Answering Islam - A Christian-Muslim dialog

Miscellaneous Issues in NT Textual Criticism



In our previous article we discussed some of the major issues in NT textual criticism that were relevant to the debate between Christians and Muslims regarding Biblical corruption.  While I think we touched upon the most basic-yet-pertinent issues in that paper, there are nevertheless a few other issues that have been raised by certain textual critics that merit more attention.  It is these issues that I deliberately did not discuss in the main paper due primarily to their complexity.  This article serves as an extended supplement to the previous article.  The viewpoints/theses expressed by four scholars in particular I find worthy of discussion:  1) Bart Ehrman; 2) David C. Parker; 3) Eldon J. Epp; 4) William Peterson.   In this article we will discuss the pertinent findings of these textual critics and how they may relate to the Christian-Muslim debate.  I will include at the end a brief discussion of textual criticism as it was relevant to the ancient Christians as well.

Bart Ehrman – “Orthodox Corruption of Scripture”

Given Ehrman’s immense popularity and prominence in the text-critical world, many are already quite aware of the arguments that he espouses in this particular book (and also to a lesser extent in his “Misquoting Jesus”).  Essentially, this represents an examination of the deliberate textual variants made by some Christian scribes in the manuscript tradition in order to make certain theologically-ambiguous texts appear more orthodox.  Ehrman’s thesis undeniably renders important observations relevant to various streams of New Testament scholarship.  It also gives us some insight into the social world of the early Christians [cf. e.g. Ehrman 1995].

The alterations that Ehrman focuses upon are those that he deems to have been made in order to render certain texts less susceptible to heretical interpretation.  The four heresies under focus are:  1) Adoptionism (the belief that Jesus was purely human and was declared “Son of God” by the Father at some point in his life, such as e.g. his baptism); 2) Separationism (the belief that Jesus, a righteous human conceived in ordinary fashion, received the “divine Christ“ who empowered him until the crucifixion, when the “divine Christ“ left Jesus and ascended to heaven); 3) Docetism (the denial of Jesus’ actual humanity, rather that he only appeared to be human and undergo suffering); 4) Patripassianism (an ancient form of modalism, in this case that Jesus was God the Father in human flesh).  In sum, many of the arguments Ehrman makes are generally accepted as convincing by scholars while others are more open to criticism.  We will discuss 3 such examples here.

Luke 3:22

and the Holy Spirit descended on him in bodily form like a dove. And a voice came from heaven: "You are my Son, whom I love; with you I am well pleased." (NIV)

Ehrman argues persuasively that the original text of the latter part of the above verse read:  “You are my Son, today I have begotten you” and that the other text was an orthodox corruption that was created in order to remove the suggestion that it was at Jesus’ baptism that he was “adopted” as the Son of God (Ehrman 1993; 62-67).  J. Neville Birdsall, however, argues that another explanation may best account for the change: 

…we may observe first of all, that the quotation of Psalm 2.7 in a christological context remained without any excision or change in Acts 13.33 and in Hebrews 1.5 and 5.5.  The substitution of the phraseology of Matthew or Mark in the Lukan passage more probably finds its explanation in harmonization, which was evidently rampant in the transmission of the text from a very early date (Birdsall 1994; 460-461).

Luke 9:35

A voice came from the cloud, saying, "This is my Son, whom I have chosen; listen to him."

While our prior example concerned the heavenly voice at Jesus’ baptism, this example concerns the same voice at the Transfiguration.  In the Markan parallel (9:7), we have “This is my Son, whom I love”, and the Lukan text underwent changes to conform to this reading.  The Lukan reading “whom I have chosen” is found in the earliest and best witnesses, including Codices Vaticanus and Sinaiticus and is considered the original reading by the scholarly consensus (Ehrman 1993; 67-68).   Now, one could once again argue that the Lukan reading was harmonized to that of Mark and Matthew.  However, Ehrman states the following, which could also serve as a reply to Birdsall’s criticism in the previous example:

If this were the case, one would expect the alternative process to have happened as well--that is, harmonizations of Mark and Matthew to the text of Luke.  The magnitude of the textual changes in Luke, coupled with the virtual absence of such changes in Matthew or Mark, suggests that the change was made for doctrinal reasons pure and simple--to eliminate the potentially adoptionistic overtones of the text (ibid 68).

Hebrews 2:9

But we see Jesus, who was made a little lower than the angels, now crowned with glory and honor because he suffered death, so that by the grace of God he might taste death for everyone.

The previous two examples we reviewed briefly were alleged “anti-adoptionist” corruptions whereas our final example is allegedly an “anti-separationist” corruption.  Per Ehrman, despite a relative paucity of external support, the original reading states: “apart from God he might taste death” rather than “by the grace of God…” (cf. ibid. 145-150).

Daniel Wallace argues against Ehrman’s conclusions on the matter:

Translations are roughly united in how they treat Heb 2.9b. The NET is representative: “by God’s grace he would experience death on behalf of everyone.” Ehrman suggests that “by God’s grace”—χάριτι θεου'—is a secondary reading. Instead, he argues that “apart from God,” or χωρὶς θεοῦ, is what the author originally wrote. There are but three Greek manuscripts that have this reading, all from the tenth century or later. Codex 1739, however, is one of them, and it is a copy of an early and decent manuscript. χωρὶς θεοῦ is also discussed in several fathers, one Vulgate manuscript, and some copies of the Peshitta.  Many scholars would dismiss such paltry evidence without further ado. If they bother to treat the internal evidence at all, it is because even though it has a poor pedigree, χωρὶς θεοῦ is the harder reading and thus may require some explanation, since scribes tended to smooth out the wording of the text. As well, something needs to explain the several patristic citations. But if a reading is an unintentional change, the canon of the harder reading is invalid. The hardest reading will be a nonsense reading, something that cannot be created on purpose. Although χωρίς is apparently the harder reading, it can be explained as an accidental alteration. It is most likely due either to a ‘scribal lapse’ in which an inattentive copyist confused χωρίς for χάριτι, or ‘a marginal gloss’ in which a scribe was thinking of 1 Cor 15.27 which, like Heb 2.8, quotes Ps 8.6 in reference to God’s subjection of all things to Christ. (Source)


So hopefully the reader that is unfamiliar with Ehrman’ thesis may gain an appreciation from the examples discussed for the kinds of argumentation he offers in his book.  We may state quite reasonably that Ehrman’s overall case of “Orthodox Corruption” is an established phenomenon, although as we’ve seen certain cases are open to debate.  

Of course, the presence of deliberate alterations in the tradition does not necessitate the conclusion that the original text cannot be reconstructed.  In fact, Ehrman generally seems to presuppose that we know what the original text said (hence he was able to demonstrate that an alteration from the original was made in the first place).  Moises Silva summarizes the issue well:

And my exhibit A is Bart Ehrman’s brilliant monograph The Orthodox Corruption of Scripture, which I consider one of the most significant contemporary works on biblical scholarship.  Although this book is appealed to in support of blurring the notion of an original text, there is hardly a page in that book that does not in fact mention such a text or assume its accessibility.  “Why is such-and-such a reading in Mark a later corruption and not original?  Because Mark (authorial intent!) would not likely have said such a thing.”  Indeed, Ehrman’s book is unimaginable unless he can identify an initial form of the text that can be differentiated from a later alteration.  (Silva 2002; 149; emphasis added)

The calculations made by David Parker in his review of Ehrman’s book are also revealing:

What impact does Ehrman’s study have on the shape of the text?  He discusses 179 units of variation in the main text of the book, and a further 36 in footnotes.  In these 179, he agrees with Nestle-Aland26  in 162...But the great majority of these are ’orthodox corruptions’ found in only one or a few manuscripts, or in a version.

Generally, Ehrman feels no need to discuss the question of the original text of such passages.  In 29 places he feels that extended discussion is necessary to substantiate his case for the original text.  And it is in 17 of these 29 places that he follows a different text from that of Nestle-Aland26.  The places of disagreement are Mark 1:1 (N-A26 has the words in single square brackets); Luke 2:40; 3:22; 22:19-20; 43-44; 24:6, 12, 36, 40, 51 f.; John 1:18, 34; Rom. 8:34 (N-A 26 has the word in single square brackets); I Cor. 10:9; 15:15; Heb. 1:3; 2:9.  The 12 passages where they agree are Matt. 1:16, 18 (twice); 24:36; 27:49; Mark 15:34; Luke 9:35; 23:42; John 19:5; Gal. 2:20; I John 4:3; 5:18.  This degree of divergence is particularly striking when one recalls that Ehrman shares the editors’ belief in the value of external evidence.  This is not the disagreement of a Kilpatrick or an Elliott.  It comes from the sober reflection of one who believes in ‘good’ manuscripts but cannot always follow them (Parker 1994 Book Review; 707-708; emphasis added).

Parker goes on from here to state that the textual divergences of Ehrman from Nestle-Aland “makes the initial chapters of Mark and Luke much more clearly ‘adoptionist’ than does the N-A26 text.” (ibid. 708)  We’ll discuss the phenomenon of the “adoptionist readings” a bit more below in the William Petersen section.

Further, while Parker mentions Ehrman’s valuing of the external evidence, Daniel Wallace, with the Hebrews 2:9 passage in mind, chides Ehrman for his going the way of a “Kilpatrick or Elliott” (I.e. thoroughgoing eclecticism) with his quest to find corruption of the text:

Second, Ehrman’s text-critical views are getting dangerously close to rigorous eclecticism. The external data seem to mean less and less to him as he seems to want to see theological corruption in the text (Source)

Yet it is not the differences of Ehrman’s conclusions with that of Nestle-Aland I find to be most intriguing, rather his agreements with Nestle-Aland.  In this case Ehrman agrees with Nestle-Aland on the original text in 162 of the 179 texts he discusses where there exists alleged orthodox corruption.  This is an interesting state of affairs regarding the textual critic that serves as champion to some overzealous critics in arguing against the preservation of the New Testament text.

Those interested in further discussion of Bart Ehrman’s writings are encouraged to visit the following web pages:
(cf. also now Holding 2009; 129-134)

The relevance of Ehrman’s thesis for our purposes is fortunately easy to summarize.  While Ehrman successfully demonstrates in many cases (some other of his arguments may be more questionable) that scribes altered texts in order to bring them more clearly in line with orthodoxy, his thesis more or less assumes certainty of the original text.  Thus while Ehrman’s work is an important contribution to textual criticism and New Testament scholarship as a whole, it does not obscure our quest for the original text.

David C. Parker - “The Living Text of the Gospels”

In Parker’s excellent and informative study, “The Living Text of the Gospels”, the author demonstrates with success the fluidity of the written Gospel tradition in the first few centuries after the initial penning of the NT.  That the NT textual tradition, and especially that of the Gospels, was prone to such alterations we acknowledged in the relevant section of our main article (cf. “The Living Text“).  The volume under discussion is primarily concerned with demonstrating this fluidity through the use of multiple test-cases.  Among the primary examples discussed by Parker are the NT traditions of “The Lord’s Prayer”, the various sayings on marriage and divorce, the story of the woman taken in adultery (when present located most often at John 7:53-8:11 in the MSS tradition), the Markan appendix (I.e. 16:9-20), and the so-called “Western non-interpolations” located in the closing chapters of Luke’s Gospel.   

While I think Parker persuasively demonstrates the fact that the Gospels were prone to alteration in the early centuries, it is perhaps the philosophy derived from his study with which, at points at least, I would disagree.  Consider the following excerpt from the conclusion of Parker‘s study:

It may seem that the argument is moving towards the conclusion that the quest for the earliest forms of the text is worthless.  But it is not, because the attempt to recover early text forms is a necessary part of that reconstruction of the history of the text without which, as this book has been at some pains to demonstrate, nothing can be understood.  But, even if very ancient, even the original texts, could be recovered or reconstructed, the ambiguity of the definitive text would not be at an end.  Even though many historical questions would be answered, our interest in the history of the text would not cease.  Theologically, there would be no resolution of the central problem.  For the heart of the matter is that the definitive text is not essential to Christianity, because the presence of the Spirit is not limited to the inspiration of the written word.  We have already approached this from the point of view of a false distinction between Scripture and tradition.  Examining it in the present context, one is struck by the fact that a belief in single authoritative texts accords to the Spirit a large role in the formation of Scripture, and almost none at all in the growth of the tradition.  Once the distinction has been abolished in the way that we have attempted above, it is possible fully to acknowledge that the very life and whole life of the church is in the Spirit. 

When we accord significance to all the text forms as a part of the tradition, are we thereby affirming a pluralism in which contradictory forms of sayings of Jesus, for example, on divorce, have equal weight?  No, because we are not attempting to ascribe to all forms of the tradition the authority which traditionally has been accorded to only one.  Instead, we argue this:  the church came into being through the Spirit, as the community of the Spirit.  The oral and written tradition together were and remain a principal element in the church’s finding its calling.  But the tradition is manifold.  There are four quite different Gospels, none with a claim of authority over the other three; there is no authoritative text beyond the manuscripts which we may follow without further thought.  There is a manifold tradition to be studied and from which we may learn.  But once that is done, the people of God have to make up their own minds.  There is no authoritative text to provide a short-cut.  Difficult though it often is in practice to accept such a situation, it at least allows us to find an alternative approach to the multiplicity of variant readings which represent the divergent interpretations of early Christianity.  Rather than looking for right and wrong readings, and with them for right or wrong beliefs and practices, the way is open for the possibility that the church is the community of the Spirit even in its multiplicities of texts, one might say in its corruptions and its restorations.  Indeed, we may suggest that it is not in spite of the variety but because of them that the church is that community. (Parker 1997; 211-212; emphasis added)

Throughout the book Parker argues that texts changed (and perhaps were even penned originally by the evangelists in such a way as to deviate from the oral tradition) in order to accommodate new situations by church communities.  Can we accept Parker’s premise that such variant readings were the result of inspiration from the Spirit?  Perhaps this may be argued to a certain extent.  We saw in Appendix I of the main article that oral tradition does not always attempt to reproduce wording in a precise manner.  That this is also the case with the traditions underlying the Gospels is evident simply from a side-by-side reading of various pericopae in the Synoptic Gospels.  If variation was the norm in ancient oral tradition, then perhaps some scribes would have felt free to alter the MSS in a similar manner.  Now, that being the case, I think we can only carry this argument so far.  If the variations of oral or written tradition remain true to the original utterance (in this case of Jesus), even if not in identical wording, then I don’t think it unreasonable to posit inspiration by the Holy Spirit.  On the other hand, variants introduced that result in contradictory facts or theology to the original cannot be considered the result of inspiration by the Holy Spirit [1].  With that in mind let’s consider some of Parker’s examples. 

Marriage and Divorce

Parker lists 7 variants of Mark 10:11-12 regarding marriage and divorce in the manuscript tradition.   We will list those variants represented by the major text-types as they are translated by Parker:  1) Caesarean; 2) Alexandrian; 3) “Western”; and 4) Byzantine:

1) If a woman divorces her husband and marry another, she commits adultery; and if a man divorces his wife, he commits adultery.

2) Whoever divorces his wife and marries another commits adultery against her, and if she, divorcing her husband marries another, she commits adultery.

3) Whoever divorces his wife and marries another commits adultery against her, and if a woman goes out from her husband marry another, she commits adultery.

4) Whoever divorces his wife and marries another commits adultery against her; and if a woman divorces her husband and be married to another, she commits adultery. 

Although the differences are there, only one potentially significant difference exists among the four variants.  In the Caesarean version, it is said that a man commits adultery merely by divorcing his wife, not by divorcing and marrying another.  Two of the remaining three variants listed by Parker also do not render important differences to the above.  The third variant, however, that is found in three Old Latin manuscripts, adds to the common double-formula of the other versions the phrase “also he who marries a woman divorced from her husband commits adultery.” (ibid. 78-79)

It is more than the differences that catch our attention, however. Parker points out that the Markan version(s) all indicate the possibility of a woman divorcing her husband (a reality in the pagan world but not in Judaism) [ibid. 79].  It is thus probable that the author of Mark adapted the original saying of Jesus to his Roman audience.  Assuming that Jesus could not have uttered the saying about women divorcing their husbands, as he preached almost exclusively in a Jewish context, could Mark’s adaptation have been Spirit-inspired?  Since Mark here is making a quite reasonable extrapolation of Jesus’ teaching in order to address a pagan target-audience, I see no reason why this could not be the case.  Of course, given the fact that Jesus in all likelihood presented the same teachings on multiple occasions (with this teaching being one that likely was oft-repeated), it may be a hazardous assumption to make that he never presented it in such a way so as to be appreciated by non-Jewish hearers (though of course we acknowledge the fact that Jesus’ target-audiences were his Jewish contemporaries, cf. Matthew 10:5-6) [On Jesus repeating teachings in variant forms see Appendix 1 of the primary article].

The parallel sayings on marriage and divorce found in Matthew 5:27-32; 19:9; and Luke 16:18 present additional issues worthy of discussion since differences exist among not only the variant readings of each passage, but also between the various passages themselves.   Perhaps the most obvious issue is the exception-clause found in the Matthean versions.  The exception-clause allows divorce only on the grounds of porneia (I.e. marital infidelity) [cf. ibid. 80-94].  In comparing the Matthean and Markan versions, the question then arises:  Is divorce acceptable on the grounds of marital infidelity or on no grounds whatsoever?  It may be that Matthew makes explicit what would have been implicit to ancient readers of Mark and Luke (cf. here). 

Regardless, this issue need not detain us any longer for a couple of reasons.  For one, this would be more of an issue of a potential contradiction between what was originally written by Mark and Matthew rather than a text-critical problem (as the exception-clause certainly was contained in Matthew’s original text).  Second, while this is an important issue regarding Christian living and morality, it is irrelevant to our purposes in this study.  Here we are concerned with our ability to get back to the original text.

The Lord’s Prayer

It is well known that the Matthean and Lukan versions of the Lord’s Prayer contain substantial differences.  Below we will list the RSV translations of the relevant texts as printed in Parker’s book:

Father, hallowed be thy name.  Thy kingdom come.  Give us each day our daily bread; and forgive us our sins, for we ourselves forgive every one who is indebted to us; (Luke 11:2-4)

Our Father who art in heaven, Hallowed be thy name.  Thy kingdom come, Thy will be done, On earth as it is in heaven.  Give us this day our daily bread; And forgive us our debts, As we also have forgiven our debtors; and lead us not into temptation.  But deliver us from evil. (Matthew 6:9-13) [cf. Parker 1997; 50]

Parker examines 7 variants of the Matthean version, the most significant feature of which is a Doxology in which the evidence suggests was eventually interpolated into the text of Matthew (reading “Because thine is the kingdom and the power and the glory for ever. Amen” with variations).  The best attested version found In such important witnesses as Codex Sinaiticus, Codex Vaticanus, Codex Bezae, Family 13, most Old Latin manuscripts, and the Vulgate do not contain the Doxology and vs. 13 ends with “evil” (as above) [ibid. 54].  Parker lists the other variants and how they differ from the authentic text as follows:

1 the shortest text (there is no Doxology, and the text ends with ’evil’)

2 ’Amen’ after ’evil’

3 the Doxology in the form ‘Because thine is the kingdom and the power and the glory for ever. Amen.’

Readings 4-7 are all variants on 3:

4 the Doxology, adding ‘of the Father and of the Son and of the Holy Spirit’ after ‘glory’

5 the Doxology, omitting ‘the kingdom and’

6 the Doxology, omitting the Amen

7 the Doxology, omitting ‘for thine is the power for ever and ever’ (ibid. 54)

It is the shorter Lukan version, however, that was subject to expansion in order to harmonize it to the Matthean version.  The “short version” (different in the addition of the last line from the RSV translation) is almost certainly the authentic one, being attested by key Alexandrian manuscripts P75 and Codex Vaticanus, along with Origen and the Sinaitic Syriac (ibid. 60-63).  It reads as follows:

Father, Hallowed be thy name, Thy kingdom come, Give us each day our daily bread, And forgive us our sins, For we ourselves also forgive each one in debt to us, And do not bring us into temptation.

The text of the “Western” Codex Bezae reads as following:

Our Father who art in heaven, Hallowed be thy name upon us, Thy kingdom come, Thy will be done, As in heaven so on earth, Our daily bread give us today, And forgive us our debts as we also forgive our debtors, And do not bring us into temptation, But deliver us from evil.

This version brings the text much closer to that of Matthew’s Gospel, though apparently there are differences here from the parallel text of Matthew in the same manuscript, according to Parker the most important worth mentioning is the addition of “upon us” after “Hallowed be thy name” in the Lukan version.  Next is the reading found in the vast majority of MSS, the Byzantine form:

Our Father which art in heaven, Hallowed be thy name.  And come thy kingdom.  And give to us bread continual of every day.  And forgive us our sins; and may we also forgive every one that is indebted to us.  And bring us not into temptation; but deliver us from evil.

Parker also briefly discusses another reading, the earliest evidence of which is found in a text of the writings of ancient theologian Gregory of Nyssa of Asia Minor in the late 4th century.  He indicates that instead of “Thy Kingdom come”, Luke wrote “May Thy Holy Spirit come upon us and purify us”.  This is also found in a couple of 11th and 12th century manuscripts (cf. ibid. 66-67). 

And so Parker demonstrates how the shorter version of Luke, preserved in the important Alexandrian witnesses as well as some others, was eventually harmonized in the manuscript tradition to be more in line with the fuller Matthean version.  Thus we have another demonstration of how the NT MSS tradition was a “living tradition”.  Given that sound principles of textual criticism, however, can strip away the layers of text added to Codex Bezae and the Byzantine text in order to determine which text was original, this is once again an issue more relevant to why the original texts of Matthew and Luke are different at this point.  Once we understand the nature of oral tradition, I think the answer becomes clear.  Not only was it not essential that the evangelists preserved given sayings of Jesus in precise wording, but it is quite likely that Jesus himself uttered many of his characteristic sayings multiple times and with multiple variations.  It is doubtful that we are going out on too much of a limb in suggesting in particular that he would have given his Instructions on “how to pray” many times to different groups of listeners.  With this being the case there would have been many forms of this prayer in circulation with differing combinations of elements contained among the various forms.  Upon reading the various forms listed above found in Matthew and Luke (as well as the variant readings of each), it becomes clear that most of the differences are trivial.    

The “Minor Agreements” and Conjectural Emendation Revisited

Another of the chapters in Parker’s book is devoted to a discussion of the so-called “Minor Agreements” in Matthew and Luke and a theory upon how these can be accounted for apart from source-critical theories.  The dominant paradigm in NT Source Criticism asserts that Mark was the first Gospel penned.  Mark’s Gospel was subsequently used as a source for Matthew and Luke.  A second source, labeled by contemporary scholarship simply as Q (short for the German Quelle, meaning “source“), was also thought to be used by Matthew and Luke.  Thus the material peculiar to Luke and Matthew that finds no parallel in Mark is thought to represent Q material.  However, there are some texts in Matthew and Luke that are paralleled in Mark in which the wording of Matthew and Luke are the same, yet is different from the wording in Mark.  Such texts are referred to as the “Minor Agreements”.  This anomaly is typically explained through source-critical theories.  For example, it is sometimes proposed that perhaps Matthew and Luke utilized a version of Mark’s Gospel (i.e. Ur-Markus) that existed prior to the revision that resulted in our current Gospel of Mark.  Another explanation (one that I personally find more likely) is that there exists some overlap between Mark and Q, and that in the case of the “Minor Agreements”, Luke and Matthew are following Q rather than Mark. 

Parker proposes a different hypothesis.  Prior to delving into it let us list a couple of examples that he gives in the text.  We will also use Parker‘s translation.  The following text is the answer that Jesus gave the disciples when the latter asked him why he spoke in parables:

And he answering said that ‘To you it has been given to know the secrets of the kingdom of heaven but for those it has not been given… (Matthew 13:11; emphasis added)

And he said ’To you it has been given to know the secrets of the kingdom of God; but for others they are… (Luke 8:10; emphasis added)

And he said to them, ’To you has been given the secret of the kingdom of God, but for those outside everything is… (Mark 4:11; emphasis added) 

In this particular example the main difference is simply the use of the plural for “secret” in Matthew and Luke against Mark (for discussion see Parker 1997; 103-112).  We’ll consider another example that Parker discusses.  Again we use Parker’s translation of the most widely-accepted original texts of each:

And some began to spit on him and to blindfold his face and to strike him and to say to him ‘Prophesy.’ (Mark 14:65)

Then they spat in his face and struck him, and some slapped him saying ‘Prophesy to us, Christ, who is it who struck you?(Matthew 26:67; emphasis added)

And the men who were holding Jesus mocked and beat him, and blindfolding him asked him saying ’Prophesy, who is it who struck you?’ And they spoke many other things against him, reviling him. (Luke 22:63-65; emphasis added)

Aside from the issue of the “minor agreement” is the question that arises when considering the Matthean version.  We understand why the mockers would ask Jesus who struck him if he was blindfolded (as we are told by Luke), but in Matthew’s version we are not told this.  According to Parker, it has been proposed by B. H. Streeter and C. H. Turner that perhaps Matthew’s Gospel did not originally contain the “who is it who struck you?” part of the text, that it found its way into Matthew’s Gospel through harmonization, even though all MSS of Matthew contain this question.  And so we are back to the concept of conjectural emendation (on which see the relevant section the main article).  Like Michael Holmes, David Parker believes that conjectural emendation might be necessary on rare occasions.  On this he writes the following:

Conjectural emendation is a weapon that, even if it need not be used, should not be surrendered.  The possibility that there are places where no manuscript preserves the original reading must continue to be entertained.  That errors were made so early as to affect all surviving copies may even be taken as more probable than possible.  So, we do not deny that conjecture may be required. (ibid. 115)


There is one set of circumstances under which conjecture may be required:  when all the manuscripts present texts which do not make sense.  Of course, we shall have to be very careful to be sure that what is nonsense to us must have been nonsense to the author.  But the requirement that the writer be allowed to make sense is paramount.  The nineteenth-century editor Haupt went so far as to claim that ‘If the sense requires it, I am prepared to write Constantinopolitanus where the MSS have the monosyllabic interjection o.’  The force of this can hardly be denied.  If we read the sentence ‘The Nicene Creed is also known as the Niceno-O! Creed’, we would quickly begin to cast around for another reading, and a little research would come up with an emendation.  The text-critical approach to ‘Who is it who struck you?’ must be to set aside the Synoptic Problem, and to deal with the question whether or no (sic) the commonly printed text of Matthew makes sense.  To quote Housman quoting Haupt again, ‘The prime requisite of a good emendation is that it should start from the thought; it is only afterwards that other considerations are taken into account.’ (ibid. 115-116, emphasis added; quotation from Housman, ‘The Application of Thought’, p. 77; cf. the whole discussion in ibid. 113-116 on Parker‘s views of conjectural emendation)

In the present example, Parker believes that emendation is unnecessary.  Following Michael Goulder, Parker raises the possibility that the author of Matthew simply took the blindfolding incident for granted, being aware of the fact from knowledge of the tradition, yet for whatever reason did not record it.  Since we cannot say that there is no way that “Matthew believed himself to be writing sense,” Parker states that “The text is difficult, but not impossible.”  Hence Parker rejects the need for emendation in this case (ibid. 116-117). 

Parker’s proposed solution to the “Minor Agreements” is to move away from the typical “documentary solution”.  He writes:

But a documentary solution requires more than the degree of detail needed to know Mark from Matthew.  It requires published editions, in which every last word, syllable and letter is known.  It is this discernible, published precision which is lacking.  The reason for the lack is not -- as it might seem I was about to conclude -- that we do not have the evidence to recover precisely what the evangelists wrote.  It is that the comparison of published editions assumes, in its two-dimensional diagrams, that there is a single point of contact between two texts, for example, the single contact when Matthew copied Mark, and there was an end of the matter.  I am proposing a three-dimensional diagram, in which the third dimension represents a series of contacts between texts each of which may have changed since the previous contact.  For example, Matthew copies bits out of Mark in reproducing a tradition; then a later copy of Mark is enriched by some of Matthew’s alterations; and next a copy of Matthew (already different from the one we began with) is influenced by something from the also changed Mark.  Add in Luke, and oral tradition, and any other sources that might have been available, at any points in the development that you please, and you have a process a good deal less recoverable than any documentary hypothesis.  It is not at all the orderly business we had hoped, and looks instead like molecules bouncing around and off each other in bewildering fashion. (ibid. 121; emphasis original)

Now, Parker does not deny the importance of source criticism in general (cf. ibid. 122-123), but simply argues that such subtle differences as the “Minor Agreements” may not be best explained in such a manner.  I don’t think Parker’s proposal represents the most probable origin of the “Minor Agreements”.  Textual criticism tells us that the original text of Mark most likely read the singular “secret” while Luke and Matthew originally read the plural “secrets”.  Did both Luke and Matthew utilize copies of Mark where Mark’s original “To you has been given the secret” was changed to “To you it has been given to know the secrets”?  Furthermore, this question would have to be multiplied to account for all of the various “Minor Agreements”.  Thus I think that Parker’s theory not only is speculative, but is also improbable.  It is more likely that Matthew and Luke were favoring another source to Mark such as Q or (as Parker alludes to above) oral tradition in order to account for the “Minor Agreements”.         

Parker’s book also contains chapters regarding the different endings of Mark’ Gospel (a subject we discussed in the main article), the propensity for verses from the final 3 chapters of Luke’s Gospel to undergo scribal alteration (included here is a discussion of the “Western non-interpolations”), and a chapter on the story of the woman taken in adultery (typically found in John 7:53-8:11).  He also writes a short chapter regarding the transmission of the Fourth Gospel where he proposes that perhaps conjectural emendation on a grander scale may be reasonable on two particular issues:  the originality of the last chapter in John’s Gospel and whether or not chapters 5 and 6 were initially inverted (ibid. 175-181).  We’ll revisit this former issue in the next section.

And so what should we conclude from Parker’s thought-provoking study?  Again, I think it is fair to conclude that Parker is successful in demonstrating that the NT textual tradition was prone to textual variation from the earliest centuries.  However, it should be noticed that the original text (or as Parker prefers to put it in a more recent publication, the “earliest attainable text” (cf. Parker 2008; 338)) [2] can be deciphered from the cluster of variants in all of the major examples that Parker discusses through the sober application of text-critical tools and principles [3]. 

Thus, as with Bart Ehrman, Parker’s study does not compel one to the belief that we should doubt our ability to reconstruct the original text.  It serves rather as a valuable discussion of some of the important changes in the manuscript tradition of the early centuries of the Christian era.

Eldon Jay Epp – “The Multivalence of the Term ‘Original Text’ in New Testament Textual Criticism”

The primary goal (though not the only goal) of NT textual criticism has classically been the establishment of the most probable original text, at least in as much as the task can be accomplished.  In our primary article we argued that approximately 99% of the wording of the original text of the New Testament can be reconstructed with virtual certainty.  This estimation is the result of a calculation made by NT textual critic Daniel Wallace that is based upon the United Bible Societies critical edition of the NT and the critical apparatus established by the scholars of the UBS textual committee with consideration of the number of B, C, and D readings (cf. the “Playing the Percentages“ section of the main article).  However, it seems a puzzling ordeal to speak of the virtual certainty of an “original text” if such a term can be defined in multiple ways. 

It is with this in mind that we consider the relevant article of Eldon J. Epp.  As the article’s title would imply, there are aspects of the history of the NT text that make a simple definition of “original text” more complex than what might otherwise have been conceived.  Prior to considering the obstacles in defining “original text” as discussed by Epp, it may be helpful to establish a definition that is useful in light of the Christian/Muslim debate over the integrity of the New Testament textual tradition.  It will be helpful to keep this definition in mind as we wade our way through the complex issues and possibilities raised by Epp. 

Typically it is argued by Muslims that the “Gospel“, however this is to be defined or elaborated, was a book (or books) inspired by God.  However, the Gospel book(s) eventually became corrupted to the point where they are no longer documents compatible with Islam’s view of Jesus and Christianity.  Most importantly, it would seem that the original inspired Gospel(s) (as viewed by Muslims) could not possibly state or imply that Jesus was truly crucified, suffered an atoning death, was resurrected, or was divine.  Perhaps some Muslims would try to argue that a non-canonical Gospel (whether still existing in extant MSS or not) was the originally inspired Gospel.  We argued that such an approach would require a solid theological argument demonstrating such a Gospel’s compatibility with Islamic beliefs in this regard as well as a solid historical argument that such a Gospel should be taken as historically more reliable than the canonical Gospels.  Otherwise, it could be argued that Mark, Matthew, Luke, and/or John was/were the Gospel(s) alluded to by the Qur’an as originally divinely-inspired and yet scribal corruptions have rendered it/them into a form theologically and historically very far removed from the autograph version(s).

We will assume the latter approach for the purpose of this article.  With that being the case, the definition that I would give of “original text” would simply be the finished form of the text prior to circulation.  It must be remembered that, unlike the Qur’an, there is no evidence that the NT was believed to have been written as the result of direct dictation from God or an angel of God (such as e.g. Gabriel).  The evangelists used sources.  They certainly utilized oral traditions that were Aramaic in original form and almost certainly used written traditions as well.  We mentioned in the last section that Matthew and Luke, for instance, probably used Mark and a sayings source given the title of Q in modern scholarship.  Furthermore, it is not impossible that the evangelists edited initial drafts of their Gospels prior to submission to their intended recipient(s), or perhaps even after submitting an initial draft.  As we will see below, it is not impossible that there were at one point multiple editions written by the same author, though this is quite speculative and impossible to prove.  If such were the case it may be more accurate, in the strictest sense of the word, to speak of inspired authors rather than inspired texts.  Yet we’re getting ahead of ourselves just a bit.  Let us say that our definition of “original text” is the text written by the evangelist in its finished form at the time of submission to its intended recipient(s).  Subsequent changes to this text (whether deliberate or accidental) are the corruptions that remove us from the “original text”.  With this in mind let’s consider some of the postulations by Epp.

‘Orthodox Corruption’ Revisited

Epp’s article contains a section discussing Bart Ehrman’s thesis of the “Orthodox Corruption of Scripture” where the latter demonstrates that early scribes altered certain NT passages in order to bring them more in line with orthodoxy.  Considering the phenomenon of such scribal alterations, Epp asks:

Therefore, which is the “original,” the texts altered by the scribes--now much obscured--or the scribes’ altered texts? (Epp 2005; 567)

I think the answer to this one is simple, at least in regards to the working definition we’ve established at the outset, that being what would seemingly be the most relevant aspect of the debate as far as Christians and Muslims are concerned.  Clearly the “original text” is that which was altered by the scribes.  The scribes’ altered texts changed what was written initially by the inspired writer and thus cannot be considered in any way meaningful to our debate as “original text”. 

Observations Regarding the Gospel of Mark

Epp next considers the thoughts of William L. Petersen.  First, we come across the following written by Petersen:

Is the “original” Mark the Mark found in our fourth-century and later manuscripts?  Or is it the Mark recovered from the so-called “minor agreements” between Matthew and Luke?  And which if any of the four extant endings of “Mark” is “original”?  And how does the “Secret Gospel of Mark”…relate to the “original” Mark?  It is clear that, without even having to consider individual variants, determining which “Mark” is “original” is a difficult--and perhaps even impossible--task (Petersen 1994; 136-137). 

I personally do not find these “penetrating questions”, as Epp refers to them (Epp 2005; 568), as difficult questions to answer at all.  The fourth-century and later manuscripts of Mark are “original” in the places where they contain readings that can be determined through the use of sound text-critical principles to be what was originally written by the original author in the first century.  The original text can only be determined by studying the text across a wide range of witnesses.  We’ve discussed the “Minor Agreements” in the previous section.  In short, the evidence as it exists at present indicates that they were not part of the canonical edition of Mark at all, but rather are original to Matthew and Luke.  We can entertain the possibility that they were present in Ur-Markus, or even a Deutero-Markus (i.e. an edition written after the canonical edition of Mark), but this is hypothetical.  We have no hard evidence of the existence of such editions (more on this possibility later though).  As for the question of the endings of Mark, it is clear that only one version could have been penned by the original author, and this appears per the evidence to be the version ending at 16:8.  Finally, recent studies on “Secret Mark” demonstrate that it was probably a 20th century creation of Morton Smith, the scholar who initially “discovered” the document (cf. Evans 2006; 94-99; there has also been a full-length book written by Stephen Carlson called “Gospel Hoax”).

Epp goes on from here to discuss some interesting findings by Petersen from early Patristic evidence.  We will skip this for now and revisit in the next section, as I think Petersen’s findings are intriguing enough to merit a section of their own.        

Multiple Original Editions of Canonical Books?

The possibility that certain NT books had multiple versions penned by the original author I find to be the most compelling and thought-provoking issue raised in Epp’s article.  When we think about what the “original text” may have contained in a given book it is easy not to even raise the possibility in our minds that there may have been more than one “original text”.  Is it possible that a given NT author released more than one version of a book into circulation?  Let’s consider the evidence for this possibility and the implications.  Epp writes the following:

When does a writing’s literary existence begin?  Can the beginning of a writing’s literary history be limited to the moment when copies were made and circulated (that is, the time of its “publication”)?  And if earlier composition levels can be detected, especially when signaled by textual variants, have textual critics not uncovered an earlier “beginning” of that writing’s literary history?  Or, to move forward in time, could not a literary process (such as revision or rearrangement of the text) have taken place after the first copies were made and released, thereby turning the earlier, copied version itself into a predecessor literary layer of the writing?  Hence, the term “beginning” begins to take on multiple dimensions, just as “original” does, and textual critics face the possibility that the text of a writing that has been transmitted, which they presume to have stood at the beginning of that particular writing’s history, now can be shown (triggered by textual variants) to have evolved from an earlier “beginning” -- an “original” has had earlier “originals.” (Epp 2005; 577)

In light of these possibilities, Epp argues that textual criticism has a role to play in compositional theories regarding the NT documents.  He gives his own definition of what falls “within the proper domain of textual criticism” (ibid. 577):

Any search for textual preformulations or reformulations of a literary nature, such as prior compositional levels, versions, or formulations, or later textual alteration, revision, division, combination, rearrangement, interpolation, or forming a collection of writings, legitimately falls within the sphere of text-critical activity if such an exploration is initiated on the basis of some appropriate textual variation or other manuscript evidence.  (ibid. 578; emphasis original)

From here Epp goes on to discuss two “categories”, namely “pre-original” compositional levels and “post-original” literary activity, that may be discernible through the study of textual variants. 

In category I (“pre-original”), Epp lists the following “subcategories”:

1.  Sources that may have been utilized by the evangelists (the quintessential example being that of the hypothetical sayings source Q).  He states that variant readings in not only the canonical Gospels, but also other sources like the Gospel of Thomas and certain agrapha (I.e. sayings attributed to Jesus not found in the Gospels) may provide help in advancing theories of such pre-compositional levels (ibid. 578). 

2. Theories regarding “varying versions, revisions, formulations, partitions, or combinations behind, or interpolations into, or collections of the Pauline letters” due to “variant readings concerning a letter’s addressees, the placement of doxologies, etc., and because of manuscript sigla indicating textual problems” (ibid. 578).

3. The two versions of Acts.

Epp also lists in this category hypotheses pertinent to the endings of Mark as well as the pericope adulterae (John 7:53-8:11), but since these latter issues do not obscure the term “original text” in any way meaningful for our purposes, we will forego discussion of these items.

Regarding subcategory #1, hopefully textual criticism will continue to prove fruitful in the endeavor to discern the various evangelists’ sources, yet I don’t think it obscures the issue of “original text” that is relevant to the Christian-Muslim debate regarding the integrity of the NT text.  While the veracity of certain compositional theories and sources underlying the Gospel tradition may prove important regarding the pertinent historical-theological questions relevant to Christianity, what concerns us here is the “original text” of the Gospels.  In other words, what did Mark, Matthew, Luke, and John originally write?  Is what they wrote substantially preserved or is it not?  Obviously, the historicity of the events they write about and the integrity of Jesus’ sayings and teachings contained therein is a crucial matter.  Nevertheless, it is important that we not confuse the two issues.  Here we are concerned about whether or not what was originally written by the evangelists has been preserved in the textual tradition, not with how well what was originally written conforms to what Jesus actually said.

Now, the question of multiple editions of a given book does have the potential to obscure what we mean when we say “original text”.  What if Mark released an early version of his Gospel (i.e. “Ur-Markus”) as well as a revised version that is our canonical version?  What if Luke is responsible for both the canonical and “Western” version of Acts, rather than the latter representing a textually corrupt version?  More on this below after we consider Epp’s 2nd category.

For category 2 (“post-original”), Epp lists the following:

1. Scribal alterations in the interest of orthodox or heretical theological viewpoints or other corruptions made in the interest of a given viewpoint, i.e. pro- or anti-Judaic sentiments or pro- or anti-female views.

2. “…rearrangements, additions, dislocations, and interpolations in already circulating writings, such as endings of Mark, portions of John or Pauline letters, etc. because of variant readings and varying positions or sigla in manuscripts.” 

3. Embellishments of such institutions as the Lord’s Prayer or Eucharist due to the “multiple forms in the textual tradition”.

4. Scribal alterations of the sayings of Jesus regarding marriage and divorce in the Synoptic Gospels (ibid. 579-580).

For our definition of what constitutes “original text” as it is relevant to the Christian-Muslim debate (once again, that which was original to the inspired author), it seems that only #2 in Epp’s list may be germane.  As we’ve argued above, those alterations made to the text in order to emphasize orthodoxy (or heterodoxy) are clearly corruptions of the original text.  The issues regarding the Lord’s Prayer and the sayings on marriage and divorce are not, as we saw in the previous section, in determining the probable original text, but rather as to why these original texts are different in the Synoptic Gospels.  And so the issue is in trying to reconcile what was originally written by the authors of the Synoptic Gospels rather than actually determining what they wrote.  The textual issue regarding the Eucharist we discussed in the main article at some length.

Thus I think we are left with dealing with the implications of the possibility that multiple versions were penned by the inspired authors.  Of the data presented, I think that there are four major issues worth discussing:

1. Multiple Gospel editions.

2. Interpolations into the Pauline epistles.

3. The questionable originality of John 21.

4. Dual versions of Acts.

Multiple Gospel Editions?

We need not spend much time on this possibility as the evidence for it is probably the most speculative.  Could there have been an Ur-Markus or even a Deutero-Markus?  Can this explain the so-called “Minor Agreements” in Matthew and Luke?  I think that such theories number among probably about a half a dozen possibilities, at least one of which I find to be more likely (i.e. that in the case of the “Minor Agreements” Matthew and Luke were following material in Q that happened to overlap with Mark).  Some argue that Luke was using Matthew at some points and Mark in others (thus negating the need to even postulate a Q document).  David Parker, as we saw in the previous section, even states that the “Minor Agreements” are simply the product of the dynamics of the textual tradition of the time, perhaps mixed with oral tradition as well.  In any event, I’m aware of no hard evidence of multiple Gospel editions.  Even if this were the case, however, would the implications be profound? 

I do not believe this would be the case.  It once again must be kept in mind that the original authors did not receive the Gospels (or other NT documents) through the medium of direct dictation by an angel or God Himself.  The Gospels, like countless other documents of the past and present, may have undergone a degree of editing by the original authors prior to their publication.  Even though we aren’t used to thinking about the issue in such a way, I would not find it a major problem to consider the possibility of multiple editions penned and even circulated by the original author.  If we accept that the author was inspired by God, can we say that one version of what was written is any less inspired than another?  As long as the versions are theologically consistent I don’t think there is a major problem, even if such an idea may make the concept of “original text” somewhat more obscure.

On the other hand, could these purported pre-Gospel editions such as Ur-Markus and/or Q be argued to represent the “original Gospel” that Muhammad referred to in the Qur’an?  I think we can close the proverbial door to the possibility that such pre-Gospel editions as the hypothetical Ur-Markus will provide much help to the Muslim case.  It is highly unlikely that an author in revising his original work (or draft?) would make dramatic changes to his own book’s theology, particularly if we view the original draft as divinely-inspired. Could the vessel through which God inspired the writing of a Gospel corrupt the very text that he/she was inspired to write?  This is not an impossibility, but is certainly highly improbable. Until some specific evidence could be offered there is simply no reason whatsoever to take such a postulation seriously.  Such purported adjustments in an individual’s theology would certainly be possible, though the burden of proof lies with the one that would make such an assertion.

Now the matter of Q is somewhat different.  Could Luke and Matthew have taken up a source (Gospel?) that once promoted a different theology and/or different emphases than the evangelists themselves?  There is some scholarly support for claims that Q did not emphasize a high Christology, posit soterological significance to Jesus’ death, or have a theology of the resurrection (cf. Kloppenborg 2008; Robinson 2005).  Regardless, it seems that making a case that extends beyond theoretical would prove a very formidable if not impossible task.  For one, Larry Hurtado has persuasively argued that in fact the Christology of Q is quite high (cf. Hurtado 2003; 217-257).  Additionally, although the portions of Q that we can reconstruct contain no explicit references to Jesus’ death, John Kloppenborg argues that a number of texts within Q presuppose it, although he argues that a different interpretation is applied to it rather than that of soteriological significance (albeit not one that is mutually exclusive with this view) [cf. Kloppenborg 2008; 73-79].  If Q truly does presuppose the historicity of Jesus’ death, then it seems to contradict the Qur’an on the matter at S. 4:157, and with that contradicts the typical Muslim-understanding that the Qur‘an indicates that Jesus did not suffer death.  

Second, despite the fact that some scholars have reconstructed Q even to the point that there now exists a critical edition (see here), without a hard copy it is impossible to establish the true extent of Q.  Regarding the retrieval of the Q material, the most that even as optimistic a scholar as John Kloppenborg (an editor of the aforementioned critical edition) can say is “there is reason to think that they [Matthew and Luke] did not omit much” (ibid. 45).  He argues this largely based on the fact that Matthew and Luke together preserve 635 of 666 verses of Mark’s Gospel and that, on average, they preserved the wording of Q better than they did that of Mark (ibid. 45).   Of course, Luke by himself uses only about 60% of Mark’s Gospel (Dunn 2003; 148, n. 36), and so much rests on the assumption that Matthew would have used Q to a similar extent that he used the material from Mark (92%).  Moreover, it is quite possible that the so-called “Minor Agreements” in Matthew and Luke (see the relevant subsection under the “David Parker” heading above) represent places where these evangelists utilized Q instead of Mark (on the theory that there is some overlap between Q and Mark).  On this theory there may well have been a passion prediction in Q (cf. subsection C of this article). 

Third, given that Q is a sayings source, it would be hazardous to conclude too much based upon the alleged lack of references to the crucifixion.  If, as Athanasius Polag argues, Q represents a sayings tradition in which the core of the collection originated with the pre-Easter circle of Jesus’ disciples, it shouldn’t surprise us to find no reference to Jesus’ death in the sayings [cf. Polag 1991; 97-105].  In this light it may have been a document or collection of oral tradition used for teaching.  Could it have been this very material that the disciples preached when they were sent to the masses by Jesus to announce the Kingdom of God? (Cf. e.g. Luke 9:1-6)

Fourth, although I think something like Q probably existed, some scholars demur (cf. Goodacre & Perrin 2004). If, for instance, Luke utilized Matthew as a source in addition to Mark, this would explain the common texts in the former two Gospels just as well as the Q hypothesis would (cf. Mark Matson’s essay in Goodacre & Perrin 2004 for an answer to one of the most problematic aspects of the theory that Luke used Matthew).  Until a hard copy of Q is found or at least until solid external evidence of its existence presents itself, it is prudent to remember that Q remains a hypothesis, even if one that has much to commend it. 

Finally, assuming the existence of a Q document as well as a Q community whose theology did not attribute atoning significance to Jesus‘ death, did not have a theology centered upon Jesus’ resurrection, and did not maintain a high Christology (despite the findings of Hurtado to the contrary), one wonders about the place of such a community within the early church.  As we explained in the main article, the belief in the atoning death of Jesus, the belief in and emphasis on the resurrection, and a high Christology can be traced back to the earliest Jerusalem community and Jesus’ disciples with a high degree of confidence.  And so the question begged by these considerations would then be:  1) Who does the Q community represent and why should we take its theology seriously?  Obviously, if such a theology did exist it didn’t prove to be very popular or enduring [4].

Interpolations into the Pauline Epistles

With these last 3 categories we deal with possibilities in which there does exist evidence within the textual tradition.  Up first is the question of the Pauline epistles.  One particularly difficult issue is the many variants regarding the doxology at the close of Paul’s epistle to the Romans.  The doxology (“Now to him who is able to strengthen you… be glory for evermore through Jesus Christ!”) [trans. Metzger in Metzger 1994; 470] is found in multiple places per the textual tradition of Romans.  It typically concludes the 16th and final chapter, but is also found in P46 at the end of chapter 15, and in some MSS following chapter 14.  In some MSS it is found after both chapters 14 and 16 and some MSS do not contain the doxology at all (ibid. 470).  This is one of the most complex (perhaps even the most complex) text-critical issues of NT textual criticism, yet we will not deal with this issue in any depth here since the authentic text is beyond our scope.  Suffice it simply to say that the UBS textual committee decided, with some uncertainty, to place the doxology in its traditional place at the end of Romans 16, and that the excision of the last 2 chapters by Marcion may have been the impetus fueling the formation of the other diverse variants (cf. ibid. 470-473 for full discussion).

The question we wish to concern ourselves with is the possibility raised by some commentators that perhaps Paul wrote two (or even three) versions of the epistle since the doxology occurs after chapters 14, 15, and 16 depending on the witness.  Could Romans initially have been only a 14 or 15 chapter book?  This would seem possible even though perhaps it is more probable that Marcion’s excision of the last 2 chapters is what resulted in the variants.  Nevertheless, even if Paul did write 2 versions of Romans, what are the implications?  Clearly no major doctrine of Christianity is lost regardless of which version we accept to be canonical.  But, once again, if we do accept that Paul was responsible for multiple versions (one ending after ch. 14 and/or ch. 15, and one after a 16th chapter was added), it seems arbitrary at best to argue that the last chapter or two are therefore not inspired.

Epp also brings to our attention some other interesting features of the epistles.  For instance, the address phrase “in Ephesus” in Ephesians is missing in some important MSS.  Likewise “in Rome” is absent from certain MSS of Romans 1:7 and 1:15 (Epp 2005; 570-571).  Epp mentions several theories for the former phenomenon.  For instance, Archbishop Ussher in the 17th century surmised that perhaps Ephesians was a “circular letter intended for several churches and that a blank was left in 1:1 for names of churches using it,…” (ibid. 570)  He also quotes a theory of Nils A. Dahl:

[T]he letter was originally issued in several copies with a special address in each of them.  In any case, the letter must have had a pre-history before it was published as part of the Pauline corpus.  The text without any concrete address is to be understood as a result of a secondary “catholicyzing,” to which we have an analogy in the textual tradition of Romans. (Epp 2005; 570; quotation of Nils A. Dahl from “The Particularity of the Pauline Epistles as a Problem in the Ancient Church”)

Again, this is certainly an intriguing possibility, yet perhaps a stronger possibility is that the address was simply removed by certain churches in locations outside of Ephesus (obviously after the epistle became universal the specific address “to Ephesus” was defunct).  But, entertaining Dahl’s theory, could Paul’s 14 chapter letter to Rome (its initial designation) later have been expanded by the apostle into a 16 chapter letter to other churches?  Epp once again asks, “What text, then, of Ephesians or Romans is designated by the term ’original’”? (ibid. 571)

I would answer that clearly both would be “original” since both versions would have been penned by the original author.  Otherwise, I think it is more likely that the “to Ephesus” was simply dropped by some later witnesses given its having been rendered defunct after achieving circulation in congregations outside of Ephesus.  Now, if Dahl is correct in that numerous copies with the different addressees (of which Ephesus was one of many) were made, then we would have to reckon with the fact that there would have been variants among all of the “original copies”.  If this was the case, then the concept of “original text” admittedly becomes obscure.  Technically the “original” would have been the first copy Paul made (or dictated to an amanuensis).  The copies, made from this “original” prior to circulation, would have contained variations (albeit probably almost exclusively trivial ones) from the beginning.  This would make the task of achieving a perfect facsimile of Ephesians all the more daunting if this is the case since there would have been a number of minor variations present among the original copies from the beginning, all of which would have been dispersed to various geographical locations across the Roman Empire, subsequently to serve as exemplars for future generations of copies. 

Now, while I think the concept of “original text” may become somewhat more obscure should Dahl’s theory prove to be correct, the overall preservation of the epistle would be that much more secure.  With multiple copies having been made from the beginning (with only minor copyist errors preventing these copies from being identical to one another) and dispersed widely, it is virtually impossible that even the minor contours of Paul’s message would have disappeared through subsequent variations in the textual tradition.

Dual Versions of Acts

There were two versions of the book of Acts that circulated in the early centuries of the church, the Alexandrian and “Western” versions, the latter being approximately 8.5% lengthier than the former (Metzger 1994; 223, cf. note 3).  Needless to say, it is in this book where the most profound differences between these two text-types is manifest.  A number of theories have been proposed for the origin of the two versions.  Metzger discusses several intriguing theories that have been proposed over time, some of which posit that both versions were from Luke’s pen.  Perhaps Luke pruned a once-longer rough draft or expanded upon a shorter one.  In any case, two versions were made and proliferated (cf. Metzger 1994; 222-236 for full discussion regarding the various theories of origin).  The Alexandrian version is, not surprisingly, generally considered to be more authentic (ibid. 223).  The United Bible Societies’ textual committee more often than not favored the Alexandrian version in their analyses of the various texts with variants, yet “proceeded in an eclectic fashion, judging that neither the Alexandrian nor the Western group of witnesses always preserves the original text, but that in order to attain the earliest text one must compare the two divergent traditions point by point and in each case select the reading that commends itself in the light of transcriptional and intrinsic probabilities.” (ibid. 235)

However, let us back up and consider the possibility that Luke did indeed author both versions.  Which one in this case would be original?  As with our other examples, I would argue that this is perhaps the wrong way of looking at the matter.  Obviously the information contained in both versions would represent the theology of the same inspired author.  Thus, as long as there are no major theological discrepancies present among the two versions I wouldn’t see a problem.  In any event, the evidence that seems to be the strongest is that the “Western” version was not authored by Luke, but rather by later revisers:

In a more recent discussion of the origin of the Western text of Acts, Barbara Aland traces the several stages in the development of this form (or of such forms) of text.  In the second century copyists introduced interpolations, omissions, and alterations in the text of Acts that tended in the direction of the Western type of text.  In the first half of the third (?) century a redactor revised a manuscript that contained a form of text that belonged to the first stage, and this resulted in a text embodying the well-known “Western” characteristics.  At the third stage the redactor’s exemplar was copied by various persons who dealt with the text in a rather free manner. 

By way of summing up at least some of the analyses of the Western text, one may conclude that it would be more appropriate to speak of Western texts, rather than of a Western text.  At the same time, one can recognize a, so-to-speak, Western tendency that is shared by many such witnesses.  In this sense, as Strange declares, “it is legitimate to refer to the Western text, as long as it is understood that what is meant is a broad stream of textual tradition, and a way of handling the text, rather than a coherent recension of the text, created at a specific time.”  Understood in this way, Codex Bezae frequently offers the most original form of the Western text.  At the same time, of course, D has a manuscript history of its own, and does not invariably preserve the earliest form of the Western text.  To ascertain that stage one must also take into account the evidence of other witnesses, both versional and patristic (ibid. 234-235, emphasis original; cf. the whole discussion in 232-235).

Metzger also draws attention to the contradictions between the two forms, making them more likely to have been written by separate pens:

Furthermore, sometimes the shorter form contradicts the longer form.  For example, having described (in the first person plural) a break in the journey from Caesarea to Jerusalem at the house of Mnason (so the Western text of 21.16), the author would not be likely to alter it so as to suggest that Mnason lived in Jerusalem (as is implied in the shorter text). [ibid. 225]

John 21

Although not discussed by Epp, another issue bearing on this discussion is that of the 21st chapter of John.  It has long been recognized that this chapter seems, on the surface at least, to be almost superfluous to the Gospel as a whole.  After all, John 20:30-31 certainly would seem a fitting way to conclude the book.  Nevertheless, with the exception of a few, textual criticism does not tend to posit the absence of the 21st chapter of John from the original text.  The reason is that no manuscript of John exists to suggest its absence.  David Parker is thus exceptional when he suggests that this may have been the case (cf. Parker 1997; 177-179).

I think the evidence allows a couple of suggestions.  One is that the author of the Gospel of John added the 21st chapter as sort of an addendum, perhaps after initially intending to end the Gospel after chapter 20.  A second possibility is that the author’s disciples (or “community” as it is commonly referred to) added the 21st chapter, perhaps after the author’s death (cf. John 21:24).  Either way the text indicates that this chapter (along with the rest of the Gospel) represents the eyewitness testimony of the “Beloved Disciple” himself.  Whichever of these possibilities may have been the case, I am inclined toward the view that the Gospel of John did not circulate without the 21st chapter.  In this writer’s opinion, this is the only reasonable option until we at least find a manuscript (or at least some testimony from an ancient church father) indicating the existence of copies that did not contain the 21st chapter.  As far as I’m aware we have as yet no such evidence.

Nevertheless, let us once again entertain that this may have been the case.  Regardless of when the Gospel initially ended, Jesus still died by crucifixion, was resurrected, and made appearances to the disciples.  From the perspective of the Christian-Muslim debates, the possibility of John ending a chapter early is irrelevant.  From a historical point of view, I think the evidence presented from 21:24 still puts us in touch with John’s eyewitness testimony as reported by his disciples (regardless of whether or not this chapter was original to the Gospel).

Concluding Thoughts on Epp’s Thesis

The material put forth by Epp is certainly thought-provoking to say the least.  One has to admit to the possibility that this may require us to qualify or redefine what we mean when we say “original text”.  We would like to close by a brief reiteration of the specific categories discussed by Epp and the implications for our current purposes.

In Epp’s section of “Proposed Dimensions of Meaning in the Term, ‘Original Text‘”, he refers to four such forms.  One is the predecessor text-form, referring to forms of text “discoverable behind a New Testament writing that played a role in the composition of that writing.” (Epp 2005; 586)  This would refer to such hypothetical documents as Q, Ur-Markus, M, L, etc [5].  It is an exciting possibility that text-critical principles may shed some light on such matters, but it is I think essentially irrelevant to the current debate regarding Biblical corruption.  The sources used by the evangelists and/or their hypothetical rough drafts do not constitute the “original text” as it first emanated from the desk of a Mark, Luke, or Paul.  It is the finished form of the writings of the inspired writers that is the “original text”, at least as far as the matter has been classically understood.  It is Epp’s second category, what he terms the autographic text-form, that we are arguing is the “original text”.

Third, the canonical text-form, is somewhat complex.  This is “the textual form of a book (or a collection of books) at the time it acquired consensual authority or when its canonicity was (perhaps more formally) sought or established, such as when a collection was made of the Pauline letters or of the four-fold gospel, or--at the level of detail--when phrases like “in Rome” or “in Ephesus” might have entered or been removed from the text.”  (ibid. 587).  To these examples could even be added any text (corrupted or original) that was treated as authoritative to any part of the church in any era of history.  Thus to most Christians who have not heard otherwise (or are even aware of the debate!), the Markan Appendix (16:9-20) and the story of the woman taken in adultery (John 7:53-811) are canonical, i.e. part of the inspired text.  Thus I would argue that although some text-forms would have been treated (and in many cases still are treated) as canonical does not mean that they are in any sense “original”.  Regardless of what this or that person may think about the original text, whether it is through ignorance or actual evaluation, it is what the evangelist actually wrote at the beginning that constitutes the “original text”.   

Finally, there is the interpretive text-form, “representing any and each interpretive iteration or reformulation of a writing--as it was used in the life, worship, and teaching of the church or of individual variants so created and used.” (ibid. 587).  It is these kinds of changes in which the variants discussed by Bart Ehrman in “Orthodox Corruption of Scripture” would most easily fall.  However, any such changes made would clearly not be “original” in the classic sense that we‘ve defined. 

Epp goes on from here to discuss how these different text-forms do not necessarily show a linear relationship, and that an autographic text-form, for example, may be an interpretive text-form.  This would be true if, say, Mark reinterpreted a source (be it oral or written) and incorporated this reformulation into his text.  Of course, this reformulation would still represent Mark’s “original text”. 

We will close this section with a note of clarification.  It is one thing to establish the original text of the Gospels (and the rest of the NT).  It is quite another to determine from this “original text” what it is that Jesus may have said and whether or not his sayings were accurately preserved by the evangelists (and the stream of oral tradition that preceded the penning of the Gospels), or that what was originally written is historically accurate.  Determining these things is obviously a very important matter.  However, what concerns us here is the preservation of the original text, not whether or not the original text is accurate.  It is important to keep this distinction in mind in order to avoid the release of red herrings.

William Petersen – “What Text Can New Testament Textual Criticism Ultimately Reach?”

In this article Petersen sets out to demonstrate the importance of the Early Church Fathers [ECFs] (i.e. patristic citations) in establishing early forms of the NT text.   In the Prologue to the article Petersen complains that the NT papyri and even more so the patristic evidence are not accorded nearly the value that they should when modern textual critics formulate critical texts.  He writes:

There is abundant Patristic evidence for the text of the New Testament, much of it very ancient.  Examples include Ignatius, Justin Martyr, the Diatessaron, and the Didache, all of which date from before 175 CE.  Nevertheless, Patristic evidence has been largely ignored, while the papyri -- all but one of which are later, and whose empirical contribution to the critical text has been nil -- have received so much attention.  Despite their superior antiquity, Patristic evidence has fared no better than the papyri in the text of N-A27/UBS4.  An examination of the apparatus of the gospels shows not a single instance where the text is based solely -- or even principally -- upon Patristic evidence.  Only when a Patristic reading is supported by the uncials does it enter the critical text. (Petersen 1994; 139-140; emphasis original)

Petersen’s contention throughout the article is that modern textual scholars are so biased toward the Alexandrian textual tradition that evidence from the papyri, but even more so that of the Patristic evidence, is only given the credence it deserves when it conforms to the Alexandrian tradition.  He argues, however, that the ECFs can take us to an earlier layer of text than the great uncials of the 4th century.   

Before making his case, Petersen acknowledges the basic problems peculiar to Patristic evidence.  We will copy what he writes in this regard:

(1) What is the precise extent of the citation -- that is, where does it begin, and where does it end?

(2) Is the citation literal, or from a perhaps-faulty memory?

(3) Is the citation only an allusion or a paraphrase?

(4) Have changes been made to suit the purposes -- homiletical, paraenetic, or illustrative -- of the person writing?

(5) Did the person citing the text “smooth” its perhaps awkward diction, or omit an irrelevant portion? (ibid. 140)

From here Petersen gives us three “exhibits”, each of which he argues represents an earlier form of the text than what is typically accepted in critical editions.  We will consider each of these. 

A Variant of the Shema

"Teacher, which is the greatest commandment in the Law?" Jesus replied: " 'Love the Lord your God with all your heart and with all your soul and with all your mind.' " (Matthew 22:36-37)

One of the teachers of the law came and heard them debating. Noticing that Jesus had given them a good answer, he asked him, "Of all the commandments, which is the most important?" "The most important one," answered Jesus, "is this: 'Hear, O Israel, the Lord our God, the Lord is one. Love the Lord your God with all your heart and with all your soul and with all your mind and with all your strength.'" (Mark 12:28-30)

"What is written in the Law?" he replied. "How do you read it?" He answered: " 'Love the Lord your God with all your heart and with all your soul and with all your strength and with all your mind'; and, 'Love your neighbor as yourself.'" (Luke 10:26-27)

Petersen cites Justin Martyr’s version in Dialogue 93:2 which reads “You shall love the Lord your God from all your heart, and from all your strength.” (Petersen 1994; 145)  This same “binary form” of “heart…strength”, according to Petersen, is found also in Dialogue 93:3 and First Apology 16:6 (ibid. 145).  While the Masoretic Text and the Septuagint (i.e. versions of the OT), along with Matthew, contain a “trinary form” (heart, soul, mind) and the Lukan and Markan versions contain a “quaternary form” (heart, soul, mind, strength), Justin’s text represents the oldest Shema per Petersen.  Petersen asks if Justin could have on 3 occasions “forgotten what the ‘real’ text was”, and cites support for a binary reading in Latin version Codex Bobbiensis (c. 400 CE) at Mark 12:30 and also the Curetonian Syriac of the fifth century at Luke 10:27, the latter of which initially reads “heart…power”.  Petersen believes that this text preserves “Justin’s deviating order” though it follows with “our present canonical version: ‘and from all thy soul and from all thy mind.’” (ibid. 146)  From this Petersen argues that Justin’s text is likely original and that the versions contained in virtually all of the manuscripts were probably brought into conformity with the Septuagint (LXX) and Masoretic Text (MT) (ibid. 146-147).   

Of the three examples discussed by Petersen it would seem that this is his best argument, though the implications are it seems the most trivial.  On the other hand, given the differences in forms (i.e. “binary”, “trinary”, “quaternary”), there were likely more than a few variations in the order and the number of elements presented as this saying (or probably more accurately these sayings, i.e. given on multiple occasions) of Jesus were passed on through oral tradition.  Thus to ask whether the “binary”, “trinary”, or “quaternary” form is original may be to miss the point.  It is hardly surprising that such variations would have been introduced into the manuscript tradition, certainly such a formula would have been subject to multiple variations at least through the medium of oral tradition. 

The Story of the Rich Young Man

Now a man came up to Jesus and asked, "Teacher, what good thing must I do to get eternal life?" "Why do you ask me about what is good?" Jesus replied. "There is only One who is good. If you want to enter life, obey the commandments." (Matthew 19:16-17)

As Jesus started on his way, a man ran up to him and fell on his knees before him. "Good teacher," he asked, "what must I do to inherit eternal life?" "Why do you call me good?" Jesus answered. "No one is good—except God alone. " (Mark 10:17-18)

A certain ruler asked him, "Good teacher, what must I do to inherit eternal life?" "Why do you call me good?" Jesus answered. "No one is good—except God alone." (Luke 18:18-19)

Petersen cites a similar quotation by Justin Martyr, writing at approximately 150 A.D.:

“One is good, my Father in the heavens.” (Justin, Dial. 101.2; emphasis original)

Based on the similarity of the original Greek between Justin’s citation of “One is good” and the Matthean version, Petersen argues that the Matthean version is being cited by Justin.  Yet Justin’s text also reveals evidence of wide dissemination:

Justin’s text is supported by Tatian’s Diatessaron, as well as Irenaeus, Hippolytus, and the Pseudo-Clementine Homilies.  In truncated form, the same variant is found in Clement of Alexandria and two of the oldest Vetus Latina manuscripts of the canonical gospels.  This is an extraordinary situation, for in strictly chronological terms, then, the oldest-known version of this Matthean pericope contains the phrase ‘my Father in the heavens.” (ibid. 142-143; emphasis original)

Petersen further writes:

While one might be tempted to regard the variant as an interpolation, and therefore secondary, it should be noted that the suppression of this indisputably primitive reading, the most ancient version of the passage we possess, solves the theological, the Christological problem caused by its inclusion. (ibid. 143; emphasis original)

Thus because Jesus by implication was denying being “good” with the inclusion of the “my Father in the heavens” clause, this clause was excised and evidence of its existence is only found in the Patristic evidence as well as a couple of early Latin versions. Petersen further argues that the inclusion of such a clause into this saying matches the theology of the early church, citing Acts 2:22, and the ECFs (such as Justin and Origen, who were adoptionist or subordinationist according to Petersen) [ibid. 144]. 

We may take issue with Petersen’s understanding of the Christology of the early church as well as the ECFs (we would agree that the NT and 2nd century Church Fathers taught a functional subordinationism, though not adoptionism or ontological subordinationism), yet this is beyond the scope of this article.  The fact is that if Petersen’s interpretation of this text as it is cited by Justin Martyr is correct, we would have at worst passages whose theology is in tension with others of a higher Christological bent.  Daniel Wallace demonstrates four passages whose authenticity is not questioned that purports the belief in Jesus’ divinity (see the relevant section of our main article on this issue).  I am nevertheless not convinced by Petersen’s interpretation.  In fact, if the passage as it stands in the citation by Justin Martyr is authentic, I would argue that it is less theologically-difficult than the parallel passage in Mark.  Mark 10:18 could be interpreted to mean that Jesus is implying that he is in fact not divine.  In fact, while I am of the persuasion that the NT constantly and consistently presents Jesus in a very exalted, high Christological light, I think that the Markan passage (on the surface at least) provides the most challenging obstacle to high Christological interpretation (though see Sam Shamoun’s observations in this article where in delving beneath the surface of this passage we find the same high Christology as elsewhere in the NT corpus).  In any event, by saying that only the Father is “good” (agathos), Jesus does not appear to imply that he is not divine as in the Markan version.  The resultant reading merely provides the disadvantage of having to explain how Jesus could deny being “good” rather than a reading that could be interpreted as an implicit denial of divinity.

The above considerations are relevant if in fact Justin is quoting from a manuscript of Matthew’s Gospel, yet this is far from clear.  Michael W. Holmes has the following to offer in way of criticism of Petersen’s argument for the originality of “my Father in the heavens” to Matthew 19:17:

Second, we may notice Petersen’s persistent tendency to assert rather than argue the source of a gospel citation.  Identifying a source of Patristic citation of Gospel material can be a difficult and sometimes impossible challenge as Petersen well knows, and he includes in one essay a fine discussion of the  methodological problems involved, but in practice he regularly ignores his own guidelines, and let me illustrate the problem.  The first involves a citation from Justin Martyr that has parallels in Matthew, Mark, and Luke…

And the text of Justin is distinctive:  “Good Teacher, why do you call me good?  There is one who is good, my Father in heaven.” Now even though Justin never identifies his sources, Petersen declares that the similarity of Justin’s phrase “There is one who is good” to Matthew shows that it is the Matthean version which is being cited, and therefore Justin preserves the earliest version of Matthew 19:17, one that includes the phrase “my Father in heaven”, proof he says that our critical text of the gospels does not correspond to the early 2nd century text of Matthew. 

But Petersen’s identification of this as a citation of Matthew is surely debatable.  The two preceding phrases “Good Teacher” and “Why do you call me good?” reflect Mark and/or Luke, not Matthew.  And furthermore Justin is known to have used a gospel harmony,… one that is based on multiple sources in addition to Mark, Matthew, and Luke.  How does Petersen know that the phrase “my Father in heaven” comes from Matthew and not one of Justin’s other sources?  He doesn’t, but yet he makes an identification anyway. (Holmes, Textual Transmission, 2008; time slice 25:08 - 27:02) 

Interestingly, in another work penned at about the same time Petersen seems to indicate that this variant in the Diatessaron is found at Mark 10:18, and that this is also the case not only for Justin Martyr’s text but also “Ephrem’s Commentary” (Petersen 1995; 91).  This essay was published the year following the article that we’re discussing.  Which one was actually written first I can’t say for certain.  In any case, did Petersen change his mind within a matter of months on the issue? 

Needless to say, it is highly questionable at absolute best that the “my Father in heaven” phrase initially belonged to the text of Matthew.  If in fact the “my Father in heaven” phrase entered the oral and/or textual tradition as a result of reflection upon the Markan/Lukan passage (even if misinterpreted), one may argue that this was a later scribe’s way of avoiding what seemed to be an implicit denial of divinity on the part of Jesus. 

Bodiless Demon

“Look at my hands and my feet. It is I myself! Touch me and see; a ghost does not have flesh and bones, as you see I have. (Luke 24:39)

This passage occurs within the context of Jesus’ post-resurrection appearance to the disciples where he is trying to convince them that he is not a spirit.  Ignatius of Antioch, writing no later than the close of the 1st decade of the 2nd century, refers to this episode where similar language is utilized:

“And when he came to those about Peter, he said to them: ‘Take, handle me and see that I am not a bodiless demon.’ And immediately they touched him and believed, being intermingled with his flesh and spirit.” (Letter to the Smyrnaeans 3:2; translation in Schoedel 1985; 225; emphasis added)

William Petersen states that Ignatius is here citing Luke 24:39, and that this variant reading is attested in other authors such as Eusebius (when quoting this passage from Ignatius’ work) of the 4th century.  Origen (writing c. 220-230), in a work entitled De Principiis, refers to this reading and attributes it to a non-canonical document entitled Doctrine of Peter.  Jerome, in Commentary on Isaiah, attests to the variant as well, attributing it to the Gospel of the Nazoraeans (Petersen 1994; 144-145).  Petersen further writes:

The variant “bodiless daemon” is clearly the most ancient extant version of Luke 24:39, for it is known not just to Ignatius -- which means it was known in the first decade of the second century -- but also to a clutch of other second and third century writers.  It is interesting how the reading -- which is apparently the standard reading for Ignatius -- is later attributed to heretical gospels by Origen and Jerome.  But it is still the standard text for Titus of Bostra, who died c. 370: “touch and see that a daemon…does not have flesh and bones” (Contra Manicheaos, IV.37, Syriac version) [Petersen 1994; 145; emphasis original].

Yet it is highly questionable that this is a variant reading at all as opposed to a separate tradition.  Michael Holmes offers the following criticism:

The second example involves a citation from Ignatius of Antioch which Petersen claims is clearly the most extant, the most ancient surviving version of Luke 24:39... Now Bill bases his claim on the fact that the two passages share five identical words, “touch me and see that”.  Five in Greek and this case nicely five in English.  “Touch me and see that”, there’s the identical phrase, but in this case the differences I think are more significant.  First, note the sharp difference between the immediately following phrases:  “I am not a disembodied demon” vs. Luke 24:39 “a spirit does not have flesh and bones like you see I have”.  Second, observe how what Ignatius says after that, “Immediately they touched him and believed” contradicts what Luke says in verse 41 where “they were still not believing”.  It seems far more likely that Ignatius is working with a parallel or similar tradition to Luke 24:39 than that he is actually citing Luke 24:39. 

Now these two examples are typical of Petersen‘s procedures throughout.  He asserts what needs to be demonstrated.  His failure to demonstrate that these early anonymous citations of Gospel tradition that he discusses are in fact citations of a specific Gospel undercuts his case.  (Holmes, Textual Transmission, 2008; time slice 27:02 - 28:29)

Petersen goes on to conclude that judicious use of the Patristic evidence can uncover a layer of text that is earlier than the text as it can be yielded from the current MSS, using the above-discussed “exhibits” as evidence for this view.  He even goes as far as to say that this ancient evidence will yield more evidence of an adoptionistic theology and that this brings us closer to the “original text” (cf. ibid. 148-151).

Needless to say, it is this kind of evidence, if valid, that has the potential to provide much-needed fuel to Muslim assertions that the Gospels do not portray Jesus as divine.  I think it is precisely these kinds of arguments that need to be presented by Muslims if textual criticism is going to be of any true value whatsoever in substantiating their assertions regarding “the original Gospel”.  In the present case, not only do we reject Petersen’s claims that the NT betrays an adoptionist theology in general (cf. the relevant sections of the main article), but we have also found, thanks largely to the criticisms by Michael Holmes, that Petersen’s argument regarding Justin Martyr’s alleged citation of Matthew 19:17 is simply insufficient in trying to demonstrate this.

Ancient NT Textual Criticism

While speaking of “miscellaneous issues” in textual criticism, it is perhaps fitting to close with a brief section that indicates the sensitivity of those transmitting the NT textual tradition in the early centuries of the church to text-critical issues.

We already saw in the main article (this section) that the Early Church Fathers Origen and Jerome lamented and/or discussed many textual variations within the manuscripts of their own day.  Yet this sensitivity to text-critical issues is also evident from certain features of the ancient manuscripts as well.  Kurt and Barbara Aland mention the fact that certain manuscripts list more than one ending for Mark’s Gospel and/or list the longer ending (16:9-20) with either critical notations or comments that question its authenticity (cf. the lengthy quote from the link above) [cf. also Parker 1997; 126-127].  The same phenomenon also occurs regarding the questionable placement of the Doxology either after chapter 14 or 16 of Romans (cf. above the subsection “Interpolations into the Pauline Epistles” in the Eldon J. Epp section).

Then there is the presence of manuscript sigla which (in some cases) may represent notations indicating textual variation.  The quintessential example of this phenomenon would be the so-called “umlauts” of Codex Vaticanus.  One key study of these “umlauts” is that of J. Edward Miller in (Miller 2003).  In this article Miller notes that there are 778 cases where Vaticanus posts an “umlaut” (i.e. two horizontal dots appearing in the margin beside a line, with a line comprising typically between 15-19 characters).  On average, there is one line marked with umlauts per 45 lines of text, which amounts to an average of about 3.5 umlauts per chapter of extant text of Codex Vaticanus (ibid. 218-219, n. 6).  This means that the scribe of Codex Vaticanus marked about 2.2% of the lines of text with umlauts.  The evidence strongly suggests that the umlauts were utilized in at least large measure to indicate textual variation at certain points in the text (cf. ibid. 224-232; Daniel Wallace further discusses Miller’s article and the implications here).  The fact that the umlauts occur in many places with well-known textual variation supports this assertion.  Now, if the “umlaut” was used exclusively to indicate textual variation, then this would indicate that a substantial number of textual variations known to the scribe that penned Vaticanus are not extant in surviving MSS [6].  It is doubtful, however, that this is the case.  Miller notes the following:

For instance, for three of the above passages, where the scribe signifies knowledge of attempts to harmonize with a parallel text, he supplies another umlaut next to the line with which it is sometimes harmonized.  An umlaut appears at Mt. 26.26 (with which Mk 14.22 is sometimes harmonized), Mt. 6.9 (with which Lk. 11.2 is sometimes harmonized) and Rom. 1.25 (with which Rom. 1.18 is sometimes harmonized).  Since no early manuscripts (apart from a few versional witnesses in the case of Mt. 26.26) show variants in the particular lines indicated by the umlaut, the marginal notations are probably marking commonly harmonized texts.  In each of these instances, the scribe was likely highlighting both facets of the harmonization (the true text and the harmonized) while offering to subsequent copyists a caveat not to defile the text further.  Nevertheless, these instances of parallel texts offer impressive evidence that the Vaticanus scribe employed the umlaut siglum to denote textually uncertain lines (ibid. 229-230).

In any event, the presence of “umlauts” in Codex Vaticanus provides us with another example of the sensitivity of ancient scribes/Christians to the matter of textual criticism.     

One final consideration we’ll bring to light here is regarding the use of ancient Greek MSS in the formation of the versions.  The translation of the Greek text into Latin can be traced to the 3rd century with the oldest surviving witnesses dated to the late 4th century.  It was about this time when Pope Damascus commissioned Jerome to create a new Latin version, hence the formation of the Vulgate based on these Old Latin manuscripts with other ancient “good Greek manuscripts” (cf. Parker 1997; 13). 

Similarly we have in the Syriac textual tradition evidence of the text being corrected against Greek witnesses.  The Peshitta version was corrected, for example in the Philoxenian translation due to the production of mistranslations in the former (Baarda 1995; 105).  Tjitze Baarda points out that, for instance, in Hebrews 5:7, compared to the “less fortunate translation of the Greek in the Peshitta, Philoxenus himself presents a Syriac text of this verse that does justice to the Greek text of the Apostolos.” (ibid. 106) [7] Thomas of Harkel later made a new collation of the Philoxenian text yet again with the use of other Greek MSS, intended as “a revision of the Philoxenian with the intent to create a Syriac text that was a most literal representative of the underlying Greek text.” (ibid. 107)  One important difference between the Philoxenian and Harklean texts from that of the Syriac Peshitta is the rendering of Hebrews 5:7 to a form more in conformation to the original Greek: “in the days of his flesh”, as opposed to the Peshitta version’s “he was clad with flesh”.  The differences are theologically significant, and as mentioned, Philoxenus thought it possible that such a change was a result of either deliberate or unintentional “Nestorianisms” (Is this an example of a heterodox corruption of Scripture?) [ibid. 109] (cf. the whole article in Baarda 1995 for full discussion).

The final version we will consider is that of the Coptic text, which is subdivided into Sahidic, Achmimic, Subachmimic, Middle Egyptian, Fayumic, and Bohairic versions (cf. Wisse 1995; 133-137 re: dating and origin of these versions).  The Coptic versions, specifically the Sahidic and Bohairic versions, are important due to their textual affinities to the Alexandrian text, though the relevant MSS do contain some Western readings (ibid. 137).  Although the formation of all of the various Coptic versions may not have been created with Greek MSS at hand for correction, it is likely that this would have been the case at least some of the time.  For instance, Wisse notes the following regarding the Bohairic version:

As far as we know, Greek had remained the official liturgical language of the Coptic church until at least the Arab conquest.  The shift from Greek to Bohairic in the church services, which appears to have begun sometime after the conquest in Alexandria (642 C.E.), would be the likely reason for the origin and spread of the Bohairic version.  Earlier versions may have been consulted for the creation of the new ecclesiastical version, but in all probability it was basically a fresh translation of the Greek text that had been in use up to that time in Alexandria. (ibid. 136-137)  

And so textual criticism of the NT is not a phenomenon restricted to the past couple of centuries.  Our current manuscript witnesses represent in many cases transcriptions or translations of exemplars with the use of multiple Greek MSS for the purposes of correction.  Codex Vaticanus, among others, utilized manuscript sigla (i.e. “umlauts”) in order to indicate points in the text where there was textual variation according to other ancient manuscripts.

We have spoken at some length in the past regarding the wealth of NT manuscripts, often met with the response that, while this is true, there are comparatively few that can be dated to the 3rd and 4th centuries, even less to the 2nd century.  In light of this, a couple of important take-home points should be made from this data regarding “Ancient Textual Criticism”.  First, we have observed that the presence of textual variation may be detected in a number of ways not restricted to the comparison of the actual readings from various MSS.  These would include: 1) Discussions by the ECFs of such variants and specific examples they provide (i.e. Origen, Eusebius, Jerome); 2) The presence of manuscript sigla, such as the “umlauts” found in Codex Vaticanus (though this category of data seems to be the least helpful in yielding any certain or specific data); 3) Marginal comments in certain MSS regarding particular variants; 4) Judicious use of quotations of Scripture by the ECFs.  Second, while the relevant data considered in this section may reveal more in the way of textual variation than exists merely from consideration of the actual readings in MSS, the places where no such indication of significant variation exists gives us that much more confidence in the integrity of the text in those particular places.  This of course would still account for the overwhelming majority of the NT text.

Conclusions and Implications

And so ends our survey of several important studies relevant to NT textual criticism.  We may summarize our findings as follows:

1) As argued by David C. Parker, the text of the Gospels was in a sense a “living text”, being subject to textual variations throughout its textual history (including the earliest centuries - owing I think in large part to the fact that the oral traditions underlying these texts were also not passed along in a precise manner, as can be demonstrated by readings of parallel Gospel passages).  Harmonization of one Gospel text to another was the most common corruption.

2) Some of the textual changes in the manuscript tradition were not only deliberate but also were theologically-motivated, as Bart Ehrman has demonstrated at length. 

3) Nevertheless, the textual tradition with its thousands of copies in Greek and other versions as well as text-critical tools utilized by all textual scholars allows for the reconstruction of about 98-99% of the original text from among the pool of variants.  Of the remaining “uncertain text”, the original is in the great majority of cases preserved in the known variants, though perhaps in a few rare cases “conjectural emendation” may be necessary to establish the original text.

4) Eldon Jay Epp’s fascinating discussion regarding the “multivalence” of the original text raises some interesting possibilities, some of which may to a certain degree obscure our definition of “original text”.  If it is true, for instance, that Paul (or more likely his amanuensis) prepared 10 initial copies of a certain epistle like Ephesians, then which copy is original?  Such postulations are speculative though possible, yet the implications for the preservation of important Christian doctrines are not profound either way.

5) William Petersen has argued for the importance of the Patristic evidence, especially that of the 2nd century, given that it could potentially provide us with a layer of text which precedes chronologically all known MSS.  On the other hand, Michael Holmes demonstrates why a great deal of diligence and above all caution is required when evaluating this Patristic evidence.

The Challenge for Christians

Put quite simply, Christians need to be aware of the primary findings of New Testament textual criticism.  That the New Testament text (and especially the Gospels, and even more particularly the sayings of Jesus) was subject to textual variation, some of which were not only deliberate but also theologically-motivated, and that important textual variations exist (such as the spurious endings of Mark) requires appreciation by Christian evangelists.  It must also be understood that the critical editions of the Greek NT are perhaps best described as “works-in-progress”.  This is evident from the changes which take place from edition to edition as new evidence is discovered and considered.  We do nobody a favor by pretending these issues do not exist.  It is important to come to grips with these issues in order to provide guidance for our brothers and sisters in the faith that may struggle as a result of these problems as well as others outside the faith who find the matter of textual variations an obstacle to converting to the Christian faith.

Regardless of the aforementioned problems, the vast majority of the original text can be reconstructed with a high degree of confidence, the remaining uncertain portions not challenging the more crucial Christian doctrines, certainly not the ones that contribute most to the rift between Christian and Islamic theology.  Furthermore, while new evidence or fresh considerations of existing evidence may provide reasons to change our opinion of the “original text” of this or that verse, the impressive body of evidence that we do have renders it unlikely that our critical texts will undergo major alterations in the future.  In regards to this Bart Ehrman writes:

Textual scholars have enjoyed reasonable success at establishing, to the best of their abilities, the original text of the NT.  Indeed, barring extraordinary new discoveries (e.g., the autographs!) or phenomenal alterations of method, it is virtually inconceivable that the physiognomy of our printed Greek New Testaments is ever going to change significantly (Ehrman 1995; 375). 

In the end I think the major Christian challenges lie not in text-critical issues, but rather the issues at the pre-literary stage, that is the time when the traditions were transmitted primarily through oral means.  The main problems with the particular passages discussed by Parker (i.e. the sayings on marriage and divorce and the Lord’s Prayer), for example, are not in the end with determining the most probable original text, but in explaining why there are differences between what was originally written by Mark, Matthew, and Luke.  These matters are important but it must be kept in mind that this is a quite separate issue from that of the preservation of what was originally written by the New Testament authors.

The Challenge for Muslims

Certain Muslim polemicists have been quite successful in bringing to light some of the major issues in New Testament textual criticism.  As I’ve expressed before, I think their efforts in this endeavor often prove to reflect the findings of the best scholarship on these important matters and should serve as an important warning against Christian complacency on this and other important aspects of scholarship. 

Nevertheless, despite the issues raised by aspects of the textual tradition it remains true that the NT has been very-well preserved.  As a result, I will reiterate the options that Muslims seem to be left with as I also had set out at the conclusion of the main article.  Since the evidence overwhelmingly indicates that the canonical New Testament texts attest to the important Christian doctrines that are at odds with Islamic claims, I argued in the last article that the only potentially intellectually-viable options for Muslim apologists are to either attempt to demonstrate that the “Gospel” alluded to by Muhammad in the Qur’an is a non-canonical Gospel or to adjust Islamic theology to accommodate these Christian doctrines.  Of course, such proposals would render rather difficult challenges, such as e.g. in the case of the former demonstrating the existence of another Gospel not only compatible with Islamic claims but also demonstrating why it should be regarded as more historically accurate than the canonical Gospels (given our current state of knowledge I think this would prove a daunting task).  In light of the discussion by Eldon Jay Epp, a similar proposal to the “non-canonical” option may be possible, however.  We discussed pre-Gospel sources such as Ur-Markus and Q (as well as the possibility of deutero-Gospel sources).  Could one argue that such a source (or sources) was/were the “Gospel(s)” alluded to by Muhammad in the Qur’an?  Given the reasons we discussed in the Epp section, I think attempting to establish this argument would be an unenviable task as well.

Outside of such “alternative proposals”, if Muslims wish to utilize sound text-critical scholarship to establish their case for substantial Biblical corruption, they need to show us specific textual variants, why such variants should be regarded as original, and simply why we should care about the finding(s) in question.  In other words, how does the variant affect major Christian doctrine, particularly the doctrines that result in the impasse between Christian and Islamic theology?  An approach similar to William Petersen’s would I think be the most relevant of the studies we’ve considered in this article.  Again, outside of any Muslim wishing to take up an “alternative proposal”, I think that delving into such specifics is the only way to move forward from this point for the Muslim-Christian debate regarding Biblical preservation/corruption.   

The Landscape of the Theological Debate

We will close with the following observations.  The data as it exists to date regarding NT textual criticism does not seem to warrant substantial change to the approach of Christians and Muslims in hashing out the important theological differences between the two faiths as it revolves around the person of Jesus.  The crucifixion and resurrection of Jesus cannot be excised from the “original NT” (assuming we’re talking here of the canonical books in their autograph forms) based on text-critical data.  These doctrines are not only confirmed or presupposed in dozens of undisputed readings across most of the canonical NT documents, but they are so prevalent that they serve as the proverbial backbone of Christian theological belief.  To a somewhat lesser extent we can say that this is similarly the case with the doctrine of Jesus’ atoning death. 

Christology is a different matter since Muslims commonly argue that even the “corrupted” canonical Gospels (i.e. what we are actually reading in our hands today) contain indications (particularly in some of Jesus’ sayings) that Jesus is not divine.  We could not disagree more but that is beyond our scope.  The important point is that the main arguments Christians utilize in order to indicate Jesus’ divinity cannot be excised based on textual variations.  Jesus is still the exalted “Son of Man” of Daniel 7:13-14.  Jesus still gives indications of his pre-existence (such as the use of what Gathercole termed the “I have come-plus-purpose formula”) [On both of these points cf. Appendix II of the main article].  The same could be said regarding the material which indicates Jesus’ divinity (or aspects consistent with divinity) in the New Testament epistles.  Again, there are simply too many undisputed readings that allude to such themes to make a reasonable argument that they were not originally present.  In fact, I would make an educated guess that virtually no aspect of Sam Shamoun’s materials on Christology is impacted by known textual variation.  The burden of proof clearly rests with those that would argue in favor of Biblical corruption when debating a given theological issue.


[1] To fully appreciate the nature of most of the differences cf. Parker 1997; 31-48.  This contains translations of Luke 6:1-11 as found in representatives of the Alexandrian (Codex Vaticanus) , “Western” (Codex Bezae), and Byzantine (Codex Athous Dion) traditions.  While the differences between the versions are many, it should be most appreciated that these are primarily differences in wording and/or ordering of material.  The content, however, is essentially the same in each version, with the exception being that in Codex Bezae there is a brief account of a certain man working on the Sabbath that is not found in any other manuscript.  This serves to emphasize the important point that while we speak of “living texts and traditions”, we should remember that generally speaking the changes do not result in essentially-different narratives or teachings, much less do they tend to have profound theological ramifications. 

[2] In light of the arguments presented by textual scholars that we summarized here regarding the “tenacity of the textual variants” as well as the intrinsic likelihood that the original text would survive somewhere among the rich manuscript tradition, I think Parker’s use of “earliest attainable text”, while technically an accurate way of putting things, remains overly and unnecessarily cautious. 

[3] Most of the relatively few places where Parker seems to deviate from the majority of scholars in specific text-critical decisions is found in chapter 9 which contains a discussion of the last 3 chapters of Luke’s Gospel.  This is most particularly the case with the so-called “Western non-interpolations” (cf. ibid. 151-157; 165-171).  These passages are Mt. 27:49; Lk. 22:19b-20; 24:3, 6, 12, 36, 40, 51, & 52.  However, this has been a very controversial issue since Westcott and Hort in the 19th century (cf. Metzger 1994; 164-166).  In fact, Metzger reveals that a minority of the UBS committee generally favored the originality of the shorter readings (supporting Parker) since “there is discernible in these passages a Christological-theological motivation that accounts for their having been added, while there is no clear reason that accounts for their having been omitted” while the majority “having evaluated the weight of the evidence differently, regarded the longer readings as part of the original text” (ibid. 166).   

[4] cf. Dunn’s thought-provoking discussion of Q and the Q community(ies) in Dunn 2003; 147-160.  Dunn elaborates upon the issues discussed here and raises other pertinent questions/issues as well.        

[5] In source-critical theory “M” denotes the material peculiar to Matthew’s Gospel while “L” denotes the material peculiar to Luke’s Gospel. 

[6] Interestingly, there are a number of lines marked with umlauts that do not correspond to variants listed in the 27th edition of Nestle-Aland.  While NA27, like other critical editions, is far from exhaustive, it contains the most prodigious textual apparatus established to date with 70,000 or so of the most important textual variations.  As such, this could indicate that the scribe of Vaticanus knew of a number of textual variations that are not found in current MSS.  If this is the case then the Alands’ claim of the tenacity of all textual variants that have been introduced at some point to the NT manuscript tradition (see the relevant section of our main article) is undermined.  Now, it would make sense that countless singular readings could have been introduced and not preserved either because the manuscripts in which these singular readings occurred were subsequently either not used as an exemplar for further copies or that when it was so used the scribes corrected the reading against another manuscript containing another reading.  As a whole I would argue that the Alands’ argument of tenacity applies (with a few possible exceptions where conjectural emendation may be required) best to the original reading given that it is this reading that would have occurred in a number of MSS bound for multiple destinations throughout the Roman Empire at a relatively early date, certainly at least within a generation or so of the time of original writing.    

[7] To be noted here is that this rendering of Hebrews 5:7 does not come directly from the translation of Philoxenus, but rather a commentary, yet as Baarda points out: “…but we know that this form of text would have been the one that he opted for as an exact rendering of the Greek text.”


Aland, Kurt & Barbara.  “The Text of the New Testament”.  2nd ed.  Transl. Erroll F. Rhodes.  Wm. B. Eerdmans Publishing Company.  1989.

Baarda, Tjitze.  “The Syriac Versions of the New Testament”.  in The Text of the New Testament in Contemporary Research.  Essays on the Status Quaestionis.  Ed. Bart D. Ehrman & Michael W. Holmes.  Wm. B. Eerdmans Publishing Co.  1995.  97-112.

Birdsall, J. Neville.  Book Review of “The Orthodox Corruption of Scripture:  The Effect of Early Christological Controversies on the Text of the New Testament” by Bart. D. Ehrman.  Theology.  97.780.  1994.  460-462. 

Dunn, James D. G.  “Jesus Remembered:  Christianity in the Making”.  Vol. 1.  William B. Eerdmans Publishing Company.  Grand Rapids, Michigan / Cambridge, U.K.  2003. 

Ehrman, Bart D.  “The Orthodox Corruption of Scripture:  The Effect of Early Christological Controversies on the Text of the New Testament”.  Oxford University Press.  1993.

Ehrman, Bart D.  “The Text as Window:  New Testament Manuscripts and the Social History of Early Christianity”.  In The Text of the New Testament in Contemporary Research.  Essays on the Status Quaestionis.  Ed. Bart D. Ehrman & Michael W. Holmes.  Wm. B. Eerdmans Publishing Co.  1995.  361-379.

Epp, Eldon Jay.  “The Multivalence of the Term ‘Original Text’ in New Testament Textual Criticism”. In Perspectives in New Testament Textual Criticism:  Collected Essays, 1962-2004.  Ed. Eldon Jay Epp.  Brill Academic Publishers.  June 2005.  551-593.   

Evans, Craig A.  “Fabricating Jesus:  How Modern Scholars Distort the Gospels”.  IVP Books.  2006.

Goodacre, Mark & Perrin, Nicholas.  “Questioning Q:  A Multidimensional Critique”.  InterVarsity Press.  2004. 

Holding, James Patrick.  “Trusting the New Testament:  Is the Bible Reliable?”  Xulon Press.  2009.

Holmes, Michael W.  “Textual Transmission in the Second Century”.  Audio version of Greer-Heart Point-Counterpoint Forum.  April 5, 2008. 

Hurtado, Larry W.  “Lord Jesus Christ:  Devotion to Jesus in Earliest Christianity”.  William B. Eerdmans Publishing Company.  2003. 

Kloppenborg, John S.  “Q the Earliest Gospel:  An Introduction to the Original Stories and Sayings of Jesus”.  Westminster John Knox Press.  2008.

Metzger, Bruce M.  “A Textual Commentary on the Greek New Testament”.  2nd ed. United Bible Societies.  1994.

Miller, J. Edward.  “Some Observations on the Text-Critical Function of the Umlauts in Vaticanus, with Special Attention to 1 Corinthians 14.34-35”.  Journal for the Study of the New Testament.    26.2.  2003.   217-236.

Parker, David C.  Book review of “The Orthodox Corruption of Scripture.  The Effect of Early Christological Controversies on the Text of the New Testament” by Bart D. Ehrman.  Journal of Theological Studies.  Vol. 45.  No. 2.  1994.  704-708.

Parker, David C.  “The Living Text of the Gospels”.  Cambridge University Press.  1997.

Parker, David C.  “An Introduction to the New Testament Manuscripts and Their Texts”.  Cambridge University Press.  2008.    

Petersen, William L.  “The Diatessaron of Tatian”. In  The Text of the New Testament in Contemporary Research.  Essays on the Status Quaestionis.  Ed. Bart D. Ehrman & Michael W. Holmes.  Wm. B. Eerdmans Publishing Co.  1995.  77-96. 

Petersen, William L.  “What Text Can New Testament Textual Criticism Ultimately Reach? In New Testament Textual Criticism, Exegesis, and Early Church History.  Ed. Barbara Aland and Joel Delobel.  Pharos.  1994.  136-152.

Petzer, Jacobus H.  “The Latin Version of the New Testament”. in The Text of the New Testament in Contemporary Research.  Essays on the Status Quaestionis.  Ed. Bart D. Ehrman & Michael W. Holmes.  Wm. B. Eerdmans Publishing Co.  1995.  113-130.

Polag, Athanasius.  “The Theological Center of the Sayings Source”.  in The Gospel and the Gospels.  Ed. Peter Stuhlmacher. William B. Eerdmans Publishing Company.  Grand Rapids, MI.  1991.  97-105. 

Robinson, James M.  “The Gospel of Jesus:  In Search of the Original Good News”.  HarperCollins Publishers.  2005.

Schoedel, William R.  “Ignatius of Antioch:  A Commentary on the Letters of Ignatius of Antioch”.  Ed. Helmut Koester.  Fortress Press.  1985.   

Silva, Moises.  “Response”. in Rethinking New Testament Textual Criticism.  Ed. David Alan Black.  Baker Academic.  2002.  141-150. 

Wisse, Frederik.  “The Coptic Versions of the New Testament”.  In The Text of the New Testament in Contemporary Research.  Essays on the Status Quaestionis.  Ed. Bart D. Ehrman & Michael W. Holmes.  Wm. B. Eerdmans Publishing Co.  1995.  131-141.