, in Christianity Today – see – the  poll, carried out by the Center for the Study of Religion and American Culture at Indiana University-Purdue University Indianapolis (IUPUI), found that 55% of Bible-readers in the United States read the KJV; the NIV was first runner-up, at 19%.  
4 – See Wieland Willker’s comment on Matthew 9:36 in A Textual Commentary on the Greek Gospels, Volume 1 – Matthew.  Willker noticed that εκληλυμενοι is in Erasmus’ text, and in Codex L, but not in codices 1 and 2.
5 – See pages 210-211 of Life and Letters of Fenton John Anthony Hort, Volume 1, at .  According to the sub-headings in the book, Hort made these references to the Textus Receptus as “vile” and “villainous” when he was 23 years old. 
6 – Wilbur Pickering, page 131, The Identity of the New Testament Text, fourth edition.
7 – This sum of 1,838 differences was offered by Wallace in his article Some Second Thoughts on the Majority Text, at , published June 3, 2004.  It should be noted, however, that his tally is based on a comparison of printed texts; if formatted in the format of ancient uncial manuscripts, with contracted sacred names and no spaces between the words, the sum would be somewhat lower.
8 – In this comparison, “Byz” is the Robinson-Pierpont 2005 Byzantine text.  Ehrman’s research in 1996 used the Textus Receptus as a major representative of the Byzantine Text, so to acquire a more accurate picture of the actual implications of Didymus’ readings, it was necessary to manually consult the RP-2005 text reading-by-reading.
9 – See page 47 of George Vance Smith’s short 1881 book, Textsand Margins of the Revised New Testament Affecting Theological Doctrine BrieflyReviewed.  Readers may also consult page 140 of Volume 2 of Life and Lettersof Fenton John Anthony Hort to see Hort’s statement, in a letter written to Westcott in August of 1870, about the “moral damage that would have been done to the acceptance of the Revision by the laity if Unitarians had been outlawed as such.”  The background of the comment is that there had been objections against the inclusion of George Vance Smith on the Revision Committee, but Westcott and Hort had insisted that he be included.
10 – See the opening paragraphs of Bruce Metzger’ article The Text of Matthew 1.16 on page 105 of New Testament Tools & Studies – Philological, Versional, and Patristic, Vol. 10 (1980).
11 – See page 251 of Eberhard Nestle’s 1901 Introduction tothe Textual Criticism of the Greek New Testament:  referring to the variant δια Ησαϊου του προφήτου, Nestle wrote, “It is certainly, therefore, genuine.”  
12 – Hort did not fully adopt the Alexandrian reading of Matthew 27:49 into his text, but stated on page 22 of Notes on Select Readings that there were two possibilities:  either the phrase “may belong to the genuine text of the extant form of Mt,” or “they may be a very early interpolation.”  He concluded:  “We have thought it on the whole right to give expression to this view [i.e., the view that the passage is an interpolation] by including the words within double brackets, though we did not feel justified in removing them from the text, and are not prepared to reject altogether the alternative position [i.e., the view that they are part of the original text of Matthew].”


Unknown said...

Hi James,

Thanks for a great series! I'm eagerly awaiting Wallace's response.

You say, "I do not subscribe to Byzantine Priority", but you do such a profound job of defending this position that I'm really curious what your position is.

James Snapp Jr said...

I explain my approach in the two posts that can be found by searching the blog (using the search-bar at the top) for "Equitable Eclecticism."

Or just read these two posts: