Tampering with the Text: Was the New Testament Text Changed Along the Way? (part 1)
- Andreas J. Kostenberger and Michael J. Kruger Authors
- 2010 7 Jul
EDITOR'S NOTE: The following is an excerpt from The Heresy of Orthodoxy by Andreas J. Kostenberger and Michael J. Kruger (Crossway).
The only way that the New Testament books (and any type of writing) could be broadly circulated in the ancient world was if they were first copied by hand. A scribe would have to sit down with the original document and copy it word for word onto a piece of papyrus or parchment.1 Of course, in our modern day, well after the time of Gutenberg's printing press, such dependence on handwritten manuscripts seems strange to us. We give little or no thought to how a book is copied and assume that whichever copy of a book we pick off the shelf will look identical to every other copy. In ancient times, however, it was quite normal (and even expected) that scribes, no matter how professional, would occasionally make mistakes.2 These scribal variations—slips of the pen, misspellings, word order changes, etc.—were an inevitable part of literary life in a pre-Gutenberg world (and even, toa lesser degree, in a post-Gutenberg world). Fortunately, as seen in the previous chapter, we have good reasons to think that early Christians possessed a solid scribal infrastructure that would have minimized the impact of such variations. Nevertheless, we still need to examine the New Testament manuscripts themselves. Are these manuscripts very different from one another? Are there reasons to think the text has been substantively changed along the way? And did the early Christian battles over heresy and orthodoxy affect the transmission of the text? It is the purpose of this chapter to answer these questions.
It is important that we begin by noting that some scholars have already given an answer. Bart Ehrman would answer "yes" to all of the above questions. In his book Misquoting Jesus, Ehrman argues that the New Testament manuscripts are so riddled with scribal errors and mistakes (some even intentional) that there is no way to have any certainty about the words of the original authors. In essence, he argues that the New Testament text has been changed—irreparably and substantially changed in the battles over heresy and orthodoxy—so that it is no longer meaningful to discuss what Paul, or Matthew, Mark, or Luke, wrote. We simply do not know. All we have are manuscripts. And these manuscripts date hundreds of years after the time of the apostles and vary widely from one another. So, what does the "New Testament" say? It depends, says Ehrman, which manuscript you read. He declares, "What good is it to say that the autographs (i.e., the originals) were inspired? We don't have the originals! We have only error-ridden copies, and the vast majority of these are centuries removed from the originals and different from them . . . in thousands of ways."3
Although Ehrman presents his who-knows-what-the-text-originally-said approach as part of mainstream textual criticism, it actually stands in direct opposition to many of his fellow scholars in the field (and even seems to be out of sync with his own writings elsewhere). Historically speaking, the field of textual criticism has not embodied the hyper-skepticism evident in Misquoting Jesus but has been more optimistic concerning the recovery of the original text (or at least something very close to it).4 In response to Ehrman, therefore, this chapter will put forward four theses that embody an approach that is more consistent with the kind traditionally taken in the field of textual criticism.
- We have good reasons to think the original text is preserved (somewhere) in the overall textual tradition.
- The vast majority of scribal changes are minor and insignificant.
- Of the small portions of variations that are significant, our text-critical methodology can determine, with a reasonable degree of certainty, which is the original text.
- The remaining number of truly unresolved variants is very few and not material to the story/teaching of the New Testament.
If these four theses are valid, then we have good reasons to think that we are able to recover the New Testament text in a manner that is so very close to the original that there is no material difference between what, say, Mark and Matthew wrote and the text we have today. Although we can never have absolute certainty about the original text, we can have sufficient certainty that enables us to be confident that we possess the authentic teaching of Jesus and his apostles. Let us consider each of these theses in turn.
Thesis 1: The Wealth of Extant Manuscripts: we have good reasons to think the original text is preserved (somewhere) in the overall textual tradition.
The first step in answering these questions about the transmission of the New Testament text is to gain a better understanding of the manuscript resources at our disposal. Discussions about whether a text has been "changed" always involve the comparison of manuscripts. After all, if we only possessed a single manuscript of the New Testament, there would be no discussion of scribal variations and changes—we would not know of such things unless we compared one copy with another copy to see where they differ.5 Although such a scenario may, on the surface, seem desirable (because then we would not need to worry about debating which variants were original!), having only one manuscript would raise a substantial problem: how would we know that we possess, in this one single manuscript, the words which were originally written by the author? If this single manuscript were simply a later copy of the original (which is most likely the case), then there is a good chance that some scribal mistakes, errors, and other variants have slipped into the text during the copying process. With only a single manuscript in our possession there is no way to be sure that no words have been lost or altered. Therefore, as scholars seek to know how much any writing of antiquity has been changed, and, more importantly, as they seek to establish what that writing would have originally said (by tracing those changes through the manuscript tradition), the more manuscripts that can be compared the better. The higher the number of manuscripts, the more assurance we have that the original text was preserved somewhere in the manuscript tradition.
But it is not just the high quantity of manuscripts that is desirable for the textual critic but manuscripts that date as closely as possible to the time of the original writing of that text. The less time that passed between the original writing and our earliest copies, the less time there was for the text to be substantially corrupted, and therefore the more assured we can be that we possess what was originally written. Unfortunately, these two components of every textual critic's wish list—numerous copies and also some with an early date—are relatively rare in the study of most documents of antiquity. As we shall see, most of our ancient historical sources are attested by few manuscripts that are often very late.
The Quantity of New Testament Manuscripts
Not surprisingly, ancient manuscripts are hard to come by. Most have perished over the ages for a variety of reasons—burned in garbage dumps, destroyed by foreign armies, rotted or decayed, damaged by insects or rodents—or have simply been lost.6 Historians never have as many pieces of evidence as they would like. For example, the writings of Tacitus from the first century, widely recognized as one of the greatest Roman historians, survive in only three manuscripts, and not all are complete.7 Consider also the writings of Gaius from the second century, a Roman jurist who is well known for his essential accounts of Roman law under emperors like Marcus Aurelius. Most of his writings are lost and his key work, The Institutes, is preserved in just three manuscripts—but the text "rests almost exclusively" on just one of them.8 The sizable History of Rome by the first-century historian Velleius Paterculus, which covers large portions of Roman history, including the life of Julius Caesar, comes down to us in a single, mutilated manuscript.9 The work Jewish War by Josephus, a trusted Jewish historian from the first century AD, is better attested with over fifty extant manuscripts, but the text is mainly dependent on about ten of them.10
By contrast, the New Testament manuscripts stand out as entirely unique in this regard. Although the exact count is always changing, currently we possess over 5,500 manuscripts (in whole or in part) of the New Testament in Greek alone.11 No other document of antiquity even comes close. Moreover, we possess thousands more manuscripts in other languages. The total for just our Latin manuscripts of the New Testament exceeds ten thousand copies, and we possess thousands more in Coptic, Syriac, Gothic, Ethiopic, Armenian, and other languages.12 Indeed, there is no exact number because there are so many of these different versions that not all have been formally catalogued. In addition to all these manuscripts, there are also a countless number of citations of the New Testament preserved in the early church fathers,13 so many, in fact, that Metzger has famously declared, "So extensive are these citations that if all other sources for our knowledge of the text of the New Testament were destroyed, they would be sufficient alone for the reconstruction of practically the entire New Testament."14
Such a scenario, from a historical perspective, is truly remarkable. As Eldon Epp has declared, "We have, therefore, a genuine embarrassment of riches in the quantity of manuscripts we possess. . . . The writings of no Greek classical author are preserved on this scale."15 If there were ever an ancient writing that had enough extant manuscripts that we could be reasonably assured that the original text was preserved for us in the multiplicity of copies, the New Testament would be it. Again it is Epp who notes, "The point is that we have so many manuscripts of the NT . . . that surely the original reading in every case is somewhere present in our vast store of material."16 Fee concurs, "The immense amount of material available to NT textual critics . . . is their good fortune because with such an abundance of material one can be reasonably certain that the original text is to be found somewhere in it."17 In other words, due to the vast number of manuscripts, the challenge of textual criticism is a different one than we might expect—it is not that we are lacking in material (as if the original words were lost), but rather we have too much material (the original words, plus some variations). When it comes to reconstructing the original text of the New Testament, the latter position is much preferred over the former.
It is here that the contrast between the New Testament and classical works becomes acute. Ehrman's hyper-skeptical approach should be challenged not by insisting the New Testament text should be treated in the same way as classical works—for he may argue that we do not know the text of the classical authors either—but by insisting that the New Testament text should be treated differently. After all, if we supposedly lack assurance regarding the preservation of the classical texts due to their paucity of manuscripts (although it is doubtful whether scholars really do treat classical works with such agnosticism), then how could we not have much greater assurance of the preservation of the New Testament text due to its abundance of manuscripts? This is precisely the sticking point for Ehrman's position. He wants to be skeptical of both sets of writings (New Testament and classical), in spite of the fact that the historical evidence for the two is vastly different. To insist that the New Testament is as unknowable as classical works is to render the historical data utterly irrelevant to the discussion at hand. Such a position, at its core, proves to be substantively unhistorical—the conclusions are the same regardless of the evidence.
It is precisely for this reason that one wonders how much textual material would be enough for Ehrman to regard a text as sufficiently knowable. Would seven thousand Greek manuscripts be enough? Ten thousand? What if we had many more manuscripts of an early date (more on this below)? Would that be enough? One gets the impression that no matter what the evidence is, it would not change the outcome. The bar always seems to be set just a bit higher than wherever the evidence happens to be—like the Greek myth of Sisyphus who thought he had finally done enough to push the boulder to the top of the hill only to find it rolled back down again. As we shall see, there is only one thing that would seem to satisfy Ehrman's requirements: the autographs themselves.
The Date of the New Testament Manuscripts
If manuscripts of ancient documents are (generally speaking) relatively rare, then early manuscripts are even more so. As noted above, the smaller the gap of time between the writing of an ancient text and our earliest copy of that text, the more assurance we have that we possess what was originally written. Unfortunately, small gaps of time are the exception and not the rule. Of the manuscripts of Tacitus, the earliest is ninth century, nearly eight hundred years after it was originally written.18 For Josephus's Jewish War, virtually all of its manuscripts are from the Middle Ages, and the earliest of these is from the tenth century, nearly nine hundred years after the original time of publication. The only manuscript earlier than this is a very fragmentary papyrus from the third century that is virtually illegible.19 The single extant manuscript of the History of Rome by Velleius Paterculus is dated to the eighth or ninth century—approximately eight hundred years after its initial publication—but was subsequently lost and now survives only in a sixteenth century copy.20 The primary manuscript for Gaius's Institutes fares a bit better and is dated to the fifth century, about three hundred years after the original.21 Such gaps of time are not unusual in the manuscript traditions of many of our classical works. As Epp sums it up, "As is well known, the interval between the author and the earliest extant manuscripts for most classical writings is commonly hundreds—sometimes many hundreds—of years, and a thousand-year interval is not uncommon."22
However, again, the New Testament situation is entirely different. The New Testament was written approximately AD 50-90, and our earliest New Testament manuscript, P52, preserves a portion of John's Gospel from c. AD 125, only thirty-five years later.23 Other early manuscripts include P90 (John, second century), P104 (Matthew, second century), P66 (John, late second century24), P98 (Revelation, second century), P4-P64-P67 (Luke and Matthew, late second century25), P46 (Pauline epistles, c. AD 200), P103 (Matthew, c. AD 200), P75 (Luke and John, c. AD 200-22526), and many others. Of course, even our major fourth-century codices, Sinaiticus (א)and Vaticanus (B), which contain nearly the entire Greek Bible (Old and New Testaments), are still quite early compared to the manuscripts of most classical works.
The brief span of time between the production of the New Testament and our earliest copies gives us access to the New Testament text at a remarkably early stage, making it very unlikely that the textual tradition could have been radically altered prior to this time period without evidence for those alterations still being visible within the manuscript tradition.27 Put differently, if a particular manuscript of a New Testament book (say, Mark) had been changed by a scribe in the late first or early second century, it is unlikely that the change would have been able to replace the original reading quickly enough so that our third- and fourth-century copies of Mark would fail to preserve the original text at all (thus creating a situation where we would not even know the text had been changed). Frederik Wisse comments:
There is no indication that the Gospels circulated in a form different from that attested in the later textual tradition. . . . If indeed the text of the Gospels had been subjected to extensive redactional change and adaption during the second century, the unanimous attestation of a relatively stable and uniform text during the following centuries in both Greek and the versions would have to be considered nothing short of a miracle.28
The textual tradition of the New Testament, therefore, has a stubborn quality about it. Although a scribe can change an individual manuscript (or an individual reading), changing the overall textual tradition is much more difficult than one might think—the fact that there are so many other copies in circulation makes this virtually impossible to do. Kurt and Barbara Aland note that "one of the characteristics of the New Testament textual tradition is tenacity, i.e., the stubborn resistance of readings and text types to change. . . . This is what makes it possible to retrace the original text of the New Testament through a broad range of witnesses."29 Again they declare:
The transmission of the New Testament textual tradition is characterized by an extremely impressive degree of tenacity. Once a reading occurs it will persist with obstinacy. . . . It is precisely the overwhelming mass of the New Testament textual tradition which provides an assurance of certainty in establishing the original text.30
In other words, Aland and Aland are arguing that the multiplicity of witnesses, combined with the stubbornness of the textual tradition and the early date of our manuscripts, make it more than reasonable to presume that the original text is preserved within our overall manuscript tradition (even though any given copy would have variants31).
However, despite the fact that the New Testament text, again, has substantially earlier textual attestation than most any other document of antiquity, this still does not seem to satisfy Ehrman. For example, he argues that we cannot know that we possess the text of Galatians because our earliest copy (P46) was written nearly 150 years after the original was composed.32 One wonders, would Ehrman's conclusions change if, say, we had a copy of Galatians from the middle of the second century (c. AD 150) or even earlier? This seems unlikely. Elsewhere in Misquoting Jesus, Ehrman argues that we can never really know what Galatians says because it is possible that one of the very first copies of Galatians could have had a mistake and maybe all of our extant copies derive from that single faulty copy.33 Thus, armed with this hypothesis about what might have happened in the early stages of the transmission (a hypothesis that cannot be proven), Ehrman is always able to claim we can never know the original text, no matter how early our extant manuscripts are. Once again, we see how Ehrman's conclusions seem impervious to the historical evidence—the date of our manuscripts does not really matter because, in principle, the text of Galatians (or any book) can never really be known.
So, in the end, Ehrman's expressed concerns over the 150-year gap of time are somewhat of a red herring; they make the discussion appear to be about the historical data when it is really about an a priori decision never to acknowledge that a text can be sufficiently known unless we have 100 percent, unequivocal, absolute certainty. In other words, we can never claim knowledge of a text unless we have the autographs themselves (or a perfect copy of them). Needless to say, if this is the standard, then it will never be met in the real world of historical investigation.
Thesis 2: The Extent of Textual Variation: the vast majority of scribal changes are minor and insignificant
Although the prior discussion has many layers of complexity, the overall point is a simple one: the impressive quantity of New Testament manuscripts, combined with the early date of many of those manuscripts, makes it historically reasonable to conclude that we possess the original text of the New Testament within the overall textual tradition (though not necessarily in any single manuscript). Therefore, as noted above, we actually have too much information—we not only possess the original text but also many textual variants. With this, we transition into the next stage of the discussion. Now we are no longer dealing with the question of whether we have the original New Testament text in our manuscript tradition but how we separate the original text from the variants. Do these variants present a considerable problem? How many of these variants are there? How different are the manuscripts we possess?
One might think we could just add up all the textual variations and we would have our answer. However, as we shall see, the answer to these questions is not as simple as providing a numerical figure. All scholars agree that there are thousands of textual variants throughout our manuscripts—maybe as many as four hundred thousand—though no one knows the exact number. Ehrman seems eager to draw attention to this fact, if not to suggest even higher numbers: "Some say there are 200,000 variants known, some say 300,000, some say 400,000 or more!"34 Indeed, numbers matter very much to Ehrman. For him, the sheer volume of variants is the deciding factor and sufficient, in and of itself, to conclude that the New Testament cannot be trusted. He even offers the dramatic statement, "There are more variations among our manuscripts than there are words in the New Testament."35 However, Ehrman's statistical enthusiasm aside, mere numbers do not tell the whole story. When other factors are considered, a more balanced and full-orbed picture of the New Testament text begins to emerge.
The Nature of the Textual Changes
All textual changes are not created equal. This fact, of course, is the fundamental reason why a numbers-only approach to textual variants is simply not viable. We need to ask not only how many variants there are but what kind of variants there are. It is a question not simply of quantity but of quality. It is for this reason that Eldon Epp and other textual critics recognize thatthere are certain kinds of textual variants that can legitimately be regarded as "insignificant."36 This term simply refers to variants that have no bearing or no impact on "the ultimate goal of establishing the original text."37 These are typically minor, run-of-the-mill, scribal slips that exist in any document of antiquity (New Testament or otherwise) and thus occasion no real concern for the textual scholar—and certainly are not relevant for assessing whether a document has been reliably passed down to us. And here is the key: these "insignificant" variants make up the vast, vast majority of variations within the New Testament text.38 Categories of insignificant variants include the following:39
1) Spelling (orthographical) differences. It turns out that scribes in the ancient world often made spelling errors/changes just like writers in the modern day. Examples of this sort of change abound. (a) If certain words ended in a nu, that nu would often be dropped by the scribe if the following word started with a vowel (this is known as the moveable nu). But scribes were not always consistent with this practice and often differed from one another, and would even change patterns within the same manuscript. (b) Scribes used a variety of different abbreviations, and not all were identical. For example, if the last word in a line ended with nu, sometimes scribes would abbreviate it by dropping the nu and putting a horizontal line in its place.40 (c) Scribes would often interchange i and ie (or ei) in the spelling of words, which was often a form of phonetical spelling rather than a formal scribal error.41 And on it goes. The variety of spelling differences in manuscripts seems endless and every one of them counts as a scribal variation.42
2) Nonsense readings. Occasionally scribes would make a mistake that would render a verse nonsensical and thus the mistake can be quickly identified as not being the original reading of the text. For example, sometimes scribes would accidentally skip a line in their copying (called haplography), and this would create incoherent readings. A well-known example is found in John 17:15 of Codex Vaticanus (B), where the scribe skipped a line and left out the bracketed portion: "I do not ask that you take them from the [world, but that you keep them from the] evil one." Needless to say, this produces a nonsensical reading that is clearly not original! Such mistakes may tell us about habits of a particular scribe, but they have no bearing on our ability to recovery the original text.
3) Singular readings. Sometimes a certain reading exists in only one Greek manuscript and no other. Such singular readings—and there are thousands of them—have little claim to be the original text and therefore are irrelevant in assessing the reliability of the manuscript tradition. For example, P66* is the only (known) manuscript where John 17:12 has Jesus declare to the Father in his high priestly prayer, "I kept them in my (mou) name, which you have given me." All other manuscripts read, "I kept them in your (sou) name, which you have given me."
4) Meaningless word order changes. One of the most common scribal changes involves word order (known as transposition). Unlike English, Greek nouns are inflected and thus their function in the sentence is not determined by word order but by their case. Therefore, the vast majority of word order changes in Greek do not affect meaning at all. For example, again in P66, John 13:1 reads toutou tou kosmou ("this world"), whereas the original likely read tou kosmou toutou ("this world")—no difference in meaning whatsoever. Another common word order change, especially in the Pauline epistles, is "Jesus Christ" for "Christ Jesus," or vice versa. Every word order change (and every various possible combination) counts as a variant.
5) Definite articles on proper nouns. Unlike English, Greek can include articles in front of proper nouns: "the Jesus," "the John," or "the Andrew." However, there is no consistency in this practice among early Christian scribes and the presence or absence of the article before proper nouns rarely affects the meaning.43 For example, a number of manuscripts (A Δ f1 f13 1241) include the article (tou) in front of the name "Simon" in Mark 1:16, whereas most other manuscripts leave it out. Either way the English translation is the same: "Simon." Every time a scribe includes or omits an article in front of a proper noun, it counts as a textual variant.
Of course, this brief overview of insignificant scribal changes is not exhaustive, and other categories could be added (e.g., scribes replacing personal pronouns with their antecedents). But the overall point is clear. Even though these types of changes are quite abundant—Ehrman is correct about that—they are also quite irrelevant. Thus, simply adding up the total textual variations is not a meaningful exercise in determining the reliability of textual transmission.
Textual Changes and the Quantity of Manuscripts
The numbers-only approach to evaluating textual variants also fails to take into account another very critical piece of data: the impressive quantity of manuscripts we possess. Obviously, if we possessed only five Greek manuscripts of the New Testament, then we would have very few textual variations to account for. But if we have over five thousand Greek manuscripts of the New Testament (not to mention those in other languages), then the overall quantity of textual variants will dramatically increase because the overall number of manuscripts has dramatically increased. The more manuscripts that can be compared, the more variations can be discovered. Thus, the quantity of variations is not necessarily an indication of scribal infidelity as much as it is the natural consequence of having more manuscripts than any other historical text.
Incredibly, then, Ehrman takes what should be positive historical evidence for the New Testament (the high number of manuscripts) and, somehow, turns the tables to make it evidence for its tendentious character—a remarkable feat, to be sure. One wonders what Ehrman's conclusions would be if we actually did possess only five manuscripts of the New Testament and thereby had very few textual variants. Would the lack of textual variants then be regarded as positive evidence for the New Testament's reliable transmission? We suspect not. One wonders if the objection would then be that we have too few manuscripts. It is a losing affair either way. Thus, once again, we see a familiar pattern emerging. Regardless of the evidence—whether the manuscripts are many or few, whether the variants are many or few—Ehrman's conclusions would remain unchanged.
For footnotes from "Tampering with the Text," click here.
EDITOR'S NOTE: Excerpt taken from The Heresy of Orthodoxy by Andreas J. Kostenberger and Michael J. Kruger (Crossway).
Read part 2 of this series, "Tampering with the Text: Was the New Testament Text Changed Along the Way?" (part 2) by Andreas J. Kostenberger and Michael J. Kruger
The Heresy of Orthodoxy
Copyright ©2010 Andreas J. Kostenberger and Michael J. Kruger
Published by Crossway Books
A publishing ministry of Good News Publishers
1300 Crescent Street
Wheaton, Illinois 60187