Saturday, March 21, 2020

Testing the Hark Chart

            It is not rare to find, among advocates of the Textus Receptus, the promotion of a chart titled Providential Preservation of the Text of the New Testament.  The central portion of the chart depicts two transmission-lines of the text; brief comments to the left and right of the chart summarize what the chart proposes:  basically, that the text of Antioch is the good text – handed down in its completely pristine form from the apostles to the KJV – while the text of Alexandria is the bad text – corrupted by Gnostics and Arians before being compiled by apostates.
            Let’s test the accuracy of some of the claims that are made in this material.   

Q #1:  Does the early text of the New Testament neatly fall into two forms?
A:  No.  This arrangement describes the base-texts of English versions of the New Testament; the KJV, NKJV, and MEV are basically Byzantine, and the NIV, ESV, NLT, and CSB are basically Alexandrian.  But the texts in Greek New Testament manuscripts and early versions fall into more than two categories:  besides the Byzantine Text (called the Antiochan text in the chart) and the Alexandrian Text, there was also the widespread Western Text, and the more limited Caesarean Text of the Gospels, and mixed texts (in which readings from more than one of the other four forms appear).

Q #2:  Was the city of Alexandria “a place where every deviant sect was represented” and where “Religious corruption and false doctrines were prevalent including Gnosticism, Arianism, pagan philosophy, etc.”?
A:  There is almost no evidence from the 100s to support the idea that the theology of residents of Alexandria was better or worse than the theology of residents of most major cities in the Roman Empire.  Egypt certainly had its share of heretics – Basilides, for example, and the Gnostics of Nag Hammadi.  However, Nag Hammadi is 450 miles to the south of Alexandria – further than the distance from Chicago, Illinois to Cleveland, Ohio.  It would be an enormous assumption to suppose that because Alexandria and Nag Hammadi are both in Egypt, the same heresies were spread in both places to the same extent. 
            Alexandria was not the only city to harbor false teachers.  Valentinus, an infamous heretic, resided at Rome for 30 years.  Marcion was from Pontus (northern Turkey).  Heracleon’s headquarters were probably in Italy.  Porphyry of Tyre, an opponent of Christianity in the 200s, worked in Rome and Sicily.  Mani, founder of Manichaeism, came from what is now central Iraq.  Bardaisan worked in Syria and Armenia.  In the mid-200s, the false teacher Paul of Samosata was bishop of Antioch for eight years.  Nestorius, after being trained in Antioch, was briefly archbishop of Constantinople, c. 430, before being denounced as a heretic by Cyril of Alexandria.  In addition, the book of Acts and the letters of Paul demonstrate that false teachers were active in Greece and Turkey and Israel in the first century of Christianity.  Alexandria, as the second-largest city in the Roman Empire, was exceptionally big, not exceptionally bad.  In addition, the contributions to orthodoxy by people from Alexandria such as Athanasius, and the piety of Pachomian monks in Egypt, should not be overlooked.
Q #3:  Is it an essential doctrine to believe that God has preserved every word of the original text “through the ages in the church”? 
A:  Adherents to the manmade creed known as the Westminster Confession of 1646 affirm that the Old Testament in Hebrew and the New Testament in Greek, “being immediately inspired by God, and, by His singular care and providence, kept pure in all ages, are therefore authentical.”  The Greek texts used by the formulators of the Westminster Confession were essentially Byzantine – specific, they were the printed editions made in the 1500s – and thus, while disputes might conceivably occur regarding differences at certain points in those printed editions – an appreciation of the historical context of this statement in the Westminster Confession might preclude an embrace of a compilation that relies fundamentally on a different form of the text:  two Greek New Testaments that differ at a thousand points affecting translation cannot, strictly speaking, both be considered “pure” forms of the text; both cannot be the original form of the text. 
            On the other hand, the term “pure” may be understood to refer to the retention of the authentic meaning of the original text, not necessarily its exact form.  Many medieval Christians who encountered the Greek New Testament did so when reading lectionaries – books in which the text was divided into segments, for which introductory phrases were added, and in which proper names were also added to identify the characters in the separated passages, and in which parallel-passages were at various points also introduced. 
            Few if any clergymen of the 1500s and 1600s would have denounced lectionaries as if they contained an “impure” form of the text, inasmuch as their meaning was no different than what could normally be found in the unsegmented, continuous text.  Yet, strictly speaking, the segmented and supplemented form of a lectionary is obviously not the original form of the text.  My point here is that a general affirmation that a specific form of the text is “pure,” inasmuch as it retains the didactic content of the original text – that is, it teaches the same thing – is not the same as a statement that a specific form of the text is the original text.
            In addition, I see no reason why anyone should feel obligated to adhere to the Westminster Confession of Faith merely because a previous generation did so.  Its formulators did not intend to make the creed itself a rival to the New Testament’s authority.  I suspect that if they were to have at their disposal the materials that are available today, they would use different words to affirm their allegiance to God’s revelation in the original text of the Old Testament and New Testament – without giving the impression that the foundation for their understanding of the original form of the New Testament (the compilations made in the 1500s, and the evidence upon which they were based) ought to be locked in place forever as the Greek New Testament and become the basis for a new “essential doctrine.”

Q #4 and #5:  Does Psalm 12:7 contain a promise that God will make all of the words of the original text available to Christians on earth in every generation?  And is such a promise fulfilled by the Textus Receptus?
A:  Those who set the KJV and the NIV or ESV side-by-side will notice that this verse has been translated very differently:  in the KJV, when Psalm 12:7 says, “ Thou shalt keep them, O Lord, thou shalt preserve them from this generation for ever,” the natural understanding is that the author is referring to the words of the Lord mentioned in verse 6 – “The words of the Lord are pure words: as silver tried in a furnace of earth, purified seven times.”
            In the NIV (2011), however, verse 7 refers to the needy individuals describes in verse 5 – “You, Lord, will keep the needy safe and will protect us forever from the wicked,” and verse 6 is presented as something parenthetical.   Likewise the ESV says, “You, O Lord, will keep them; you will guard us from this generation forever.”  The NLT seems designed to preclude the interpretation that verse 7 refers to the words of the Lord:   “Therefore, Lord, we know you will protect the oppressed, preserving them forever from this lying generation.”  
            I leave it to others to consider the arguments for both renderings – Kent Brandenburg engaged Doug Kutilek’s interpretation in 2010; Jon Rehurek argued for the NIV’s rendering in 2008; in 2017 Bryan Ross engaged the approach that had been promoted by William Combs in 2000.  Other passages – Proverbs 30:5, Isaiah 40:7-8, Matthew 24:35, etc. – do not leave much room for the idea that the inspired words of the original text of Scripture have passed away.  At the same time, I see nothing in any of those passages that suggests that God is obligated to make all of those words available to any particular believer, or group of believers, in every generation.  To put it another way:  the veracity of these promises is granted but there are three very different ways they can be interpreted:  it is one thing to propose (a) that every word of the original text is perfectly known by our omniscient God.  It is something else to propose (b) that every word of the original text is preserved intact in some extant witness, somewhere.  And it is something else to propose (c) that every word of the original text of the New Testament has been preserved intact in the Textus Receptus in every generation.
            It is that third proposal that the Hark Chart (so called because it was prepared by H. N. Arkell of Hark Ministries) advocates.  However, there is not much evidence that this is the case – and there is plentiful evidence that this is not the case.  Hundreds of reading in the Textus Receptus are not supported by the majority of Greek manuscripts.  For some readings in the Textus Receptus, Greek manuscript-support is very sparse.  And for some readings in the Textus Receptus, there is no Greek manuscript-support at all.  Acts 9:5-6, for example, is part of the Textus Receptus, but, as far as I can tell, it is not in the Greek text of any manuscript made before the 1500s.  
            Although the Hark Chart gives its readers the impression that 5,210 manuscripts stand behind the KJV’s New Testament base-text, this is untrue at hundreds of points where the Textus Receptus has a reading that is not in the Byzantine Textform.   Here are 10 examples of readings in the Textus Receptus that are not supported by the majority of manuscripts in the Byzantine (Antiochan) Text, drawn from the Gospel of Matthew:
            3:8 – most MSS support “fruit,” not “fruits”         
            3:11 – most MSS do not have “and fire” at the end of the verse  
            5:27 – most MSS do not support “by them of old time”   
            5:47 – most MSS support “friends” instead of “brethren”
            6:18 – most MSS do not have “openly” at the end of this verse   
            8:5 – most MSS do not specifically name Jesus here
            8:25 – most MSS do not refer to the disciples here as “His” (αυτου)
            10:8 – most MSS do not have “raise the dead” after “cleanse the lepers”
            12:35 – most MSS do not have “of the heart” here
            26:38 – most MSS specifically name Jesus near the beginning of this verse

            Many other translation-impacting textual contests in which most manuscripts disagree with the base-text of the KJV can be found in a 2018 article by Luke Wayne at the CARM website.   My point here is not to defend all of these majority-readings, but rather to make clear that the KJV’s base-text and the Byzantine (Antiochan) Text are two different things.  The Hark Chart’s claim that “every word” has been passed down from the original New Testament documents, to the Textus Receptus, with the support of 5,210 manuscripts, is not true.

Q #6:  Did Clement of Alexandria and Origen have a strong impact on the New Testament text?
A:  Although Origen, in the first half of the 200s, did some thorough research on the text of the Old Testament – compiling his Hexapla, for example – his comments about textual variants in the New Testament are rather mild; he mentions relatively few textual variants, and occasionally emphasizes how each rival reading can be edifying.
            Clement of Alexandria does not appear to have attempted to standardize the New Testament text either.  He certainly does not appear to have strictly followed the Alexandrian form of the text:  Carl Cosaert, in his 2008 research-book The Text of the Gospels in Clement of Alexandria, acknowledged that in the Synoptic Gospels (Matthew, Mark, and Luke), “no single text-type played a dominant influence on Clement’s text.”          
Q #7:  Do Codex Sinaiticus and Codex Vaticanus contain corruptions due to the influence of false teachers such as Origen?
A:   In Codex Sinaiticus (À), in the first seven chapters of the Gospel of John, the text is rather more Western than Alexandrian.  It is as if the copyists who made À relied on a damaged Gospels-manuscript as the basis for most of the Gospels, but for these chapters, a different exemplar was used.  Bart Ehrman has shown that in this portion of John, À has some affinities with the text of John used by the second-century heretic Heracleon (who was answered, in the 200s, by Origen).  It does not seem improbable to me that the copyists who made À were working at Caesarea, at the same library which Origen had accessed a century earlier.  If that were the case, they might have used the same text of John that Heracleon used (and possibly edited).  In addition, if À were made at Caesarea in the mid-300s, then it seems likely that its copyists worked for bishops Acacius and Euzoius, both of whom were Arian.  That could have had an effect on some decisions about which readings to adopt when encountering different readings in different copies of New Testament books – but this is a difficult scene to bring into focus from such a chronological distance.
            In Codex Vaticanus (B), there are a few readings that suggest Marcionite influence; in Romans 1:16, for example, the word “first” is strangely absent.  However, in the Alexandrian Text as a whole – setting aside readings peculiar to individual manuscripts – while there are some readings that are unquestionably errors (seen, for example, in Matthew 27:49 and Acts 27:37), almost all of them can be explained as having arise by scribal carelessness (for instance, via the confusion of sacred-name contractions) rather than as the result of a doctrinal agenda.  Another interesting feature of Codex B is that its copyists tended not to contract references to the Holy Spirit; this may reflect Binitarian, rather than Trinitarian, theology.  However this does not reduce the importance or value of the Alexandrian Text itself, which is extant in other witnesses.
            To summarize:  while both Vaticanus and Sinaiticus may have emerged from less-than-orthodox backgrounds, relatively few readings in the Alexandrian Text pose any sort of obstacle to Christian orthodoxy.     
Q #7:  Did Westcott and Hort, who were responsible for a heavily Alexandrian revision of the New Testament text released in 1881, undertake that revision with a pro-Roman-Catholic, pro-evolution agenda?
A:  Hort said some things in his many letters that are the basis for valid concerns; he was rather racist; he opposed democracy; he expressed sympathy with Roman Catholicism to an extent, and he was far from conservative.  At the same time, there is not much evidence that any of that had an effect on his text-critical research.  Many elements of Hort’s approach were already expressed – before European scholars were aware of Codex Sinaiticus – by J. J. Griesbach (d. 1812) and by Granville Penn (d. 1844).  In addition, Samuel Tregelles (d. 1875), whose theological school of thought tended to be much more conservative than Hort’s, issued a multi-volume compilation of the Greek New Testament which anticipated in many respects the compilation later issued by Westcott and Hort. 
            In short, while we cannot read the minds of Westcott and Hort, it seems precarious to attribute a doctrinal agenda to their text-critical decisions when we can observe advocates of other doctrinal schools of thought making the same, or very similar, decisions.

Q#8:  Compared to the traditional text, do modern versions such as the NIV and ESV add 306 words that should not be in the New Testament text, and omit 2,987 words (including 20 verses) that should be there?
A:  This claim is based on data in E. W. Fowler’s book Evaluating Versions of the New Testament, which I do not have, so it is not easy at present to discern exactly what words the maker of the Hark Chart had in mind.  An exact and permanent account of the differences between the Textus Receptus and the critical text is not easy to obtain, because the Nestle-Aland edition of the critical text continues to undergo revision.  However, an estimate provided by Jack Moorman, a prominent promoter of the KJV, in 1988, sums up the differences:  the Nestle-Aland text is shorter than the Received Text (i.e., the Textus Receptus) by 2,886 words.  That does not mean that there are only 2,886 differences:  in many places, both the Textus Receptus and the Nestle-Aland compilation have a word, but it is not the same word.   KJV-advocate D. A. Waite offered some more detailed calculations:  using the Textus Receptus as the basis of comparison, there are 5,604 changes in the 1881 revision of Westcott & Hort, of which 1,852 are omissions, 467 are additions, and 3,185 are other changes (such as word-exchanges and transpositions). 
            Thus, there seems to be some incongruity about just how much the Textus Receptus differs from the KJV’s base-text:  Fowler apparently claimed that 2,987 words – just a little less than 3,000 – are omitted; Waite claimed that 1,852 words – fewer than 2,000 – are omitted.  Perhaps different editions of the critical text were used.   No matter which report one favors, the differences are significant.
            A key factor to consider in such statistics is the use of the Textus Receptus as the basis of comparison.  What if, instead, one used the readings favored by the majority of Greek manuscripts as the basis of comparison?  In that case, the Textus Receptus would be found to deviate from the majority-reading in over a thousand places.  (Daniel Wallace has claimed to have found 1,838 differences between the Textus Receptus and the Hodges-Farstad Majority Text; however many differences in his list (such as differences in the spelling of David’s name, which is often contracted in manuscripts) are untranslatable.  Michael Marlowe has offered a better estimate of 1,005 differences.)  
            The Hark Chart gives readers no hint that the Textus Receptus contains 1,005 readings which diverge from the Byzantine (Antiochan) Text.  Instead, it makes it looks as 100% of the Textus Receptus is allied with 5,210 manuscripts, and with the Peshitta, and with the “Old Latin and Syriac of the Originals” and with papyri from 150-400.  Such an impression is far different from what one finds in the Peshitta and the Old Latin and in the papyri when they are studied in detail; they contain many readings that diverge from the Textus Receptus.       
            In conclusion:  the Hark Chart is not an accurate depiction of the transmission of the text of the New Testament.  It is KJV-Only propaganda.

Readers are invited to double-check the data in this post.



Tadodot Pedodot said...

Thank you very much James Snapp Jr. for the informative article. I encountered that chart once because of a Baptist that says that the KJV is an accurate translation in english. I wanted to know the professional opinion of a textual critic so I can learn from both sides their idea regarding about that chart.

But all the more, thank you for the informative article.

God bless

Maurice A. Robinson said...

Also keep in mind that appeals to the 5000+ MSS at all points of variation is equally invalid, since even in the Gospels at any point there are usually less than 2000 MSS extant, in the Acts and Epistles less than 600,and only about 300 in Revelation. Inflated claims do nothing to improve the (generally) false assertions being made in such cases.

Steven Avery said...


Just commenting on the Hark Chart, this is something that I have been writing about for c. 10 years, this chart and others that are similar "two lines" or "two streams" analysis.

James has occasionally pointed out the problems, even back to 2014 or earlier. And two AV-contras, Doug Kutilek and Rick Norris, really helped in the correction, by pointing out the fact that you cannot put the Old Latin in a different line than the Vulgate.

The original problem goes back to Benjamin Wilkinson (1872 -1968), who mangled the excellent theories of Frederick Nolan (1784-1864)to try to fit them in with the Adventist (and sometimes Baptist) super-focus on the Waldenses. Thus they wanted the Waldenses to be on the good Old Latin line. All this was passed down by David Otis Fuller (1903-1988) to many of the name AV defenders. Note, this is not in the writings of Edward Freer Hills (1912-1981).

Bryan Ross neatly followed up and expanded on my writings:

The Two Streams of Bibles Model of Transmission: Its Origins & Accuracy

Paul B said...

I submit there is no Textus Receptus. Show it to me. Is it found in some library somewhere? If so where? In fact a comparison of William Tyndale's translation as found in the Coverdale Bible shows that much of the King James Bible came from that and not from some ancient manuscript that by astonishing co-incidence ended up being translated word for word the same as Tyndale did. It's mathematically beyond mere chance.