FOREWORD<\/p>\n
Daniel B. Wallace<\/p>\n
We no longer live in a black-and-white world. We never did really, but those who are embroiled in debates about the Bible have often viewed things in such binary hues. These achromatic ideologies can be found on both sides of the theological aisle.
\nMany who have abandoned the unreflective beliefs they grew up with now cling\u2014just as unreflectively\u2014to unmitigated skepticism toward the New Testament text. The Dan Browns and Kurt Eichenwalds of our world can liken, with a straight face, the scribal copying of Scripture to the parlor game of Telephone. To them, the text has been corrupted so badly that attempting to recover the original wording is like looking for unicorns. It\u2019s an impossible task because the search is for something that does not exist.
\nOn the other hand, some apologists for the Christian faith speak of (nearly) absolute certainty when it comes to the wording in the New Testament. And laypeople routinely think of their Bible as the Word of God in every detail. They are blissfully unaware that Bible translations change\u2014because language evolves, interpretations that affect translation become better informed (and all translation is interpretation), and the text that is being translated gets tweaked. Biblical scholarship is not idle. Yet even the publisher of the ESV translation, extremely popular among evangelicals, contributed to this fictive certitude when it declared in August 2016 that \u201cthe text of the ESV Bible will remain unchanged in all future editions printed and published by Crossway.\u201d The next month it admitted, \u201cThis decision was a mistake.\u201d When a publishing house tries to canonize its Bible translation, what does this say to the millions of readers who know nothing of Greek, Hebrew, or Aramaic?
\nThese two attitudes\u2014radical skepticism and absolute certainty\u2014must be avoided when we examine the New Testament text. We do not have now\u2014in our critical Greek texts or any translations\u2014exactly what the authors of the New Testament wrote. Even if we did, we would not know it. There are many, many places in which the text of the New Testament is uncertain. But we also do not need to be overly skeptical. Where we should land between these two extremes is what this book addresses.
\nThe new generation of evangelical scholars is far more comfortable with ambiguity and uncertainty than previous generations. They know the difference between core beliefs and those that are more peripheral. They recognize that even if we embrace the concept of absolute truth, absolute certainty about it is a different matter.
\nOne word kept coming to mind as I read this book: nuance. The authors understand what is essential and of vital significance in the Christian faith and what is more peripheral. As Stephen Neill argued over fifty years ago and Peter Gurry affirms in this book, \u201cThe very worst Greek manuscript now in existence \u2026 contains enough of the Gospel in unadulterated form to lead the reader into the way of salvation.\u201d Andrew Blaski shows that the patristic writers, too, recognized this. Origen, whose concern to recover the original wording of the Bible was worked out with indefatigable exactness, had an even deeper concern. Many Fathers understood that the New Testament\u2014highly valued, revered, even apostolically authoritative\u2014nevertheless pointed ultimately to what is more revered, more authoritative, and more central to our faith: our God and Savior, Jesus Christ.<\/p>\n
THE CHASM BETWEEN SCHOLARS AND APOLOGISTS<\/p>\n
The authors in this book offer a necessary corrective to decades of overly exuberant apologetic arguments\u2014arguments that have actually hurt the Christian faith. The writers are refreshingly honest, and they do not pull their punches. They observe poignantly that apologetic works on the reliability of the New Testament text have been drifting away from a proper, well-researched, accurately documented scholarship that is anchored to actual data. Apologists have had a tendency to regurgitate other apologetic works, which in turn are based on other apologetic works. Meanwhile, the scholarship that is supposedly behind the popular declarations in many an evangelical trade book is out of date, misunderstood, or simply ignored.
\nA classic example of the disconnect between scholarship and apologetics is how textual variants are (mis)counted. A steady stream of apologists for more than half a century have been claiming that variants are counted by wording differences multiplied by manuscripts attesting them. Neil Lightfoot\u2019s How We Got the Bible, a book first published in 1963 and now in its third edition with more than a million copies sold, seems to be the major culprit. Lightfoot claims:<\/p>\n
From one point of view it may be said that there are 200,000 scribal errors in the manuscripts, but it is wholly misleading and untrue to say that there are 200,000 errors in the text of the New Testament. This large number is gained by counting all the variations in all of the manuscripts (about 4,500). This means that if, for example, one word is misspelled in 4,000 different manuscripts, it amounts to 4,000 \u201cerrors.\u201d Actually, in a case of this kind only one slight error has been made and it has been copied 4,000 times. But this is the procedure which is followed in arriving at the large number of 200,000 \u201cerrors.\u201d<\/p>\n
The only problem with this statement is that it is completely wrong. Chief among the errors, as Elijah Hixson and Peter Gurry point out, is that textual critics \u201care not counting the number of manuscripts that attest a variant; we are counting the number of variants attested by our manuscripts.\u201d If variants were actually counted the way Lightfoot suggests, the number of variants among the Greek New Testament manuscripts would be in the tens of millions. That this miscalculation has seeped its way, unchecked, into several apologetics books for more than five decades is a telling indictment on the uncritical use of secondary sources by many in this field.
\nAn example of using out-of-date statistics is found in the comparative argument\u2014that is, the argument that compares the number of New Testament manuscripts with those of other Greco-Roman authors. As James Prothro notes, several popular apologists have claimed that there are only 643 manuscripts of Homer\u2019s Iliad. This number got into apologists\u2019 hands, according to Prothro, most likely via a technical book Bruce Metzger authored in 1963. But Metzger repeats the same number of Iliad manuscripts in the far-more-accessible first (1964), second (1968), and third edition (1992) of his The Text of the New Testament. The fourth and latest edition (2005), coauthored with Bart Ehrman, continues to speak of only 643 manuscripts for the Iliad! It is not just apologists, then, but sometimes even top-flight scholars who have added to text-critical myths. On the other hand, in the latest edition of Evidence That Demands a Verdict, published in 2017 and coauthored with his son Sean, Josh McDowell did his due diligence to update the number of Iliad manuscripts by consulting classicists and the Leuven Database of Ancient Books\u2014exactly the right approach.
\nOther myths that are often touted get some schooling. Jacob Peterson goes into impressive detail on why the \u201cofficial\u201d number of Greek New Testament manuscripts (i.e., the tally made by adding all the catalogued numbers of papyri, majuscules, minuscules, and lectionaries by the Institut f\u00fcr Neutestamentliche Textforschung), often cited as the actual number, is way too generous. But the stats are not static. Peterson commends the work of the Center for the Study of New Testament Manuscripts for adding significantly to our fund with its digitizing of dozens of newly discovered manuscripts.
\nGregory Lanier bursts some bubbles about the supposed inferiority of later manuscripts, as though age necessarily corresponds to intrinsic value. He adds that even the later Byzantine manuscripts speak well of \u201cthe fidelity of the entire textual tradition.\u201d Lanier gives a quite helpful table on the most significant minuscules and, in the spirit of G\u00fcnther Zuntz, puts forth diagrams that illustrate many facets of transmissional fidelity.
\nThe number of versional manuscripts (those written in other than Greek) has been routinely specified without documentation by apologists and some scholars. Jeremiah Coogan reels back the sensationalism and grounds the numbers in what is known. He also addresses some key issues in what the versions can and cannot do to aid us in recovering the autographic wording of the Greek New Testament.
\nDemonstrating sensitivity to the priorities of scholars, translators, and faith communities, Edgar Ebojo discusses the dialogue that takes place behind the scenes. The scholarly guild has a say, but it is not the only say in how translations should look and what texts should be included in the translation.<\/p>\n
MYTHS AND MISTAKES MADE BY SCHOLARS<\/p>\n
Hixson and Gurry tell us that the authors of this book \u201cwrite primarily as a self-corrective to Christian speakers and writers.\u201d As I was perusing the manuscript, I came to the conclusion that the editors have defined their readership too narrowly. Precisely because the contributors are up-and-coming scholars\u2014with PhDs (earned or in process) from Birmingham University, Cambridge University, Dallas Seminary, University of Edinburgh, New Orleans Baptist Seminary, University of Notre Dame, and Southern Baptist Seminary\u2014they are up to date on the state of their disciplines. Many have written their doctoral theses on the very topics they explore in this book. These young scholars have something to say\u2014not only to Christian speakers and writers but to non-Christian speakers and writers, and even to New Testament scholars of all stripes.
\nI have been working for several years on an introduction to New Testament textual criticism. Many of the topics discussed in Myths and Mistakes are those I have felt needed some treatment in such an introduction. I was happily stunned to see the depth of discussion, the candid examination, and the up-to-date bibliography in each chapter. Although Myths and Mistakes is written in clear, user-friendly prose, the contents are well grounded and perspicacious. I intend to use this volume unapologetically in my introduction as a primary source for several analyses.
\nThere are no sacred cows here. Occasionally, even scholars who have delved into the realm of apologetics have been a bit too enthusiastic, naive, or biased. All of us can learn something from this volume. Craig Evans\u2019s view on the longevity of the autographs is perceptively analyzed by Timothy Mitchell; Michael Kruger\u2019s link between canon and codex is critiqued by John Meade; Philip Comfort\u2019s early dating of papyri is challenged by Elijah Hixson; and I am not immune from censure. Chief among such criticisms (but by no means the sole issue) is my mention of a first-century fragment of Mark\u2019s Gospel in one of my debates with Bart Ehrman. I had it on good authority that the date was firm and that the papyrus would be published in a year. But at the time I had not seen the manuscript, which should have been critical for me in making any statements about its date. Six years later (!) the fragment was published (April 2018), and it turned out not to be from the first century but was dated to the second or third century by the editors.
\nBart Ehrman, a first-rate scholar and an outspoken skeptic about recovering the original New Testament text, comes in for some specific criticisms too. His \u201cearly orthodox corruptions\u201d are seen to be less frequent and less severe when Robert Marcello applies a more rigorous method to some key textual problems.
\nEhrman\u2019s claim that the early scribes were not professionally trained and therefore did not make careful copies is handled by Zachary Cole. Ehrman\u2019s view is overly simplistic, presenting a multicolored reality as black and white, and is often factually wrong.
\nPeter Malik boldly takes on E. C. Colwell, whose studies on method are legendary, by documenting corrections in papyri that \u201cshow that scribes strove to improve and revise their work before they handed it to posterity.\u201d It is not just what the scribe originally penned but the corrections he or she made to the codex before releasing it to other readers that demonstrate this care.<\/p>\n
REQUIEM FOR A DISCIPLINE?<\/p>\n
Forty years ago, Eldon Epp published a disturbing article in the Journal of Biblical Literature, in which he predicted the end of text-critical studies in America. \u201cNew Testament Textual Criticism in America: Requiem for a Discipline\u201d canvasses the trends of a downward spiral pertaining to detailed study of the text and the lack of opportunity for writing a dissertation in textual criticism. He concluded his essay with this gloomy outlook: \u201cI may have pushed too far the figure of speech in the subtitle of this paper when I chose the expression, \u2018Requiem for a Discipline.\u2019 Yet, that ominous eventuality is all too likely should the clear trends of the recent past continue even into the near future.\u201d
\nJust a decade back, the field appeared almost desolate. At a two-day colloquium held in August 2008 at the Institut f\u00fcr Neutestamentliche Textforschung in M\u00fcnster, at most a few dozen New Testament textual scholars were present. I inquired from one of the organizers of the event about the list of those invited. She informed me that all textual critics worldwide were on the list and that only one had declined the invitation. To be sure, the American representation had improved since Epp\u2019s requiem, but the numbers were still small.
\nSamuel Clemens, when rumors that he was on his deathbed were circulating, wrote, \u201cThe report of my death was an exaggeration.\u201d The same can be said of American\u2014as well as international\u2014scholarship in New Testament textual criticism. A sea change has transpired in the last ten years. Not all the contributors to Myths and Mistakes are Americans, of course, but most are. Further, evangelicals in particular have dedicated themselves to this discipline.
\nEpp spoke of \u201cthe growing lack of concern and support for NT textual criticism in America.\u201d I was finishing my ThM degree when he wrote these words. As I\u2019m sure that several other graduate students did, I took his requiem to heart. It was a sobering and swift kick in the derri\u00e8re! I am delighted to report that, forty years later, the scenery has improved markedly. Four of my former students have contributed to this publication: Peter Gurry, Zachary Cole, Robert Marcello, and Jacob Peterson. They either interned at CSNTM, wrote their master\u2019s thesis on an aspect of New Testament textual criticism, or both. Elijah Hixson also worked for CSNTM on a digitizing project. Peter Malik collaborated with CSNTM at the Chester Beatty Library as we digitized P47, the topic of Malik\u2019s doctoral thesis. Two other interns, Matthew Larsen and Brian Wright, whose doctoral dissertations earn a shout-out or rebuttal in Timothy Mitchell\u2019s chapter, also earned their PhDs in New Testament textual criticism or its kin. A certain paternal pride comes with these declarations, but I am hardly alone. Other American professors who specialize in textual studies can claim a measure of mentorship to several of these authors.
\nOne of Epp\u2019s complaints in his requiem is that \u201cthe company of trained collators rapidly has disintegrated.\u201d Collation is an accurate recording of the exact wording of each manuscript via registering its differences from a base text. At the time that Epp filed this complaint, the number of New Testament books whose manuscripts had been completely collated was one. Only the Apocalypse received this honor, a monumental task accomplished by Herman Hoskier in 1929 after thirty years of painstaking labor. Furthermore, virtually nothing has been published on the text of the great majority of codices of the New Testament. A look at Elliott\u2019s Bibliography of Greek New Testament Manuscripts reveals that less than one-fourth of all extant Greek New Testament manuscripts have even a paragraph published on them. Collations are necessary for every one of these documents.
\nA collation of a given manuscript not only reveals the differences between said codex and the chosen base text but also unmasks scribal proclivities. The latter is a methodological sine qua non for assessing theological and other tendencies among the manuscripts, as Robert Marcello articulates in his chapter, \u201cMyths About Orthodox Corruption.\u201d
\nComplete collations are not only necessary for individual manuscripts; they are also necessary for each New Testament book. Since Hoskier\u2019s work on Revelation some ninety years ago, exhaustive collations have been produced for only two other New Testament books. Tommy Wasserman published his doctoral thesis on the text of Jude in 2006, and Matthew Solomon completed his dissertation on the text of Philemon in 2014. Solomon summarizes his findings in the ninth chapter of Myths and Mistakes. Among other observations, he reminds readers that the NA apparatus displays a small fraction of the textual variation in the manuscripts.
\nCollations of individual documents, when coupled with those of known manuscripts, can reveal something of the rich tapestry of textual history seen in each codex. The textual relations often hint at generations of mixture and influence, opening up intriguing questions on the document\u2019s transmission history. One of the priorities in collations is work on newly discovered codices. CSNTM posts the images of many such manuscripts, often before they are given a Gregory-Aland number. Graduate students interested in doing original research in the New Testament are encouraged to collate these documents. A recent collation of one of these discoveries, a tenth-century Gospels text, was even used by the editors of the Tyndale House Greek New Testament, published in 2017.<\/p>\n
COMPLIMENTS AND CAVEATS<\/p>\n
I have touched on just a few highlights in Myths and Mistakes. There is much, much more here than this bird\u2019s-eye view can display. \u039a\u1fe6\u03b4\u03bf\u03c2 to Hixson for conceiving this work, to both Hixson and Gurry for selecting the contributors, and to all for their unstinting devotion both to this arcane discipline and especially to \u201cthe faith that was once for all entrusted to the saints\u201d (Jude 3 NET). The takeaways at the end of each chapter summarize well its relevance for apologetics and anchoring the Christian faith in the text.
\nYet the authors do not advance a lock-step apologetic. No doubt, there are several points in this volume that any careful reader will take issue with. More than that, I am sure that not one of the authors will completely agree with all the others. That is part of the book\u2019s strength. The pursuit of truth holds greater capital than unity in presentation. The very nature of such a compilation models what the editors intend for the readers to grasp: we may not have an absolutely pure text, nor can we have certainty about everything we do have, but \u201ceven the most textually corrupted of our manuscripts and editions still convey the central truths of the Christian faith with clarity and power.\u201d
\nAs Michael Holmes has articulated and Zachary Cole attested, the New Testament manuscripts exhibit a text that is overall in excellent shape, but certainly not in impeccable shape; it manifests \u201cmicrolevel fluidity and macrolevel stability.\u201d What the authors of Myths and Mistakes insist on is that it is neither necessary nor even possible to demonstrate that we can recover the exact wording of the New Testament. But what we have is good enough.<\/p>\n
ACKNOWLEDGMENTS<\/p>\n
This book started life as an idea during our PhD programs. We love the Bible and are fascinated by how it came to be, especially at the level of its textual history. But as we progressed in our studies, we began to see a troubling trend among others who also loved the Bible and wanted to explain how it came to be. What we saw repeatedly were statistics, facts, and arguments meant to bolster confidence in the Bible that were actually having the opposite effect because they were misinformed, misapplied, or misstated. From that experience, Elijah had the idea of putting together a book to help reverse the trend.
\nOne thing became clear: someone needed to produce a good resource to correct these errors and provide updated information. Such a task, we quickly realized, was too complex for a single person to be able to handle all of the issues well, and it would be too important to settle for less. We resolved to produce such a book, and we decided that a team effort was the only way to approach the task.
\nWe will say much more about our goal in the following pages, but here we simply want to say thanks to the many others who helped us along the way.
\nIn the first case, our editors at InterVarsity Press have been encouraging from day one\u2014and that despite some hurdles presented by our approach to the book. A special thanks to all who attended and gave us feedback when we presented a preview of the book at the Evangelical Theological Society annual meeting in Rhode Island in 2017, including two distinguished guests from Germany, Holger Strutwolf and Klaus Wachtel. We were especially helped by the feedback from our esteemed panel, which included Peter J. Williams, Michael J. Kruger, Charles E. Hill, Peter M. Head, Timothy Paul Jones, and Daniel B. Wallace. These latter two deserve special thanks for being some of the first to see value in such a book as this and for going the extra mile in helping two greenhorns navigate the wild world of publishing. Other people who encouraged the idea deserve a mention as well, including Amy Anderson, Jeff Cate, Jeffrey D. Miller, and Tawa Anderson. Still others\u2014too many to mention by name\u2014gave of their time and expertise to read individual chapters and offer suggestions. For that, the contributors and we are grateful. Naturally, none of those scholars are responsible for anything they dislike about the resulting book.
\nLast and most important, we must mention our wonderful wives, whose patience, steadfastness, and joy has come in measure equal to their husband\u2019s long hours, eccentricities, and occasional discouragement. For them we are grateful well beyond any words we could write here.<\/p>\n
CHAPTER ONE<\/p>\n
INTRODUCTION<\/p>\n
Peter J. Gurry and Elijah Hixson<\/p>\n
WHY THIS BOOK?<\/p>\n
Perhaps, like us, you\u2019ve had this experience when driving to a new place. You set off, confident that your map or GPS has you headed in the right direction, and you begin thinking about other things. Soon, however, the roads are all the wrong names, and the signs do not seem right. Slowly, you begin to discover that you are lost. But where did you go wrong? Was it the last turn or the turn before that? Was it because you were on the phone, or are the directions wrong? If you\u2019re lucky, you manage to answer these questions, get back on the right track, and find your destination. This experience of thinking you know where you are going, only to realize you\u2019re lost, can be disorienting and frustrating. It can leave you wondering what else you may be wrong about. Are you sure you turned off the stove? Was the back door locked, or did you leave it cracked again? One doubt easily leads to another.
\nThe problem of getting to the right place by the wrong route is what we address in this book. Not about driving, of course, but about the Bible and about defending its credibility. Unfortunately, some defenders think they know how to get us to the proper destination when in fact they\u2019ve taken us through several wrong turns along the way. For those who discover that the route is wrong, the realization can be disorienting. Once-trusted guides can turn out not to be as reliable as once thought, and, in the case of defending the Bible, this can sadly lead to greater doubt in Scripture\u2019s reliability.
\nChristians believe and trust the Bible as God\u2019s special revelation. That belief is basic to the Christian faith. So, naturally, serious challenges to the trustworthiness of this book are significant and need a response. One challenge to the Bible that has risen to new prominence is the claim that we can\u2019t trust the New Testament because we do not even know what it says. This, we are told, is the case because the manuscripts\u2014handwritten copies of the New Testament\u2014are so corrupt from miscopying that we simply cannot know what the original text was. As Bart Ehrman, the scholar whose bestselling book Misquoting Jesus has done more than any other to bring this issue to the forefront, has said, \u201cHow does it help us to say that the Bible is the inerrant Word of God if in fact we do not have the words that God inerrantly inspired, but only the words copied by scribes\u2014sometimes correctly but sometimes (many times!) incorrectly?\u201d For Ehrman, the answer is clear: it is not much help at all, a conclusion that contributed to his much-publicized loss of evangelical faith.
\nAs Ehrman\u2019s public profile has risen, this part of his argument has gained greater traction, often without the benefit of his years of research in the subject area. Just before Christmas in 2014, for example, Newsweek published a long-form essay by Kurt Eichenwald titled \u201cThe Bible: So Misunderstood It\u2019s a Sin.\u201d Among a series of provocative claims, Eichenwald tells us, \u201cNo television preacher has ever read the Bible. Neither has any evangelical politician. Neither has the pope. Neither have I. And neither have you. At best, we\u2019ve all read a bad translation\u2014a translation of translations of translations of hand-copied copies of copies of copies of copies, and on and on, hundreds of times.\u201d
\nThis notion that the New Testament has been miscopied to the point of near oblivion has reached beyond national news magazines to capture certain parts of the popular imagination. Sometimes it crops up in unexpected places, such as popular fiction. In the bestselling Jack Reacher series written by Lee Child, we find an unexpected presentation of the idea that the original wording of the New Testament is hopelessly lost. In one of his stories, Child presents us with an Anglican priest who meets the protagonist on his way to Yuma, Arizona. On the drive there, the priest offers this lesson on the book of Revelation:<\/p>\n
Most of the original is lost, of course. It was written in ancient Hebrew or Aramaic, and copied by hand many times, and then translated into Koine Greek, and copied by hand many times, and then translated into Latin, and copied by hand many times, and then translated into Elizabethan English and printed, with opportunities for error and confusion at every single stage. Now it reads like a bad acid trip. I suspect it always did.<\/p>\n
There you have it. A trippy book made worse by thousands of years of miscopying and mistranslation so that now we do not even know what the original was. As anyone with a basic introduction to the New Testament knows, the problems here are obvious and plentiful. For starters, the book of Revelation was not translated into Greek for the simple reason that it was written in Greek. The many translations we do have of it\u2014both ancient and modern\u2014are almost all taken directly from Greek. It is true that opportunities for error do come from copying anything of length by hand, but these have also been accompanied by opportunities for correction and clarification. In short, our traveling priest\u2019s view of the matter is about as wrong as could be. The point here is not to pick on fiction (the appropriate genre for such misinformation, after all) but to show that views like these are all too easily consumed and accepted by popular audiences who lack the expertise to see through them. Indeed, it is not too much to say that the view expressed by the priest in this Jack Reacher novel is held by more and more people today.
\nTo be sure, most trained scholars ignore such popular nonsense and go about their work unfazed. Still, when these kinds of conspiratorial claims find their way to the New York Times bestseller list or the cover of Newsweek and Wired magazine, Christian scholars and apologists who care about Christianity\u2019s reputation take note. In their justified zeal to defend the Bible against such misinformation, they have naturally produced a growing number of books, articles, chapters, study Bibles, and blog posts in response. With such a proliferation of material, what justification could there be for yet another publication on the subject?
\nAs it turns out, that very proliferation has caused an unintended problem, and it is the one this book particularly addresses. A survey of literature reveals a growing gap between good scholarship on the transmission of the New Testament and its appropriation in the literature aimed at nonspecialists. In some cases, the misinformation is actually more severe on the side of those who want to defend the Bible\u2019s reliability (perhaps because they write more often on it). Such treatments often repeat bad or outdated arguments from other authors. In many cases, the treatment ends up worse than the ailment: arguments meant to encourage confidence in the Bible make it look untrustworthy through ignorance, negligence, or worse. This is troubling for those of us who love the Bible and want to know whether it can be trusted.
\nThe contributors to the present volume are convinced that the Bible should be loved and that its text can be trusted. Like many of those we critique in what follows, we are convinced that the New Testament text provides a more than adequate foundation on which to build the Christian faith. In that, we quite agree with them against Christianity\u2019s media-savvy critics. But we often find their reasons inadequate. From our own research, we know that studying the Bible\u2019s textual history can be intimidating. For the New Testament, it requires a knowledge of Greek and other ancient languages. It demands experience in reading ancient manuscripts. It draws on elements from classics, church history, and biblical studies. If that were not enough, some of the most important research is published in languages other than English. Those who write for popular audiences should not be faulted if they lack expertise in all these areas, and we certainly do not fault them here. However, the fact remains that many who address the topic from an apologetic angle construct their arguments from information that is at best outdated and at worst patently wrong.<\/p>\n
EXAMPLES OF THE PROBLEM<\/p>\n
Minor mistakes should be avoided, but misleading errors must be corrected because they discredit those who make them. At its worst, misinformed apologetics can have the opposite of their intended effect. Although the full story is surely more complicated, atheist Robert Price traces his rejection of Christianity back to this very issue. Despite becoming a Christian at eleven years old and engaging in fervent evangelism, devotional life, and church membership, Price writes, \u201cIronically, my doubts and questions were a direct outgrowth of this interest in apologetics.\u201d He continues,<\/p>\n
Obviously, at first I thought the arguments I was picking up from reading John Warwick Montgomery, F. F. Bruce, Josh McDowell, and others were pretty darn good! But once it became a matter of evaluating probabilistic arguments, weighing evidence, much of it impossible to verify, much of it ambiguous, I found it impossible to fall back on faith as I once had.<\/p>\n
That statement is sobering and serves as a warning against irresponsible apologetics. Price traces the beginning of his \u201cdeconversion\u201d to bad arguments presented by apologists. Granted, we do not think the question of the textual transmission of the New Testament leaves one\u2019s faith hanging in the balance. One could adopt almost any available text of the New Testament and still build a robust, orthodox Christian faith on it. Still, the Bible is worth defending, and that means it is worth defending well. Unfortunately, when it comes to the transmission of the New Testament, misinformation abounds. We can illustrate the problem with three examples.
\nOutdated information. The first example springs from a problem we all face: keeping up with the deluge of information. Thankfully, textual criticism is a field of study that regularly benefits from new manuscript discoveries. But this blessing becomes a curse for authors who have not kept their arguments updated. We can illustrate from some of our earliest material evidence. The papyri are those manuscripts made using papyrus, a reed plant that flourishes in the Nile River. We get our English word paper from this writing material. For the New Testament, the standard scholarly edition (the twenty-seventh edition of the Nestle-Aland Novum Testamentum graece, or NA) published in 1993 included all the papyrus manuscripts then known. These were numbered up to P98 (P = papyrus). Fast forward to 2012, and the newest edition (NA) lists papyri up through P127. That is almost thirty new papyri in fewer than twenty years. Moreover, these numbers are already out of date, because more papyri have been added to the official registry of New Testament manuscripts since then. In other words, our knowledge of manuscripts is constantly growing, and it can be hard to keep up. It is understandable when authors do not have the latest and greatest numbers. What is not as understandable and, in fact, a real problem is when the author does update the information, but only the part of it that favors the New Testament.
\nThis problem of selective updating has become common in one of the most widely used arguments to defend the New Testament. The argument involves a comparison between the number of New Testament manuscripts and the number of manuscripts for other ancient literature. One of the classic statements of it is found in F. F. Bruce\u2019s little book The New Testament Documents: Are They Reliable?, in which he tried to demonstrate the reliability of the New Testament using the same methods applied to other ancient documents. In one of his chapters, he directly addresses our concern here about whether the New Testament has been copied reliably. In that context, he shows that many other important works from antiquity lag far behind in comparison to the abundance and quality of the material we have for establishing the New Testament text. Our evidence is both earlier and more abundant. As a trained classicist himself, Bruce was calling skeptics to account for their double standard. As he put it, \u201cIf the New Testament were a collection of secular writings, their authenticity would generally be regarded as beyond all doubt.\u201d In other words, skeptics were not being consistent, and Bruce\u2019s comparison between the New Testament and classical works was meant to expose just how much this was the case. In this use, the argument has a long pedigree. One finds it used centuries ago by one of the great classicists, Richard Bentley (1662\u20131742). It is not surprising, therefore, that the comparison continues right down to the present. Today it is hard to find a book on apologetics that does not at some point make much of this basic comparison.
\nThe problem with the argument as used today is that the double standard has been reversed so that now it is defenders of the Bible who are guilty of being unfair to the classical literature. In many cases, the number of classical manuscripts is still taken from Bruce\u2019s sixty-year-old data, so that comparisons that were once accurate have become inaccurate and thereby misleading. There have been attempts to bring the comparison up to date, but these have gone unnoticed, and in any case they too are now in need of updating. Too often, authors simply take Bruce\u2019s numbers for the classical literature for granted even though they dare not do the same for the New Testament. One praiseworthy exception to this trend is Josh McDowell. In the most recent edition of Evidence That Demands a Verdict (coauthored with his son Sean McDowell), McDowell and McDowell cite both classicists and the Leuven Database of Ancient Books (LDAB) to give manuscript counts that are even more up to date than Clay Jones\u2019s 2012 article.
\nTo give an example of the problem, Bruce tells us that there are only a few papyrus fragments of Herodotus\u2019s famous Histories, dating nearly four hundred years after he wrote. Beyond that he reports only eight complete copies dating from nearly fourteen hundred years later. For someone familiar with New Testament manuscripts, that is sparse evidence indeed. Today, however, a few minutes with a modern manuscript database such as LDAB reveals forty-three manuscripts for Herodotus\u2019s Histories, one of which dates as early as the second century BC. For a more famous author such as Homer, the number of manuscripts swells to well over two thousand, the majority of which are papyri. These newer discoveries have not kept some from adopting Bruce\u2019s outdated numbers wholesale without any attempt at updating. The problem is made worse by the fact that these same authors do not adopt Bruce\u2019s numbers for the New Testament. Instead they rightly try to find more accurate numbers. In the case of Stanley Porter and Andrew Pitts, their zeal for the New Testament seems to have gotten the best of them, as their total is overstated by more than fifteen hundred manuscripts. How that happened is hard to say, but it is not hard to imagine how a fair-minded reader, to say nothing of a skeptical reader, might think that Christian authors have stacked the manuscript deck in favor of the New Testament. Regardless of intention, the result is lost credibility.
\nAbused statistics. A second problem in the debate is the widespread abuse of statistics. This is particularly unfortunate because, while statistics can never tell the whole story, they do offer much-needed perspective for nonspecialists. It is, for example, helpful to know how many manuscripts of the New Testament we have, how many differences there may be between them, and how many of the differences affect English translations. However, such information is helpful only if it is accurate and used responsibly. It is at both points that another problem confronts us. The numbers cited are frequently wrong, abused, or both.
\nThe most commonly cited statistic is the number of textual variants. This statistic has been referenced by scholars for over a century, but the number became a staple in the popular literature after Ehrman began claiming that there may be as many as 400,000 variants. Ehrman was one of the first and by far the most prominent scholar to suggest a number that high. Since then, the proposed number has risen even higher, with one prominent scholar venturing that it may be as high as 750,000. This is quite a shocking number, and it becomes even more so when compared to the number of words in the typical Greek New Testament, which is just over 138,000. As Bart Ehrman is fond of pointing out, this means that \u201cthere are more variations among our manuscripts than there are words in the New Testament.\u201d The impression is clear: the original New Testament has been lost in a sea of variants. Little wonder, then, that Christians have tried to address this disturbing statistic. Unfortunately, they have not always done a good job.
\nWhen we look closely, we find a myriad of problems that accompany all estimates about the number of variants. To begin with, until very recently, no one had bothered to give a reliable account of how they arrived at their number. Estimates were most often rehashed, sometimes accompanied by impressive-sounding but empty phrases such as \u201csome scholars say\u201d or \u201cthe best estimates are.\u201d What exactly makes one estimate reliable and another unreliable? No one seems to know. Consequently, it is rare that authors explain what it is they count in their estimated number of variants. Do they count spelling differences? Do they count cases in which the scribe has made an obvious mistake and produced something meaningless? Is the estimate meant to include ancient translations and scriptural citations from earlier Christian writers, or is it limited to Greek manuscript evidence alone? These are basic questions one should ask whenever seeing numbers such as 400,000 or 750,000 variants brandished in arguments about the New Testament text.
\nRecently, however, these problems were addressed in a detailed study with a clear method and open data. The study concluded that there are probably about half a million nonspelling differences among our Greek New Testament manuscripts. Every qualification in that sentence is important. This estimate does not include variants from non-Greek manuscripts or from the quotations of early church writers. Nor does it include what must be a vast quantity of trivial spelling differences. Finally, it estimates differences between manuscripts rather than differences from the original text, since that original is usually what is up for debate in the first place. (We should emphasize here that it is an estimate, not a count, since most of our manuscripts remain unstudied in detail.)
\nWith all the proper qualifications in place, what can this estimate actually tell us? At this point it becomes crucial to think about context. We have already seen Ehrman\u2019s statement that variants outnumber the actual words in the New Testament. This juxtaposition, which is numerically true, is offered to support his view that the text comes loaded with uncertainty. The problem is that the comparison itself is meaningless. It makes little sense to compare the number of supposed variants in all our Greek manuscripts to the number of words in only one manuscript or printed edition. That is a bit like comparing car accidents in the 1930s to car accidents today without factoring in the total number of cars on the road in each era; a crucial variable is left out. As it is, Ehrman\u2019s comparison tells us very little about how scribes did in copying the New Testament text.
\nUnfortunately, many apologists have followed Ehrman in his original comparison without thinking it through, and the results have only made the problem worse. Craig Blomberg, for example, tries to dull the force of Ehrman\u2019s estimate by claiming that all our variants are \u201cspread across more than twenty-five thousand manuscripts in Greek or other ancient languages,\u201d which, he says, results in a mere sixteen variants per manuscript. This claim is simply not true. As we have already noted, the half a million variants are only among our Greek manuscripts, of which there are closer to fifty-three hundred. Even if he were right, the comparison of \u201cvariants per manuscript\u201d is not helpful because our manuscripts are of such different sizes. Some are fragments no bigger than your hand, while others are complete copies of the New Testament. As before, the comparison is intrinsically a bad one, as it leaves out a key variable. You do get a number in the end, but it explains next to nothing about the real world. Despite this problem, the comparison continues to be made, and the results can approach the absurd, as, for example, when Stanley Porter\u2019s attempt at the comparison draws the conclusion that New Testament manuscript production \u201cnearly rivals that sometimes found today in modern print!\u201d Such a claim certainly grabs attention, but it discredits the larger point.
\nAnother problematic way the number of variants has been addressed by apologists is by addressing how the variants are counted. Some have claimed that the number is not as big as it might seem because every textual variant is recounted for every manuscript in which it occurs. If one manuscript has \u201cPeter said \u2026\u201d and two thousand others have \u201cSimon Peter said \u2026,\u201d then we have two thousand variants. If this were true, it certainly would be significant. This way of counting is found in the work of B. B. Warfield and continues down to the present. But, again, it is not right. In the example above, we should not count two thousand variants; we should count just two. The reason is that we are not counting the number of manuscripts that attest a variant; we are counting the number of variants attested by our manuscripts.
\nAlas, the problems do not end here. More recently, the number of variants has been further mistreated by writers making unfounded claims about where these variants occur in the New Testament. Craig Blomberg, in the same book cited above, says that the hundreds of thousands of textual variants \u201ccluster\u201d in \u201conly 6 percent of the New Testament.\u201d The implication is that they are less of a problem for Christians since they affect such a small portion of the New Testament. It is true, of course, that some passages lend themselves to variation more than others either because of their significance (such as the words of Jesus) or because of their form (lists are especially easy to mix up). However, one needs to spend only a few minutes with any good, critical edition of the Greek New Testament to see that variants occur throughout. They certainly do not \u201ccluster\u201d in anything like 6 percent of the text.
\nHow did such an erroneous claim get started? In Blomberg\u2019s case, the cause appears to be a misreading of his source\u2014a source, it turns out, that itself is unreliable at precisely this point. The original statistic is found in a book by Paul Wegner, who says nothing about variants \u201cclustering\u201d in a certain percentage of the text. Instead, he says that significant textual variants affect only about 6 percent of the text. But even this number is unreliable as given. In the only place where Wegner gives absolute figures for it, he says it is 7 percent\u2014and the numbers behind it are wrong! Referring to the well-known UBS fourth edition of the Greek New Testament, he says that it \u201cnotes variants regarding approximately 500 out of 6,900 words, or only about seven percent of the text.\u201d As a matter of fact, the UBS notes variants affecting over 1,500 words out of a total of 138,020, or only about 1 percent. In this case, the true numbers actually support Wegner\u2019s position better than his own numbers. But neither number supports Blomberg\u2019s claim about variants \u201cclustering\u201d in 6 percent of the text, whatever that might mean. The best we can say about this example is that it does offer a nice illustration of how texts become corrupted in transmission\u2014and how they can be corrected.
\nAll of this shows how easily statistical problems can spin out of control when left unchecked. In the popular literature, the results can be embarrassing. In the case in Greg Gilbert and his attempt to \u201cdebunk silly statements about the Bible,\u201d he not only repeats these errors from Blomberg and Wegner but also makes a series of \u201csilly statements\u201d of his own. These bad statistics might be easy enough to ignore if they happened occasionally. And they are certainly excusable in isolation. But when they occur repeatedly and without correction, they become a serious problem that needs to be addressed. In the worst cases, these confident claims about scribes, variants, and manuscripts discredit the very thing they are meant to support: our confidence in the biblical text.
\nSelective use of evidence. A final error is what we might bluntly call \u201cbelieving what we want to be true.\u201d One place this occurs is in the appeal to early manuscript evidence. In late 2011 Scott Carroll, at the time a director of the extensive Green Collection of biblical artifacts, tantalized the internet by announcing on Twitter that \u201cFor over 100 years the earliest known text of the New Testament has been the so-call[ed] John Rylands Papyrus. Not any more. Stay tuned.\u2026\u201d This news went mostly unnoticed until February 2012, when Dan Wallace said during a debate with Bart Ehrman that he could confirm the existence of a first-century fragment of Mark\u2019s Gospel, one dated by a renowned (but unnamed) paleographer. This was startling, not only as it would be our earliest New Testament manuscript but also because Mark\u2019s Gospel, while probably our earliest, has the poorest attestation when it comes to early manuscript evidence. Wallace went on to report that this new papyrus would be published the next year and suggested that, like our other known papyri, it too would reaffirm the text of the New Testament as we now know it. In time, it became clear that Carroll and Wallace were referring to the same manuscript. Ehrman raised the appropriate question of how much of an impact a first-century copy of Mark\u2019s Gospel would make on scholarship. The answer: maybe very little. As one might expect, however, the possibility of new information about early Christianity proved too much to resist, and speculation ran beyond the reach of good judgment.
\nThe yet-unpublished fragment had already begun to appear in published apologetics handbooks as evidence of the reliability of the New Testament. In the most recent update of the classic work Evidence That Demands a Verdict, Josh and Sean McDowell cite a lecture by Wallace for the following statement: \u201cA recently discovered portion of a Mark manuscript may date as early as c. AD 85\u2013125. However, we must wait for this discovery to go through peer review and publishing before we can establish the dating of this portion with more certainty.\u201d It is admirable that they are hesitant with regard to the date\u2014rightly so. However, the date of the fragment was not its only uncertain aspect. A genuine, early fragment of Mark\u2019s Gospel could possibly contain a text that differs in some significant aspect from other witnesses of Mark\u2019s Gospel, which could be seen as an argument against the reliability of the New Testament rather than for it.
\nThe early Mark fragment was finally published in 2018, and it is not quite so old as many hoped. To the surprise of many, it turned out not to be an item from a private collection but instead was part of the renowned Oxyrhynchus Papyri collection. The early Mark papyrus was edited by Dirk Obbink and Daniela Colomo as P.Oxy. 83.5345, and it was assigned the number P137. Obbink and Colomo admitted, \u201cDating [its] hand presents even more difficulties than usual,\u201d but they assigned it to the late second or early third century. The Egypt Exploration Society issued a statement reporting that Bernard Grenfell and Arthur Hunt had excavated it probably in 1903, according to the inventory number. While some were disappointed that P137 is not a first-century manuscript, it is still a significant discovery\u2014the oldest manuscript of Mark 1 and likely the oldest manuscript of any part of Mark.
\nWhat might be surprising to those outside the circle of textual criticism is that P137 is not the first manuscript of the New Testament claimed to be from the first century. Brent Nongbri points out Sch\u00f8yen manuscript 2630, a \u201cfirst-century Mark\u201d papyrus published in 1986 by Anton Fackelmann, son of the Austrian papyrus conservator of the same name. In the original publication, Fackelmann claimed to have deciphered notes used in the composition of Mark\u2019s Gospel from the undertext of a rewritten papyrus manuscript. No such undertext has been verified by scholars who have seen the fragment, and the visible writing\u2014the writing that would have been written after the notes Fackelmann described\u2014dates to the Ptolemaic period (second or third century BC). Thankfully, no apologists we know of have adopted Fackelmann\u2019s outlandish claims about Sch\u00f8yen manuscript 2630 as evidence of the reliability of the New Testament. The same cannot be said of another alleged \u201cfirst-century Mark,\u201d however.<\/p>\n
Figure 1.1. The first-century date assigned to P.Oxy. 83.5345 (P137) before its publication was widely touted by apologists as evidence for the reliability of Mark\u2019s account of Jesus; when published, however, it was dated to the second or third century<\/p>\n
In 1972, Jos\u00e9 O\u2019Callaghan, a papyrologist at the Pontifical Biblical Institute in Rome, argued that a manuscript containing Mark 6:52\u201353 was among the papyrus fragments of the Dead Sea Scrolls found in Cave 7 of Qumran. The caves of Qumran were abandoned in AD 68, which means that any manuscripts found in them must have been written before that date, including the fragment in question (designated \u201c7Q5\u201d for Cave 7, Qumran, fragment 5). In the case of 7Q5, the date of this tiny fragment is not what is disputed; the debate is whether it is actually a fragment of Mark\u2019s Gospel.<\/p>\n
Figure 1.2. 7Q5 is a small Greek fragment among the Dead Sea Scrolls at Qumran that some still reference as a first-century copy of Mark\u2019s Gospel, although scholars today reject this identification<\/p>\n
The fragment itself is tiny, \u201csmaller than two standard U.S. postage stamps,\u201d as Wallace describes it. The visible text on the papyrus consists of just a handful of letters (see fig. 1.2). The only complete, legible word is \u03ba\u03b1\u03af (kai, \u201cand\u201d), which is not a helpful word for identifying the text given how common it is. Some partial traces of letters are visible, but what ink is left of them is consistent with more than one letter. It was largely on the basis of these uncertain letters that O\u2019Callaghan identified the fragment as Mark 6:52\u201353. In addition to his reliance on uncertain letters, O\u2019Callaghan\u2019s identification requires two unique changes to the text of Mark in order to account for the few letters of 7Q5 that are visible. These changes would make 7Q5 the only manuscript in existence that leaves out the three-word phrase \u1f10\u03c0\u1f76 \u03c4\u1f74\u03bd \u03b3\u1fc6\u03bd (epi t\u0113n g\u0113n, \u201cto the land\u201d) in Mark 6:53 and contains a possible but rare misspelling of another word. What makes it highly unlikely is that only two letters of the misspelled word are visible, and one of them constitutes part of the misspelling. Almost immediately, Gordon Fee, among others, raised problems for O\u2019Callaghan\u2019s identification of 7Q5 as Mark\u2019s Gospel. Carsten Peter Thiede, on the other hand, continued to defend O\u2019Callaghan\u2019s position on 7Q5, despite its rejection by most scholars. In a scathing review of Thiede\u2019s book, J. K. Elliott makes an important observation about the temptation to follow the evidence that best suits our desired conclusions:<\/p>\n
Many of those who accepted O\u2019Callaghan\u2019s position did so because they were encouraged to date the canonical writings as early as possible on the basis of the specious argument that this would safeguard the veracity of their contents. More than 50 articles were published in the 4 years following the article in Biblica [the academic journal in which O\u2019Callaghan\u2019s original article appeared]. Most popular writers agreed with O\u2019Callaghan; most scholarly articles rejected his conclusions.<\/p>\n
Today, the wider academic community is in almost unanimous agreement that 7Q5 is not a copy of Mark\u2019s Gospel. Recently, Hans F\u00f6rster published a thorough summary and critique of the position, showing how speculative the argument is. Dead Sea Scrolls scholar Timothy H. Lim also rejects O\u2019Callaghan\u2019s claims, adding that others have identified the text of the manuscripts in question as 1 Enoch. Lim concludes, \u201cPrudence should guide one to be cautious in reading so much significance out of so little evidence.\u201d
\nUnfortunately, the identification of 7Q5 is not the only one of O\u2019Callaghan\u2019s and Thiede\u2019s claims that was received by popular audiences despite its rejection by the academic community. O\u2019Callaghan also identified several other fragments from Qumran as New Testament texts. Among the scraps of papyrus he also claimed to find manuscripts of Mark 4; James 1; and 1 Timothy 3\u20134. Additionally, he argued that Acts 27; Mark 12; Romans 5; and 2 Peter 1 were probably attested.
\nIn each case, the tiny fragment in question has only a handful of letters, and one is often required to allow multiple rare or even unprecedented mistakes and changes to the text to \u201cmake it work\u201d as a fragment of the New Testament. The unusual errors and changes that O\u2019Callaghan had to propose to account for all the letters on the fragments pose their own problem for Christian apologists tempted to follow him. If O\u2019Callaghan were correct about the fragments in question, it is true that we would have very early New Testament manuscripts, but they would have a text that is unlike anything we have in any of our other copies. Ironically, then, O\u2019Callaghan\u2019s view gives him remarkably early manuscripts, but they come at the cost of the text\u2019s earliest reliability.
\nDespite the textual problems these manuscripts would create if they were early witnesses to the New Testament, and despite an academic consensus rejecting O\u2019Callaghan\u2019s claims, some Christian apologists continue to cite them. Amy Orr-Ewing, for example, appeals to no fewer than nine of the Cave 7 fragments identified by O\u2019Callaghan in her book on the reliability of the Bible. She does note the presence of debate but still favors the conclusion that supports her larger argument:<\/p>\n
Critics have argued that these are not fragments of New Testament manuscripts but are instead writings produced by the Qumran community which sound similar to New Testament passages. The fact that the fragments are so small makes it extremely difficult to be certain either way. However, if, as seems highly possible, these fragments are pieces of the New Testament, these discoveries are potentially hugely significant.<\/p>\n
The shift from what is \u201cextremely difficult to be certain\u201d about in one sentence to what is then deemed \u201chighly possible\u201d in the next is rather stunning here. To the honest reader, it looks as if a desire for evidence to confirm her own view has gotten the best of her.
\nMore problematic still is Brian Edwards\u2019s use of O\u2019Callaghan\u2019s work in his book on the same subject. Edwards appeals specifically to 7Q5 and 7Q4 (\u201c1 Timothy\u201d), but he dismisses those who object to their identification as New Testament texts as if their own lack of scientific expertise, their critical biases, and even spiritual defects have prevented them from seeing the truth: \u201cGenerally, it should be noted, the opposition came from New Testament critical scholars and theologians and not from the scientists.\u201d Not to put too fine a point on the matter, but Elliott\u2019s warning is again appropriate: \u201cPots should be particularly self-critical before accusing kettles of blackness.\u201d
\nThe controversy over 7Q5 was not the first time someone claimed to have discovered a first-century manuscript of the Gospels. Over a century earlier, Constantine Simonides published an edition of some first-century papyrus manuscripts of the New Testament he had \u201cdiscovered\u201d in Liverpool. Today Simonides is remembered as the most notorious manuscript forger in history. Not long after his \u201cdiscovery\u201d in Liverpool, it came to light that he had forged the papyri himself. Dan Wallace gives good advice regarding the use of new, untested discoveries as evidence: \u201cWhen the next sensational archaeological find is made, should not conservatives and liberals alike ask the question: Will we fairly examine the evidence, or will we hold the party line at all costs?\u201d
\nThese three cases show the extent of the problem. It is not just that the arguments discussed are wrong; it is that they are misleading. More than that, they are widespread. The result is that the Bible is discredited by the very people who think they are defending it. Their aim is certainly noble, but their knowledge of the subject is not. What celebrated classicist Richard Porson (1759\u20131808) said about the major Bible controversy of his own day applies here as well: \u201cHe, I apprehend, does the best service to truth, who hinders it from being supported by falsehood. To use a weak argument in behalf of a good cause, can only tend to infuse a suspicion of the cause itself into the minds of all who see the weakness of the argument.\u201d What we hope to do in the present volume is offer reliable arguments that can be used on behalf of what we heartily agree is a good and noble cause.<\/p>\n
RELIABLE ENOUGH FOR WHAT?<\/p>\n
If defending the Bible\u2019s textual integrity is a noble cause, it remains to offer a working definition of what we mean when we say it is reliable. Simply put, we believe the textual evidence we have is sufficient to reconstruct, in most cases, what the authors of the New Testament wrote. We cannot do this with equal certainty in every case, of course, and the following chapters will discuss places where doubt remains significant. Nor do we think that God has preserved the original text of the New Testament equally well at every point in history or at every place in the world. Some times and places have had better manuscripts, editions, or translations at their disposal than others. This is true today, and it was true in the past. Nevertheless, we do think that even the most textually corrupted of our manuscripts and editions still convey the central truths of the Christian faith with clarity and power. In every age, God has given his people a text that is more than reliable enough to know the saving work he has accomplished through Jesus Christ.
\nTo be sure, the incredible discoveries of the last several hundred years and the enormous labors of textual critics from past eras have done much to identify God\u2019s inspired words and to increase our confidence in their recovery. In textual criticism, as in so much else, we stand on the shoulders of giants. Perhaps it would be best, then, to let one of those giants clarify what we mean when we say the text of the New Testament is reliable. Johann Albrecht Bengel (1687\u20131752) was a student in Germany who became troubled by the variants he encountered in the New Testament. After a lifetime of study and the production of his own important edition of the Greek New Testament, he concluded that these variants shake no pillar of the Christian faith. He writes:<\/p>\n
God\u2019s testimony concerning his Son Jesus Christ is truly abundant and worthy of our respect: the main thrust of what God wants us to learn never hangs on one single particle or word. The faith of the saints, accordingly, rests on sure and true foundations. But in the same way that a grain of gold, no matter how small, is nonetheless gold, so the smallest portion of the word which comes from the mouth of God is divine.\u2026 For this reason, whoever holds in reverence whatever comes from the mouth of God will be bound, in consequence, to seek out the most accurate reading of the New Testament Scripture as well.<\/p>\n
We agree with Bengel both in his confidence in the text\u2019s ability to support our faith and in the need to continue studying that text to ferret out every \u201cgrain of gold\u201d where doubts remain. Our text is more than adequate for what we need, yet the nature of God\u2019s Word requires us to seek out its original form to the full extent of our God-given abilities.<\/p>\n
PREVIEW<\/p>\n
This book is our attempt to bridge the gap between critical scholarship and those who address a popular audience. We do not write primarily for other textual critics; they will know most of what is presented here. Nor is our primary concern to answer Christianity\u2019s critics. Although some of that naturally occurs in what follows, others have already done that job ably. Instead, we write primarily as a self-corrective to Christian speakers and writers. For them and for their audiences, we want to offer an up-to-date, responsible guide to understanding the remarkable history of the New Testament text.
\nUnfortunately, the task of compiling and presenting correct information is easier said than done. It would be nearly impossible for a single person to gain the required expertise in all the matters we wish to address in this book. For this reason, we have brought together a team of scholars who are willing to contribute to our cause. The contributors are all (relatively) early career academics and all actively involved in academic scholarship of the text of Christian documents. Most received academic training at a time when Bart Ehrman\u2019s work was making headlines and topping bestseller lists. Some of the contributors, in fact, were introduced to the discipline of textual criticism through his writing. Though our denominational affiliations and personal convictions differ, we all agree that the New Testament is crucial to the faith and practice of Christians and that our present Greek editions, imperfect as they are, are more than adequate for those needs.
\nWe have organized the book into three broad categories. The first part deals generally with manuscripts, the second with the process of copying, and the third with translation, citation, and canonization. To begin at the beginning, Timothy N. Mitchell explores the starting point of the textual tradition, considering what the autographs were and how long they might have lasted. From there, Jacob W. Peterson plumbs how many manuscripts we have, explaining the difficulties involved in obtaining an accurate count as well as the distribution of manuscripts by content and by age. James B. Prothro picks up on this thread by looking at the widely used comparative argument that pits the vast number of our New Testament witnesses against the supposed pittance of other ancient literature. He details carefully how not to make this comparison and how it can still be of value. Moving on to the question of manuscript dates, Elijah Hixson shows how New Testament manuscripts are assigned a date and warns against falsely accepting our preferred date. Having considered our earliest manuscripts, Gregory R. Lanier provides a fresh evaluation of our later manuscripts, showing why the common view that \u201clater is worse\u201d is a caricature that needs to be retired. The upshot is that the appeal to having so many manuscripts by Christian apologists is not entirely \u201can idle boast\u201d simply because \u201cthose numerous MSS are not utilized [by them] to restore the original text,\u201d as some have claimed.
\nTurning from the manuscripts themselves to those who copied them, Zachary J. Cole challenges the claim that our earliest scribes were some of the most careless and \u201cwild.\u201d Peter Malik also looks at the work of scribes, particularly what followed their completion of manuscripts. He gives a fascinating tour of how scribal corrections can reveal much more to us than simple mistakes. Panning much wider across the tradition, S. Matthew Solomon provides the rare perspective of one who has collated all the Greek manuscript data for a particular New Testament book, offering insights from his rich study of Philemon. Next, Peter J. Gurry asks about how many variants there are and how much they matter. He qualifies the popular claim that our New Testament text is \u201c99 percent certain\u201d by showing how some variants in that 1 percent really do affect Christian theology and practice. Robert D. Marcello then looks at one type of variant\u2014namely, theologically motivated changes or \u201corthodox corruptions,\u201d showing the validity of the category but raising doubts about how often such occurred.
\nThe final section explores issues beyond the New Testament\u2019s Greek form. This includes patristic citations, the canon, and translations ancient and modern. Many popular works address patristic citations of the New Testament, but few address the patristic theology that enlightens their approach to such citation. Andrew Blaski explores the way early Fathers thought about textual criticism, how they dealt with textual variants, and how we can (and can\u2019t) use the Fathers in our own textual criticism. John D. Meade answers why the contents of our codices do not always line up with what became the New Testament canon and warns against looking for a canon in the codex. Although many popular-level treatments mention the early translations of the Greek New Testament into other languages, few such discussions are written by someone who works with ancient manuscripts written in those languages. Jeremiah Coogan addresses early Latin, Syriac, and Coptic manuscripts as well as early approaches to translation to assess the value these manuscripts have for recovering the Greek text. Finally, Edgar Battad Ebojo draws from his experience as a trained textual critic and a Bible translator in the field to explain how modern translators work with the textual evidence to produce accurate, reliable translations for Christians around the world.<\/p>\n
WHAT THIS BOOK IS NOT<\/p>\n
We have explained what this book is, and we have explained why this book is, but we still need to say a few words about what this book is not. It is not meant as a general introduction to New Testament textual criticism. Other books already cover that ground adequately, and readers seeking such an introduction are encouraged to turn to them. This book is also limited to the New Testament. Although we believe these issues arise for the Old Testament as well, few of us are qualified to treat it with the attention it deserves. Indeed, we hope that this book might spur others to take up that task. This book is also not a defense of the New Testament\u2019s claims about Jesus, his resurrection, or the apostles\u2019 teaching. Rather, this book addresses the more fundamental question of whether we can know what those claims are. Finally, this book is not the last word on the subject. The information presented here will become outdated sooner or later, of that we can be sure. More manuscripts will be discovered, methods will be refined, and new editions will be published. That is all to the good of the discipline and, we expect, to the Christian faith as well.<\/p>\n
CONCLUSION<\/p>\n
It is our sincere hope that this book can help Christians who love the Bible but find the arguments floating around popular culture disturbing. To be sure, there are numerous textual problems that remain unsolved. There are cases where the Bible\u2019s text has been corrupted by friend and foe alike. These cases make textual criticism an important discipline for Christians to engage in. If we believe that God inspired the particular words of Holy Scripture, then it is incumbent on us to do our best to identify those words so that we can preach, teach, treasure, and obey them.
\nYet we also hope this book will be helpful to those who have been disturbed to find that not all the arguments they were given about the reliability of the New Testament text are themselves reliable. To them, we hope this book serves as an antidote to cynicism and skepticism. Finally, to our friends who engage in apologetic ministries, we want to express our appreciation for your important work. We hope this book will be a trusted guide and reliable resource as you go about your work. What Richard Porson aimed to do two hundred years ago we want to do again today: to offer a service to truth by keeping it from being supported by falsehood. With that aim in mind, we hope to clear away the myths and mistakes that have grown around the New Testament text and let it shine for all its worth.<\/p>\n
Key Takeaways
\n\u25ba Information about the reliability of the New Testament in apologetic handbooks is often outdated.
\n\u25ba Statistical arguments for the reliability of the New Testament are often unqualified or factually incorrect.
\n\u25ba Christians need to avoid the tendency to \u201cbelieve what we want to be true\u201d about the early manuscripts of the New Testament.
\n\u25ba We should be careful not to appeal to exaggerated or sensationalistic claims.<\/p>\n
CHAPTER TWO<\/p>\n
MYTHS ABOUT AUTOGRAPHS<\/p>\n
WHAT THEY WERE AND HOW LONG THEY MAY HAVE SURVIVED<\/p>\n
Timothy N. Mitchell<\/p>\n
In the midst of nearly every discussion of the textual reliability of the New Testament writings, references to the \u201cautographs,\u201d or more simply \u201coriginals,\u201d frequently surface. Noted New Testament scholar and author Bart Ehrman does just this in his bestselling book Misquoting Jesus: Behind Who Changed the Bible and Why, when he writes, \u201cRather than actually having the inspired words of the autographs (i.e., the originals) of the Bible, what we have are the error-ridden copies of the autographs.\u201d Ehrman emphasizes the lost \u201coriginals\u201d in response to evangelical doctrinal statements, which place God\u2019s act of inspiration on the autographs of the New Testament and not on any one manuscript or manuscript tradition.
\nThe Chicago Statement on Biblical Inerrancy is representative of these types of doctrinal statements. Formulated in 1978, the International Council on Biblical Inerrancy held a series of meetings in Chicago, which resulted in the formulation of the Chicago Statement on Biblical Inerrancy. A central proposition of this statement is found in article X: \u201cWe affirm that inspiration, strictly speaking, applies only to the autographic text of Scripture, which in the providence of God can be ascertained from available manuscripts with great accuracy. We further affirm that copies and translations of Scripture are the Word of God to the extent that they faithfully represent the original.\u201d
\nWith a desire to defend a high view of Scripture and in response to the criticisms of scholars such as Bart Ehrman, apologists and theologians often advance arguments advocating that the New Testament manuscript tradition faithfully represents the autographs. In his book Why Trust the Bible?, popular author Greg Gilbert writes, \u201cIt\u2019s more than a little likely that the originals, penned by the authors themselves, would have been preserved and used to make countless new copies over decades or even centuries before they were lost.\u201d Gilbert is not the first to make such claims. As he mentions earlier in the chapter, he draws mainly on the work of Craig L. Blomberg of Denver Seminary. While addressing Ehrman\u2019s arguments in Misquoting Jesus, Blomberg writes, \u201cSecond- and third-century New Testament manuscripts may well be copies of the very autographs, or at least copies of those copies.\u201d
\nA similar but much more detailed form of this argument can be found in Craig Evans\u2019s article \u201cHow Long Were Late Antique Books in Use?\u201d In this study Evans, a professor at Houston Baptist University, contends that because of \u201cthe probability that the autographs and first copies circulated and were in use for one century or longer, there really is no justification for supposing that the text of the NT writings underwent major changes in the first and second centuries.\u201d This argument is used prominently in the Faithlife documentary Fragments of Truth, hosted by Evans, which appeared in American theaters on April 24, 2018.
\nIn contrast to the optimism of Evans, Matthew Larsen of Yale University has recently argued against the concept of a finished autograph altogether. Instead, Larsen contends, \u201cWe can no longer simply assume that a text was finished and published, especially texts that are not high literature. Unless we can determine that a text was finished, closed and published, which would exclude many of the texts now called the New Testament, traditionally conceived modes of textual criticism may be a square peg for a round hole.\u201d
\nLarsen is arguing that the New Testament writings, more specifically the Gospels, were not the kind of writings that had originals. Thus, Larsen\u2019s thesis challenges the very foundation of Evans\u2019s optimistic assertions by questioning whether such a thing as a finished autograph existed for the New Testament writings. The conclusions of Evans and Larsen will be addressed throughout the discussion in this chapter, and a rejoinder will follow in the concluding comments.
\nWith all of this talk of autographs and originals, it is of first importance to define these terms. In order to define an autograph, it is paramount to understand the Greco-Roman milieu in which the New Testament writings were produced.<\/p>\n
GRECO-ROMAN PUBLICATION<\/p>\n
The term book, used in this chapter to describe the Greco-Roman medium for literature, refers to both the roll or bookroll (often referred to more popularly as a scroll) and the codex, the ancient predecessor to the modern book form. The bookroll was written entirely by hand in columns of continuous text on long sheets of parchment (prepared animal skin) or papyrus and then rolled up with the text on the inside. The codex, in contrast, was made from leaves of parchment or papyrus stacked together, folded down the center, and then stitched along the folded spine with the writing (also by hand) on both sides of the pages.<\/p>\n
Figure 2.1. P.Oxy. 18.2192 shows postscripts in the right column in two different hands<\/p>\n
Modern conceptions of publication, with its copyright laws, rules of plagiarism, the mass production of books by printing presses, and marketing through bookstores, did not exist in the Greco-Roman world. An adequate understanding of the transmission of the New Testament writings in the earliest stages must therefore incorporate the techniques of book production available to the authors of the New Testament and the early communities who copied, circulated, and read these books.
\nIn the Roman world of the first and second centuries, two main avenues were used by authors to produce and distribute books of literature. Though likely small in scale, there was a commercial industry of distribution and circulation composed of copyists and artisan craftsmen who sold works of literature in bookshops and city marketplaces. Social contacts and networks of friends and associates who circulated books often played a more pivotal role in the dissemination of literature.
\nThe remains of a letter found in the Roman city of Oxyrhynchus dating to the second century illustrate both the commercial and private circulation of books in striking detail. P.Oxy. 18.2192 is a fragmentary papyrus in which the main body of text no longer survives. All that remains are two postscripts written in a different hand from the text of the letter (see figure 2.1). The first postscript is written by the sender. It is in a different hand likely because the sender, though obviously literate, used a secretary to compose the letter. The first postscript reads, \u201cHave a copy made of books six and seven of Hypsicrates\u2019 Men Who Appear in Commodies and send it to me. Harpocration says that Pollio has them among his books, and probably others may have them too. And he also has prose epitomes of Thersagorus\u2019s Myths of Tragedy.\u201d
\nThe other postscript is in a different hand and was made by the recipient of the letter, who then dispatched it back to the original sender in reply. It reads, \u201cDemetrius the bookseller has them, according to Harpocration. I have ordered Apollonides to send to me some of my books\u2014which ones you\u2019ll find out from him. And if you find any volumes of Seleucus\u2019s work on Tenses\/Metrics\/Rhythms that I don\u2019t own, have copies made and send them to me. Diodorus\u2019s circle also has some that I don\u2019t own.\u201d This fragmentary papyrus gives a glimpse into the circulation and copying of books among social networks. Here two circles of scholarly acquaintances are exchanging and borrowing books in order to procure copies of them. A bookseller is briefly mentioned, but the circulation of literature occurs mainly by appropriating through acquaintances and making personal copies from these borrowed books. Notice that there is no mention of legal or monetary restraints preventing these scholars from simply transcribing a complete book in order to possess a copy.
\nBoth the commercial and private circulation of books can be seen in the letters of Pliny the Younger (ca. AD 61\u2013115), who was governor of Bithynia, a region that is now part of Turkey. He wrote to the famous Roman historian Suetonius, urging him to publish his work. In this letter Pliny declares that he wants to hear that the \u201cbooks of my dear Tranquillus are being copied out, sold, and read\u201d (Ep. 5.10). Here each avenue of distribution is alluded to. Copying appears to be a reference to the circulation of the work among Suetonius\u2019s associates. Selling must be an allusion to book shops. Reading is likely a reference to the reading out of the work in a communal setting.
\nAlong with mentioning \u201cpublication\u201d more generally, Pliny gives one of the most detailed descriptions from the Roman era of the editing and releasing process that a work of literature underwent.<\/p>\n
In the first place, I revise my composition in private, next I read it to two or three friends, and then give it to others to annotate; if I doubt the justness of their corrections, I carefully weigh them again with a friend or two. Last of all I recite the piece to a numerous assembly, and this is the time, if you can believe me, when I exercise the most rigid criticism; for my attention rises in proportion to my solicitude. (Ep. 7.17)<\/p>\n
Pliny released draft versions to associates and friends in phases of editing, sharing the edited piece with friends, rewriting in light of the feedback, and reading out a final draft to a larger group of associates for comment. During these rewriting stages Pliny assumed that the work in progress would not be circulated beyond this immediate group of friends. Strict control over the draft version did not always occur, and unfinished writings would sometimes circulate before the author desired. Pliny again illustrates this in his letter to Octavius, warning him that portions of his works began to circulate without Octavius\u2019s consent (Ep. 2.10).
\nDuring the draft stages, Pliny often dictated his compositions and notes to a scribe. He wrote to his friend Fuscus that he would close the windows in his Tuscan villa and work on an initial draft, after which, he wrote, \u201cI call my secretary, and, opening my shutters, I dictate to him what I have composed, after which I dismiss him for a little while, and then call him in again and again dismiss him\u201d (Ep. 9.36).
\nOnce Pliny was satisfied with the piece, he then \u201creleased\u201d the work to be circulated by his associates and acquaintances, either by sending a copy to a dedicatee of the work, giving or lending a copy to a friend upon request, sending a copy to a bookseller, or depositing the book in a library. At this point of relinquishing control of the piece, Pliny signaled that the work was complete and ready to be copied and circulated.
\nWriting a couple of decades earlier than Pliny, the epigrammatist Martial vividly describes Roman imperial life in the last decades of the first century and often alludes to the publication of his work throughout his epigrams. Martial gives insight into the problem of plagiarizing at the end of the first century. After accusing an anonymous person of plagiarizing his work, Martial warns them, \u201cYou must look for private, unpublished work,\u201d and this is because \u201ca well-known book cannot change author,\u201d and then concludes by declaring that \u201cwhoever recites other men\u2019s productions and seeks fame thereby, ought to buy\u2014not a book, but silence\u201d (Epigr. 1.66).
\nNotice here what Martial is declaring: that once a work had been circulated in a community, that writing could not be significantly transformed, either by plagiarizing or altering its text, without these actions becoming widely known. Therefore, an aspiring author who wished to publish another author\u2019s work as their own would have to do so from writings not yet in circulation.
\nWriting at about the same time as Martial, prominent rhetorician Quintilian began to compile his magnum opus, Institutes of Oratory, which was a \u201clengthy dissertation on the finer points of raising a gentleman to the art.\u201d In the preface to book one, while dedicating the work to his friend Marcellus Victorius, Quintilian indicates that some of the material in his Institutes might be found in an unedited form circulating under his name:<\/p>\n
Two books on the Art of Rhetoric are already circulating in my name, though they were never published by me nor prepared for this purpose. One is a two days\u2019 lecture course which was taken down by the slaves to whom the responsibility was given. The other lecture course, which spread over several days, was taken down by shorthand (as best they could) by some excellent young men who were nevertheless too fond of me, and therefore rashly honored it with publication and wide circulation. (Inst. Or. 1.prologue.7\u20138)<\/p>\n
Quintilian learned from his circle of acquaintances that his lectures had been copied and were circulating in an unedited form. He knew (or assumed) that Marcellus had read these transcribed lectures and was careful to indicate that the Institutes was a new work that supplanted these crude notes.
\nSeveral of his speeches had also been copied down by scribes who were expert in shorthand. These scribes made many blunders during the copying process and circulated these poorly transcribed copies for a profit (Inst. Or. 7.2.24). The circumstances in which his speeches were copied were different; they were specifically transcribed in order to be sold for a profit. Though Quintilian disliked these crudely copied lectures circulating under his name, he was powerless to control the dissemination and copying of them.
\nAnother example comes from physician Galen of Pergamum, who flourished in the last half of the second century, practicing medicine for Rome\u2019s social elite and authoring hundreds of treatises, of which around 170 separate works survive. Because many of these treatises circulated haphazardly and without Galen\u2019s consent, he wrote two works discussing the composition and publication of his many writings, De libris propriis (On My Own Works) and De ordine librorum suorum (On the Order of My Own Books). In De libris propriis, Galen provides a unique glimpse into the circulation practices of Greco-Roman society when he complains, \u201cMy books have been subject to all sorts of mutilations, whereby people in different countries publish different texts under their own names, with all sorts of cuts, additions, and alterations\u201d (De libr. propr. 19.9).
\nGalen writes that the corruptions of his books occurred when the pupils died and the works were stolen and began to circulate under the names of these plagiarizers. Fortunately, Galen writes, \u201cAll these were eventually caught, and many of those who then recovered the works affixed my name to them. They then discovered discrepancies between these and copies in the possession of other individuals, and so sent them to me with the request that I correct them\u201d (De libr. propr. 19.10). Apparently, many of Galen\u2019s compositions were not meant to circulate as formal works of literature with his name affixed. Instead they were merely Galen\u2019s unedited lecture notes. Once these transcripts were given to his students, however, he lost control over their fate, and these writings were then misappropriated and altered by his students\u2019 successors.
\nGalen gives a vivid account of a public presentation he facilitated concerning several ancient medical writers. The presentation escalated into a debate between Galen and a follower of Martialius (a contemporary physician). Galen\u2019s critical response to the follower of Martialius resulted in admiration from the crowd. As a consequence, Galen writes, a friend begged him to dictate his speech to a scribe trained in shorthand. Galen agreed, and after returning to Rome years later, he discovered that it \u201cwas now in the possession of a large number of people\u201d (De libr. propr. 19.14\u201315). Galen was complaining because his lecture notes had circulated and were read more broadly than he had intended. As a result, he vowed never to give a public demonstration for fear that his words would be transcribed and misused (De libr. propr. 19.15). He was completely at the mercy of this community to obey his instructions not to circulate a particular writing (De libr. propr. 19.42\u201343).
\nThe sources above reveal that publication, or more accurately, the releasing and circulation of books, occurred only after the author edited and rewrote several times over. It was often a community effort that involved some of the author\u2019s closest associates, who gave constructive criticism, suggested changes, and at times used the services of a scribe or secretary to copy down dictation. Once the document was complete, the author signaled to associates that the piece was finished by releasing the work to be copied and circulated through networks of friends and acquaintances, sending the work to a dedicatee, reciting the piece to the community, or by sending a copy to a bookseller or library. Therefore, the concept of \u201cauthor\u201d can include more than just a single person and more likely included scribes, friends, and others who contributed to the composition. Though these authors used their associates, literate slaves, and scribes to aid in the editing process, the writing was still considered to be the author\u2019s own work, a product of his or her creative mind.
\nIn contrast to this view, Matthew Larsen contends that writings such as Galen\u2019s accidentally published lecture notes would be better described as a \u201c \u2018living\u2019 textual tradition, rather than as a final and fixed book with clean lines of progression that lead back to a pure originary moment.\u201d Thus, it is paramount that this \u201coriginary moment,\u201d or rather, \u201cautograph,\u201d be properly understood and defined.<\/p>\n
A DEFINITION OF AUTOGRAPH<\/p>\n
Doctrinal statements, apologists, and theologians are typically careful to distinguish between the authorial copies of the New Testament books and the subsequent various textual forms and alterations introduced throughout their transmission history. On the other hand, these same theologians and doctrinal statements are not always as clear in distinguishing between the sources used by the New Testament authors, the various stages of composition, and the final form these writings took before being released for dispatching to their intended recipients (the Epistles), or for circulation and copying (Gospels, Acts, Revelation).
\nThe word autograph means simply \u201csomething written or made with one\u2019s own hand.\u201d Therefore, strictly speaking, included in this definition are any possible early versions and rewritings of a composition while under the control of the author. These draft copies would presumably be in a state of textual fluidity because the author is in the process of composing and editing the document.
\nMultiple autographs of a petition to Egyptian prefect Publius Ostorius Scapula (ca. AD 3\u201310\/11) illustrate the textual instability of an autographic text. Though the topic at hand concerns literary compositions, P.Mich.inv. 1436 and P.Mich.inv. 1440 provide a rare glimpse of multiple draft copies of the same work (see figs 2.2 and 2.3). Although both papyri were written by the same person, inv. 1436 contains several alterations that favor its identification as the first draft of inv. 1440. Though the text of both papyri is fragmented, lines 2\u201310 of inv. 1436 are repeated in lines 11\u201317 of inv. 1440. The scribe revised the text of inv. 1436 above lines 6 and 8, and marked line 9 for deletion; despite this, these alterations were not integrated into the text of inv. 1440.<\/p>\n
Figure 2.2. P.Mich.inv. 1436 showing authorial changes<\/p>\n
These two papyri reveal that at the level of \u201cautograph\u201d the author\u2019s text can be fluid. The difficulty in this case is that it is impossible to know the completed form of the petition with certainty. This uncertainty would change, of course, if a copy of this petition were known to have been dispatched and was received by the prefect Scapula. Even if a copy were inadvertently released before the author was satisfied with its form, the composition would still be effectively completed since then the author would lose any control over the fate of the document at that point.
\nWith regard to Galen\u2019s lecture notes, once they were given over to his students, he lost control over their fate. He could no longer affect the textual form of these writings in the same way without the consent of the students who possessed them. The same is true of the writings of Quintilian; once his students transcribed his lectures and released them for circulation, Quintilian could no longer control the shape of the text as he could before. In both instances, Galen and Quintilian relied on the participation of the community. With Galen\u2019s writings, his followers and students informed him of the altered and plagiarized writings, then retrieved and returned them so Galen could edit them (De libr. propr. 19.10). Quintilian required the cooperation of Victorius, the dedicatee of his Institutes of Oratory, to disregard the other crudely transcribed lectures that were in circulation (Inst. Or. 1.prologue.7\u20138).
\nThis loss of control is important because, in contrast to the views of Larsen, the point at which a document was released beyond the immediate control of the author, whether accidentally or intentionally, effectively ended the draft and rewriting stages of the document. This can be seen in the warnings of Pliny to Octavius that his writings were already circulating against his desire (Ep. 2.10), in the frustrations of Galen that his writings were widely circulating to unwanted readers (De libr. propr. 19.14\u201315), in Quintilian\u2019s explanations of multiple versions of his lectures (Inst. Or. 1.prologue.7\u20138), and in the accusations of Martial to the plagiarizer of his epigrams (Epigr. 1.66). The text was fluid in some respects, as Larsen contends, but the authors (in this case Martial, Quintilian, Galen, and Octavius) clearly distinguished between these altered texts and their initially released versions. Even the followers and students of Galen made distinctions between the composing of these documents by Galen and the alterations and mutilations made to them after their release and circulation (De libr. propr. 19.14\u201315).<\/p>\n
Figure 2.3. P.Mich.inv. 1440 showing that authorial changes in P.Mich.inv. 1436 were not incorporated into the text of this copy of the petition<\/p>\n
In light of this, the New Testament writings can be said to be \u201ccompleted\u201d once they were released by the authors and began to circulate as works of literature. These documents were no longer under the control of their authors and would have circulated as distinct writings. Therefore, in reference to the New Testament, the \u201cautograph,\u201d as it is discussed by apologists, theologians, and doctrinal statements, is best defined as the completed authorial work that was released by the author for circulation and copying, and this can and should be distinguished from earlier draft versions or layers of composition.<\/p>\n
AUTHOR\u2019S COPIES<\/p>\n
Once a New Testament writing was completed and released for circulation, there was a possibility that further copies would be made under the control of the author or authors. Strictly speaking, these copies would also be considered \u201cautographs,\u201d for they would be produced by the hand of the author, or at least under the author\u2019s direction. The practice of retaining a copy of a writing once published is sometimes referenced by Greco-Roman authors. Though multiple copies of a work may have been produced by the author before or after circulation, as will be seen below, this was not always the case.
\nThe fourfold Gospel. Once completed, a duplicate of a composition might be retained in the personal collection of the author or the author\u2019s community. With regard to the Gospels Matthew, Mark, Luke, and John, this is one possibility. The Shepherd of Hermas, likely composed in Rome sometime in the early to mid-second century, may help to illustrate the circumstances surrounding the publication of each of the four Gospels and perhaps Revelation. At one point the narrative features an \u201celderly woman\u201d (who represents the church) instructing Hermas to distribute copies of the message contained in a \u201clittle book\u201d he had been given through very specific means:<\/p>\n
Therefore you will write two little books, and you will send one to Clement and one to Grapte. Then Clement will send it to the cities abroad, because that is his job. But Grapte will instruct the widows and orphans. But you yourself will read it to this city, along with the elders who preside over the Church\u201d (Shepherd of Hermas Vision 2.4).<\/p>\n
In this reference, three copies of the \u201clittle book\u201d were produced by Hermas. A copy was dispatched to Clement, who then made duplicates to be distributed to various cities. A second copy was sent to Grapte to use for instructing widows and orphans. A third copy was retained by Hermas in order to read communally in \u201cthis city\u201d (possibly a reference to Rome). Therefore, it is possible\u2014though we do not know\u2014that multiple copies of the Gospels were made by their authors in order to distribute them through the Christian communities.
\nIn some circumstances, however, an author may not have produced duplicates of a work. It was common practice in the Greco-Roman world for an author to dedicate a composition to a superior, a patron, or a friend. Galen shares a circumstance in which he composed several works dedicated to a friend, Boethus (De libr. propr. 19.16). This friend left Rome for Syria with these writings in his possession, apparently before Galen felt they were finalized. It was some time later, after Boethus\u2019s death, that Galen was able to reacquire these books and complete them.
\nIn another work, Galen laments the loss of an entire collection of his books, which were kept in storehouses in Rome. A fire on the Sacred Way destroyed several libraries, archives, and book collections along with Galen\u2019s own. He writes that he had made duplicates of some of his compositions, which he had sent to friends, but others were not so fortunate. Apparently, he had lost several of his own writings, of which he possessed only a single copy (De indolentia 20\u201322).
\nPaul\u2019s letters. In the time of Paul in the first century, it was common practice for a copy to be made of a letter before it was dispatched and retained in the sender\u2019s archives. From these archived copies, if the author was famous or influential, a collection of letters might be gathered, edited, and circulated as a literary unit. With regard to his collection of letters, Pliny the Younger writes to Septicus, \u201cYou have frequently pressed me to make a select collection of my Letters (if there be any which show some literary finish) and give them to the public. I have accordingly done so; not indeed in their proper order of time, for I was not compiling a history; but just as they presented themselves to my hands. Farewell\u201d (Ep. 1.1). The first nine books of letters, as Epistle 1.1 indicates, Pliny collected and circulated on his own initiative, whereas the final book of correspondence between Pliny and Emperor Trajan was likely published posthumously. Pliny illustrates how Paul\u2019s epistles could have been gathered together and circulated as a collection.
\nAlthough Evans writes, \u201cIn late antiquity, no one produced a single exemplar of a work and then circulated it,\u201d it is quite possible that Paul did not have copies of his more personal letters, such as those to Timothy and Titus. This occurred in the case of Cicero\u2019s (106\u201343 BC) personal correspondence with his longtime friend Atticus. Apparently these letters were collected from the family archives of Atticus and published around a hundred years after Cicero\u2019s death. Cicero\u2019s other letters had already been collected from Cicero\u2019s personal copies and published in books by his secretary Tiro. Thus, it cannot be said for certain that there were multiple autographic copies of every New Testament book produced by the authors.<\/p>\n
THE LONGEVITY OF THE AUTOGRAPHS<\/p>\n
Having considered the question of what the autographs were, we turn now to consider how long they may have been in use. George Houston, in his work Inside Roman Libraries, surveys book collections in antiquity and analyzes their contents, the date of composition, and the rough date of the discarding of the collection or the last known period of use. From these data Houston concludes that the useful life of papyrus bookrolls was on average 100 to 125 years and in extreme cases 300 to 500 years. Using Houston\u2019s research as a foundation, Craig Evans argues that the New Testament autographs were likely in use into the second and third centuries. From this, he suggests that the autographs probably had a controlling influence on the textual transmission of the New Testament. To properly evaluate this claim, we should consider both the longevity and the loss of ancient manuscripts.
\nLongevity of ancient manuscripts. In the ancient Roman city of Oxyrhynchus in Egypt (modern el-Behnesa), thousands of fragments of papyri were recovered in the late nineteenth and early twentieth centuries. Some of these excavations were headed up by Bernard Grenfell and Arthur Hunt, and the findings were published in the famous series the Oxyrhynchus Papyri. During their excavation of the trash mounds and in later discoveries made by others, a few concentrations of documents were revealed that probably represented several distinct collections. Though Grenfell and Hunt did not reveal how they ascertained the date these collections were discarded, the paleographical dating of the bookrolls gives a general period of time these books were used. This period is roughly from 75 to 250 years.
\nHouston also examined the remains of the extensive library found in the Villa of the Papyri in the Roman city of Herculaneum, which was destroyed during the eruption of Vesuvius in AD 79. Many papyrus bookrolls were preserved by carbonization in the intense heat. Judging by the paleographical date of when the books were copied and when they were destroyed by the eruption, a general useful date for these rolls ranges from 120 to 350 years. Taking the data from these two studies, Houston, as previously mentioned, calculated the average lifespan of the papyrus bookroll at around 100 to 125 years and in extreme cases 300 to 500 years.
\nEvidence from the Caesarean library tentatively supports Houston\u2019s findings. Living in Bethlehem in the later part of the fourth century, Jerome wrote that the successors of Eusebius of Caesarea labored to copy the library\u2019s aging and damaged papyrus bookrolls onto parchment codices. Jerome writes, \u201cEuzoius was educated as a young man at Caesarea along with Gregory, the bishop of Nazianzus, under the rhetor Thespesius and later became bishop of the same city; with very great toil he attempted to restore on parchment the library of Origen and Pamphilus that had been damaged\u201d (Vir. ill. 113). Some of these manuscripts that were restored in the middle of the fourth century, if they were from Origen\u2019s original collection, would have been nearly 150 to 200 years old. If some of these damaged bookrolls were from the hand of Pamphilus or from his collection, the bookrolls would have been 75 to 100 years old at the time of their replacement by Euzoius. At least in this instance, when a book collection was kept in a large library, the useful life of a papyrus bookroll lasted from 75 to 200 years. This is in general agreement with Houston\u2019s study.
\nLoss due to climate, persecution, and so on. Though Houston\u2019s findings reveal a long useful life for bookrolls, these averages likely represent best-case scenarios. Papyrus did not fare as well in the more humid environments of the rest of the Mediterranean world. Galen expressed some exasperation at his attempt to make personal copies of important books stored in the libraries on the Palatine in Rome. Galen writes, \u201cThese (books), then, did not cause me a small pain when copying them. As it is, the papyri are completely useless, not even able to be un-rolled because they have been glued together by decomposition, since the region is both marshy and low-lying, and, during the summer, it is stifling\u201d (De indolentia 19).
\nIt is apparent that papyrus bookrolls did not survive well in the humid environment of Rome. Even in the case of the Caesarean library mentioned above, these century-or-more-old books that were still in the library were damaged and in need of replacement. There is no telling how long these books were in this state of disrepair before Euzoius set about making replacements.
\nAlong with wear and tear, natural disasters often proved to be formidable to book collections and libraries. In AD 192, a terrible fire broke out in Rome, which destroyed many storehouses, libraries, and book collections. Galen incurred terrible losses to his collection of books and store of medicines. The eruption of Vesuvius in AD 79 destroyed both Pompeii and Herculaneum in Italy. As was discussed above, this disaster destroyed an extensive collection of books in the Villa of the Papyri in Herculaneum.
\nImperial persecution also played a role in the destruction and banning of books, and not just Christian books. The great poet Ovid, who flourished during the reign of Emperor Augustus, was exiled to Tomis on the Black Sea (modern Constanta, Romania), and his books were banned from imperial libraries. He even composed a poem that listed the libraries that banned his works (Tristia 3.1.65\u201372). In the beginning of the first century AD, Titus Labienus\u2019s books were burned by a senatorial decree, and the instigator of this decree eventually had his own books burned as well. This practice of book burning culminated in the edict of Diocletian in AD 303 \u201cordering the confiscation and burning of Christian books.\u201d
\nWhen considering the localized persecutions of Christians early on in the first and second centuries, it is no stretch of the imagination to visualize the confiscation, loss, or even destruction of the \u201cautographs and first copies\u201d of the New Testament writings. One must also remember that many of these New Testament manuscripts discovered in the sands of Egypt were cast aside in the trash heaps of Oxyrhynchus, some of them torn to shreds before doing so. This reveals that Christians sometimes threw away biblical manuscripts after a period of use, likely after being replaced with a new copy, rather than being retained for hundreds of years. This is because it was the text of the autographs that was important. Once a good copy of the text was produced, the physical autograph could then be discarded.<\/p>\n
CONCLUSION<\/p>\n
This chapter has given us reason to doubt that the autographs of the New Testament lasted hundreds of years. This obviously challenges Evans\u2019s further claim that these long-lasting autographs of, say, Matthew \u201cwould have exerted influence on the text of Matthew.\u201d Though Evans is confident of the autographs\u2019 influence over the transmission of the New Testament writings, it is further evident from the papyri that closeness in proximity to the physical autographs does not necessitate a reliable or more accurately copied text.
\nA fragment of Julius Africanus\u2019s (AD 160\u2013240) work illustrates this point. P.Oxy. 3.412 is a well-known bookroll portion containing Africanus\u2019s encyclopedic piece titled Kestoi. This work was part of the Oxyrhynchus book collections examined by Houston. Though a good-quality copy of Africanus\u2019s work, it was not valued for long by its owners. The work had to be written no later than the AD 220s, and the roll was reused on the reverse for a will dated to AD 275\u2013276. Thus this roll was used for fifty years at most before the material was reused for a will. Though this copy was produced within twenty years or less of the time Africanus\u2019s Kestoi was first written, \u201cseveral lines of the incantation especially are clearly corrupt, and one of them is incomplete.\u201d
\nTherefore, both of Evans\u2019s assertions\u2014that the autographs may have lasted hundreds of years and that they probably stabilized the textual transmission\u2014fail to take the contrary evidence into consideration. Along these same lines, his argument risks conflating the importance of the autographs as physical artifacts and the text of these autographs. In reality, a later manuscript could faithfully reproduce the text of the autograph despite being removed by hundreds of years in time from the physical autograph. Likewise, an earlier manuscript could poorly reproduce the text of the physical autograph, due to scribal error and other issues, despite being directly copied from or close in time to the physical autograph (as noted above with P.Oxy. 3.412). For these reasons, we should not follow Evans\u2019s argument for trusting the stable transmission of the New Testament text. Reasons for trusting it do exist, but they lie elsewhere, as detailed in other chapters.
\nQuite different from Evans\u2019s claims are Larsen\u2019s, which blur the lines between the composition of the autographs by the author and the alterations made by readers and scribes after the New Testament writings began to circulate. As discussed above, the authors and readers of these accidentally published works still clearly distinguished between the author\u2019s version and the alterations and plagiarizing of subsequent versions. In the case of the Gospels themselves, the earliest known reference to both Mark and Matthew, made by Papias of Hierapolis (late first to early second century), views these Gospels as distinct texts rather than \u201can open textual tradition\u201d (Eusebius, Hist. eccl. 3.39.15\u201316).
\nFrom this survey of publication in the Greco-Roman world, then, we should conclude that when the New Testament writings began to circulate is also when the text of these documents was finalized. From that point on, there exists a stream of copying and distribution that is distinguishable from the earlier stages of composition.<\/p>\n
Key Takeaways
\n\u25ba For the New Testament, the autographs should be conceived of as the completed authorial work that was released by the author for circulation and copying, not earlier draft versions or layers of composition.
\n\u25ba Once a work had been circulated in a community, that writing could not be significantly transformed, either by plagiarizing or altering its text, without these actions becoming known.
\n\u25ba Once a New Testament writing was completed and released for circulation, there may have been a number of copies made under the control of the author or authors.
\n\u25ba It is unlikely that the New Testament autographs still existed and influenced the text by the time of our earliest copies. Even if they did, this alone would not guarantee that the existing manuscripts are reliable.<\/p>\n
CHAPTER THREE<\/p>\n
MATH MYTHS<\/p>\n
HOW MANY MANUSCRIPTS WE HAVE AND WHY MORE ISN\u2019T ALWAYS BETTER<\/p>\n
Jacob W. Peterson<\/p>\n
Numbers are both vital and prone to manipulation. They are meaningful and trivial, concrete and pliable. Without reliable numbers, humankind never could have reached the moon. Without malleable numbers, companies such as Coca-Cola never could have sold the drink Enviga on the premise it contained \u201cnegative calories,\u201d causing the consumer to lose weight just by drinking it.
\nWhen it comes to the number of manuscripts, the New Testament documents are in rarefied company among the wider world of classical authors. As a general rule there are more of them, many are earlier, and they tend to be more complete. But the bulk, while more complete, are not early. The early ones are typically fragmentary and few in number. What should we make of the numbers? Are they meaningful? Are the numbers useful?
\nAuthors who appeal to the number of New Testament manuscripts often get into trouble because of a lack of nuance or by overstating what the numbers actually mean and represent. Variations of the following quote appear repeatedly in the popular literature: \u201cTwo factors are most important in determining the reliability of a historical document: the number of manuscript copies in existence, and the time between when it was first written and the oldest existing copy.\u201d These authors go on to cite the \u201c24000+\u201d manuscripts of the New Testament as sure proof of its reliability.
\nAside from the conflation of textual reliability with historical reliability, such claims commit the logical fallacy of assuming that a larger number and an earlier date necessarily equate to more reliability. The lack of qualification for the manuscript counts is a recurring issue as well. The problem typically occurs when the New Testament is compared to other classical works with respect to the number of manuscripts and the date of the earliest copy. For the New Testament, numbers from five thousand to twenty-five thousand are given, depending on what the author includes, and a date in the early second century is reported. This is usually followed by a summary, such as that of Norm Geisler: \u201cNo other book is even a close second to the Bible on either the number or early dating of the copies.\u201d
\nThe problem is that, even when the number and dates of the manuscripts are right, the claims typically lack crucial context about the chronological distribution of the manuscripts, the fragmentary nature of the earliest manuscripts, and the unequal representation of some parts of the New Testament in the manuscripts. The goal of this chapter is to provide some of that much-needed context. In doing so, we will look more closely at the number of New Testament manuscripts and the significance attached to that number. First, we will tackle the issue of how manuscripts are counted and who counts them. This is followed by a discussion of all the factors that make counting manuscripts difficult. From there, we will discuss the increasing number of manuscripts by considering what it means for a manuscript to be discovered, how they are discovered, what kinds are being found, and how many may still be out there. The chapter will then shift to discussing the problems associated with talking about the number of manuscripts. In particular, we will uncover why simply counting them is not enough and how the manuscripts are distributed by age, content, and size. The chapter will conclude with a practical turn as we deal with how to talk about the manuscript witness to the New Testament text.<\/p>\n
COUNTING MANUSCRIPTS<\/p>\n
Making sense of the number of manuscripts first requires understanding how manuscripts are categorized and counted, and who does the counting.
\nCategories of manuscripts. Most New Testament students are generally familiar with the four categories of manuscripts: papyri, majuscules, minuscules, and lectionaries. These categories are part of the Gregory-Aland (GA) system, devised by Caspar Ren\u00e9 Gregory and continued by Kurt Aland. In the GA system, papyri are designated by a capital P followed by a number (e.g., P100 or ?100), majuscules by a number with a prefixed zero (e.g., 042), minuscules by a plain number (e.g., 1739), and lectionaries by a lowercase l followed by a number (e.g., l249 or \u2113249). Gregory created the system to improve on the varied and often confusing conventions of past scholars. The goal was a uniform means of referencing texts scattered around the world without having to remember and keep up with changing library shelf numbers. It is much more efficient to refer simply to \u201c022\u201d than to its nine shelf numbers! Likewise, the designation \u201cP40\u201d does not require remembering that its shelf number changed from \u201cP.Heid. Inv. no. 45\u201d to \u201cP.Heid. Inv. G. 645.\u201d
\nDespite their many benefits, it is easy to overlook how inconsistent the categories are as a classification system. The first category, papyri, is a materialbased category. Majuscule and minuscule are script-based categories, whereas lectionary is a content-based category. Inconsistencies arise from multiple types of overlap. For instance, a lectionary can be either majuscule or miniscule. Papyri feature majuscule script, meaning the difference between them and majuscules is entirely one of material. This division often leads to the incorrect impression that the two categories exist as chronologically distinct points in the transmission of the New Testament text when, in fact, the two overlap for several centuries.
\nWho counts them and how? As previously mentioned, Kurt Aland continued Gregory\u2019s system for cataloguing manuscripts. In 1959, Aland founded the Institute for New Testament Textual Research (INTF), which is now responsible for the official registry of New Testament manuscripts. One of INTF\u2019s many projects is maintaining the Kurzgefa\u00dfte Liste der griechischen Handschriften des Neuen Testaments (or Concise List of Greek Manuscripts of the New Testament; Liste for short) as a record of all known manuscripts. The Liste provides basic information about the manuscript\u2019s location, its date, its contents, and various physical features, such as line counts and page dimensions. Thankfully, the Liste is now freely available and continually updated online at ntvmr.uni-muenster.de\/liste.
\nINTF is notified when a new manuscript is discovered and provided with images and as much bibliographical detail as possible. Its task then, especially with fragmentary manuscripts, is to ensure that the newly discovered manuscript is not part of some previously identified manuscript. If INTF believes the manuscript is a new one, it is assigned a new GA number and has its information added to the online Liste.<\/p>\n
WHY COUNTING IS DIFFICULT<\/p>\n
Searching the Liste reveals the following manuscripts as the most recently added, as of April 2019, in each category: P139, 0323, 2940, and l2483. Adding those numbers results in a count of 5,885 manuscripts. Yet the typical approximation for how many Greek New Testament manuscripts we have is a bit north of 5,500. Why the hesitancy by most to provide a precise total? In part, as will be discussed later in this chapter, the number would almost certainly be outdated by the time of publication. A main reason, however, lies with how difficult and laborious it is to come up with an accurate count. The 5,885 total just cited, for instance, is surely incorrect, as will be shown. The following discussion covers a few of the reasons why counting manuscripts is so difficult and why we should be content with approximations rather than worry about having the most precise count.
\nDouble counting. As mentioned, one of INTF\u2019s tasks is determining when two samples are from the same manuscript. An unfortunate fact of history is that the papyri have become fragmentary, and some manuscripts were disassembled and sold as pieces. Accordingly, pieces of the same manuscript can turn up in collections all over the world. Matching these pieces is a difficult task, so it is no wonder there are many places in the Liste where portions of a manuscript were assigned two or more GA numbers only to have subsequent scholarship reveal them to be from the same manuscript. One of the more famous examples of this concerns P64 and P67, which were revealed to be two parts of the same manuscript of Matthew. Today, you will see this papyrus referenced in any number of ways, such as P64(=67), P64\/67, and P[64+67], to indicate they are a single manuscript.
\nWhen using the online Liste, numbers in parentheses next to an entry identify this issue. The first time part of a manuscript is listed, all of the other parts are in parentheses and preceded by a plus sign. All subsequent entries for that manuscript have only the initial number in parentheses after an equal sign. For example, here\u2019s how GA 070 and its constituent parts are presented:<\/p>\n
070(+0110+0124+0202)
\n0110(=070)
\n0124(=070)
\n0202(=070)<\/p>\n
However, a scroll through the majuscules reveals several gaps in the numbering, for instance from 0177 to 0181. It is only by consulting the print version of the Liste that it becomes clear 0178, 0179, 0180, 0190, 0191, and 0193 are also reported to be part of 070. To confuse matters more, 0194 is reported as being the same as 0124, which we already know to be the same as 070. A complete record of 070, therefore, includes eleven different GA numbers and is properly 070(+0110+0124+0178+0179+0180+0190+0191+0193+0194+0202). Whenever a manuscript is found to be the same as another, INTF does not reuse the old number. This prevents confusion in later scholarship between the new manuscript with that number and the old one.
\nThe reverse circumstance has happened several times when two or more manuscripts were assigned the same number. This typically happened when multiple distinct manuscripts were bound together in a single codex. One example is a minuscule from the Vatopedi Monastery in Greece with the shelf number 889. It was entered into the Liste with five separate entries featuring the number 2306 followed by the letters a through e. These five manuscripts now have their own GA numbers: 2306 for 2306a, and 2831\u20132834 for 2306b\u20132306e.
\nThere is yet another complication in the seemingly straightforward numbering system in the Liste. Some manuscripts have had sections or quires replaced or added in, with the missing material added by a secondary scribe. In the case of Codex Vaticanus (03) a fifteenth-century scribe added the end of Hebrews (from Heb 9:14 on) and Revelation. Quite sensibly, this supplementary material received a new GA number: 1957. In a number of manuscripts, the supplementary material is not given a new GA number but instead is suffixed in the Nestle-Aland text with an s (e.g., 032 at Jn 1:1\u20135:11 and 1241 for sections in Paul and the Catholic Epistles).
\nIt should be clear now that GA 0323 as the most recently catalogued majuscule does not mean there are 323 majuscules. The case of 070 alone demonstrates the number is inflated by at least ten. Yet double counting is not the only reason why getting an accurate total is difficult.
\nLoss. Another factor affecting our manuscript count is loss. This phenomenon is observable in the Liste, with the location or institute listed as \u201cBesitzer unbekannt\u201d (\u201cowner unknown\u201d), which has been attributed to 136 manuscripts. There are a number of ways this happens, ranging from accidental to illegal. It might sound rather farfetched that a manuscript could be lost in the same way that one loses one\u2019s keys, yet this seems to be the case with a number of manuscripts found in old library catalogues (e.g., 0174). Naturally, manuscripts have a high value both materially and as religious objects, so it is no surprise that many have been stolen over the years. The problem with the two above scenarios for the task of counting manuscripts is that if the manuscript was not well documented, it could reappear and be recatalogued with a new GA number, resulting in it having two entries in the Liste.
\nManuscripts are also lost through more natural causes such as fires, floods, and insects. For instance, many manuscripts from Turin are recorded as destroyed by fire. Then there are manuscripts, such as 1257\u20131259 from a school in Izmir, that are listed as \u201cburnt?\u201d Do these survive and do we count them? An unfortunate result of war has been the loss of several manuscripts as well. From World War II we know that manuscripts were badly damaged or lost completely in the firebombing of Dresden (e.g., 241 and 2039). There are numerous early manuscripts, such as 062, catalogued in Damascus, Syria, that are already listed as \u201cowner unknown.\u201d With the ongoing Syrian wars, their location and survival are even more uncertain. Although different from the types of loss above, many manuscripts have been sold to private owners, which makes tracking them more difficult and all but eliminates using them for research (e.g., 0258, a fourth-century copy of John sold by Sotheby\u2019s).
\nSometimes manuscripts are recorded as lost in the Liste when they still exist. This is the case with 0229, which is reported as \u201czerst\u00f6rt\u201d (destroyed) and formerly located at the Biblioteca Medicea Laurenziana in Florence. However, it turns out the manuscript is housed at the Istituto Papirologico Girolamo Vitelli at the University of Florence and has been known since 1970.
\nIncorrectly added. Finally, there are some manuscripts that were simply incorrectly added to the Liste. In some instances, this was nothing more than a category mistake or a change in cataloguing procedure. This was the case with 0152 and 0153, which are an amulet and ostracon, respectively. Gregory classified them as majuscules; Ernst von Dobsch\u00fctz reclassified them as an amulet and ostracon, respectively; and Kurt Aland subsequently removed them from the Liste. On more than one occasion, manuscripts have made it into the Liste that were not written in Koine Greek. The understandable cases are several manuscripts, such as 2114, that were written in Modern Greek and were initially included. The more confusing cases are those such as 1151 and 1825, written in Slavonic and Syriac, respectively, that made it in.
\nSometimes a manuscript is determined to have some dependence on a print edition. These could be either forgeries or just someone with a desire for a handwritten copy; imputing motive in these cases is undoubtedly difficult. It is rare, but occasionally such a copyist fools people and the manuscript is given a GA number. One of the more famous examples is 2427 or \u201cArchaic Mark,\u201d which was initially dated to the thirteenth century but was shown by Stephen Carlson in an SBL paper to be a nineteenth-century copy of Philipp Buttmann\u2019s 1860 edition of the New Testament. Again, these are likely very rare, but it remains a possibility that further complicates our cataloguing efforts.
\nIn addition to these types of errors, other manuscripts were once included in the Liste only to be taken out because they were shown to be commentaries, homilies, or any other number of types of texts.
\nMaterial bias. There are also manuscripts that, for whatever reason, most scholars agree should not be counted but nonetheless maintain a number. This primarily affects the New Testament papyri. For example, P99 is a fourth-or fifth-century manuscript containing what appears to be a school exercise with Latin and Greek vocabulary and phrases taken from the Pauline Epistles. Surely this should not count as a New Testament manuscript. J. K. Elliott describes this general phenomenon as giving \u201can undue significance to New Testament writings on papyrus, as if the very writing material was itself so important.\u201d Part of the problem with really fragmentary papyri is that it is extremely difficult to determine whether it was a continuous text or something that would not count, such as an amulet, a song, or a commentary. Aland, who was responsible for maintaining the Liste, observes, \u201cAmong the ninety-six items which now comprise the official list of New Testament papyri there are several which by a strict definition do not belong there,\u201d and he goes on to ascribe some of these to \u201cthe occasionally uncritical attitude of earlier editors of the list.\u201d In total, he lists fifteen examples of this among the ninety-six papyri known to him and says, \u201cThese peculiarities are on the whole negligible.\u201d I am not sure how 15 percent of all papyri being potentially incorrectly categorized counts as negligible, but it shows the generally accepting attitude for papyri despite myriad problems in their classification and inclusion in the Liste.
\nSummary. The above discussion serves to illustrate how difficult it is to provide an accurate total for our New Testament manuscripts. An accurate count is clearly not simply a matter of adding up the numbers for the most recently registered manuscripts. A quick look at the majuscules, a comparatively easy group to count, provides plenty of evidence against bean counting. If we just count the number of unique entries in the Liste, there are 282 majuscules. This means that of the 323 items given a majuscule number, forty-one have been stricken from the Liste; most were to consolidate those that were assigned multiple numbers. This number still includes all of the manuscripts with unknown owners and locations and those whose earlier location was known but are presently doubtful, as with the Damascus manuscripts. The number drops to 261 if all of these are removed. Additional questions are raised about what it means to \u201chave\u201d a manuscript, since thirteen of those were preserved on microfilm. Do we count a manuscript that might not physically exist but is preserved via images? What about 0254? It has microfilm images, but they are totally worthless for reading this palimpsest manuscript. At best, we can say that there are between 261 and 282 majuscules. Arriving at a range this precise for the thousands of minuscules and lectionaries would be a monumental task and one that, as will hopefully become clear in what follows, is not worth undertaking\u2014at least not for apologetic purposes.<\/p>\n
DISCOVERING MANUSCRIPTS<\/p>\n
Another factor that makes attempts at providing precise counts of New Testament manuscripts difficult is the continued discovery of more manuscripts. While it is certainly nice to have an ever-increasing supply of witnesses to the New Testament text, we must be cognizant of what is behind this increase. Most importantly, we must ask, What does it mean to \u201cdiscover\u201d a manuscript, and what kinds of manuscripts are being found?
\nHow discoveries happen today. There are a number of possible meanings of discovery in play when a new manuscript is announced. Of course, there is the traditional sense in which a manuscript that was lost for centuries is found again and thus, more properly, rediscovered. Famous examples of this are the papyri from Oxyrhynchus, the Chester Beatty papyri, and the Bodmer papyri, which for various reasons were buried or discarded only to be unearthed in Egypt in the late nineteenth and early twentieth centuries. Although the excavations have stopped, the Oxyrhynchus collection continues to produce new discoveries as scholars work through the thousands of unidentified fragments. More controversially, New Testament papyri have allegedly been found through dissolving mummy cartonnage to separate the layers of papyrus used in its construction. Another team is working to virtually dissect mummy cartonnage with advanced imaging technology. While no New Testament texts are reported to have been found this way, it is illustrative of new ways that scholars are working to find texts in new places.
\nAnother way that manuscripts are discovered involves working with various libraries around the world to bring to light manuscripts that were previously unknown to Western scholarship. In most cases, New Testament text critics do not know about a manuscript unless it has been catalogued by the previously mentioned INTF. Perhaps the leader in this type of discovery is the Center for the Study of New Testament Manuscripts (CSNTM). In our work cataloguing and digitizing manuscripts at monasteries and libraries around the world, CSNTM often \u201cdiscovers\u201d New Testament manuscripts that the library has known for centuries in many cases but about which INTF was unaware. In this sense, a \u201cdiscovery\u201d is the official registration of a manuscript by INTF that was unknown to the wider New Testament scholarly community but was known to its owner or holding library.
\nThere is another type of manuscript discovery that falls in between the two just mentioned that involves New Testament manuscripts hiding in plain sight. Frequently, New Testament manuscripts are found within other known manuscripts, New Testament or otherwise. This can take many forms, such as leaves or quires that were replaced for any number of reasons, a leaf being repurposed as binding material or a flyleaf for another manuscript, or New Testament writings being bound together with non-biblical texts, as is often the case with Revelation being found with other liturgical writings. Along the same lines, new discoveries have been made by recognizing biblical texts underneath other writings in palimpsest manuscripts. The most exciting recent work on this front is the Sinai Palimpsests Project at St. Catharine\u2019s Monastery in Egypt, which is online at http:\/\/sinaipalimpsests.org.
\nRecent discoveries. As someone who has been involved in almost every type of discovery mentioned above, I am acutely aware of just how easy it is to get excited about new finds. New manuscripts are exciting because they increase our overall witness to the text and, in many cases, you are the first person to see the text in hundreds of years. However, not all discoveries are created equally, so it is important to understand what is being discovered today.
\nOn one end of the spectrum for generating interest and excitement, several New Testament papyri have been found in recent years. As mentioned in this book\u2019s introduction, there were 98 papyri listed in NA, published in 1993. That number increased to 127 in the twenty-eighth edition, published in 2012, and since then an additional eleven papyri have been catalogued by INTF. This is a fairly impressive amount of papyri discovered in the past twenty-five years. At the other end of the spectrum is a replacement leaf I recently discovered inside an eleventh-century lectionary at the University of Edinburgh. Whereas newly discovered papyri are studied intensely and get cited in the next edition of critical texts, this newly discovered lectionary replacement leaf will perhaps only ever be mentioned here. My general impression of the majority of recent discoveries is that they tend to align more with this latter camp of interesting but text-critically unimportant manuscripts than with ones that will contribute greatly to our understanding of the New Testament text.
\nThe reason for this impression is that none of the papyri just mentioned are even remotely close in importance to the great finds of the past. Most of them are small fragments with only portions of a few verses surviving. CSNTM, which was founded in 2002, has discovered almost seventy manuscripts in the intervening seventeen years. The most important manuscript among their finds is 0322, a seventh- to ninth-century palimpsest containing a few chapters from Mark. Its work at the Albanian National Archives in Tirana resulted in their biggest find in terms of content, with dozens of relatively complete medieval manuscripts being discovered. Altogether CSNTM\u2019s finds amount to more than twenty thousand pages of text. While these manuscripts should be studied, chances are, the bulk of them are generally representative of the later, abundantly attested Byzantine text. The intention here is not to denigrate any recent discoveries or to suggest the work is unimportant. Rather, the point is to demonstrate that a wide range of texts are being discovered in terms of age, extent, and quality. No one rightly expects every manuscript discovered to be a Codex Sinaiticus, so we should be careful not to give the impression that all new discoveries are of equal significance for textual criticism.
\nHow many are left to be discovered? All of this raises an interesting question about how many manuscripts are left to be discovered. As far as the discoveries from archaeological digs and the like that tend to result in papyri are concerned, it would be impossible and imprudent to venture a guess. As it stands, about one papyrus manuscript has come to light every year since the twenty-seventh edition of the Nestle-Aland text was published in 1993. Of course, there are some years where a handful are published in quick succession, but a one-papyrus-per-year rate is generally reflective of the overall trend since P1 was published in 1898.
\nAlthough CSNTM is not the only organization working with manuscripts, its rate of discovery can be used to gain a rough estimate of how many manuscripts might be sitting on the shelves of libraries around the world waiting to be \u201cdiscovered.\u201d CSNTM has digitized 668 manuscripts since 2002, and 67 of those were new discoveries. Assuming that rate continues for the approximately catalogued 5,300 manuscripts, then there are about 525 manuscripts left to discover. This is an exciting prospect, but we must keep in mind that manuscripts do not have to abide by theoretical calculations, and the rate of discovery could change drastically in either direction. These numbers are little more than a fun thought experiment.<\/p>\n
WEIGHING MANUSCRIPTS<\/p>\n
With the above caveats and difficulties as a background, we can proceed to discuss some of the issues that come up when talking about the number of manuscripts. As a point of entry, issues arise in discussions of manuscript numbers when there is a failure to nuance the point being made. Yes, there are about fifty-three hundred Greek New Testament manuscripts, but they are not all created equally, and that number should not be used as though they are. It is not sufficient to simply count manuscripts; they must be weighed. This is the battle cry of text critics when engaging Majority Text advocates, but it also has potency for those who wish to tout the trove of New Testament manuscripts as a death knell for skeptics.
\nDates of manuscripts. It is generally well recognized that not all of our manuscripts are early. The relationship between date and manuscript importance will be taken up in more detail in other chapters of this book, dealing specifically with why early manuscripts are not necessarily better and why later ones are not necessarily worse, but I want to take an initial look here at the dates of Greek New Testament manuscripts that survive from the first millennium.
\nIt is important to remember from the outset the earlier points about how difficult it is to count manuscripts with complete precision. Accordingly, the numbers in all that follows should be understood in that light rather than as gospel truth. In particular, manuscripts that are missing without images but are still in the Liste are not counted in the following analyses. Further, in this section on manuscript dates, the numbers will be biased toward the early end of the spectrum to avoid doubling. For instance, P3 is dated to the sixth\/seventh century but will be counted only for the sixth. With these considerations in mind, the manuscript distribution by century is shown in figure 3.1 below.<\/p>\n
Figure 3.1. Distribution of Greek New Testament manuscripts by century for the first millennium<\/p>\n
It is clear from this distribution that although there are a good number of early manuscripts, the bulk are still considerably late, with a sizable percentage being lectionaries. Furthermore, this first-millennium window represents only a fraction of the total amount. If there are approximately fifty-three hundred manuscripts in total, 83 percent of them come from the year 1000 or later. This does not mean the number of manuscripts from the second to tenth centuries is insignificant; clearly the New Testament is well attested in this period. Rather, it is intended to be the first step toward putting the many thousands of manuscripts into proper perspective.
\nContents of manuscripts. The next step is to understand what these manuscripts actually contain. There are just over sixty manuscripts of the whole New Testament. This leaves the other 99 percent containing only portions of the text. The standard divisions of the New Testament in the manuscript tradition are the Gospels (represented by e), Acts and the Catholic Epistles (a), the Pauline Epistles (p), and Revelation (r). It is very common to find the Gospels transmitted alone and Acts and the Catholic Epistles together with the Pauline Epistles, but these are hardly rules, and almost any combination can be found. The following chart shows how well the various sections are represented in the first millennium, with a separate column for the lectionary count (see fig. 3.2).<\/p>\n
Figure 3.2. Distribution of Greek New Testament manuscripts by contents for the first millennium<\/p>\n
The important point here is that when speaking of the 889 manuscripts from the first millennium, there are not 889 complete New Testament manuscripts. Indeed, while the Gospels are numerically well represented, the manuscripts containing other parts, especially Revelation, are a small fraction of the total. Breaking these divisions down by century is also helpful, lest we forget that most of these are not early. As with figure 3.1, the numbers in figure 3.3 are biased toward the earliest possible date for each manuscript.
\nDiffering sizes of manuscripts. Even these numbers need to be put into proper perspective. What does it mean to have x number of manuscripts for a particular part of the New Testament? Just as 889 New Testament manuscripts from the first millennium does not mean 889 complete New Testament manuscripts, 158 manuscripts of the Pauline Epistles still does not mean there are 158 manuscripts of that entire corpus. Furthermore, it does not mean that even an entire epistle is contained in a manuscript. On the extremes, Codex Sinaiticus contains the whole New Testament, whereas P12 has parts of a single verse from Hebrews.<\/p>\n
Figure 3.3. Greek New Testament manuscript contents by century<\/p>\n
So, how complete is the earliest witness to some books? Of the Gospels, Mark is numerically the least attested, and the following chart in 3.4 shows how much is preserved prior to the fourth century as percentages of each chapter.
\nThe only manuscripts of Mark from this period are P45, which is absent for 78 percent of that Gospel, and P137, which contains only parts of six verses from Mark 1. John, on the other hand, is very well represented in this earliest period.
\nJohn is found in several significant early manuscripts (i.e., P45, P66, and P75) that cover the majority of the Gospel and then several smaller ones that fill in the gaps. Only fourteen verses in the whole Gospel do not find a witness in this period as seen in figure 3.5.
\nSummary. The point here is not to cause despair about books such as Mark but to point out that not all manuscripts are equal when it comes to early attestation and that even when a section (e.g., the Gospels) has numerous manuscripts, it does not mean that every Gospel has equal representation or that what exists is early and complete. When speaking of the fifty-three hundred manuscripts of the New Testament, these clarifications of the data are the types of things that must be taken into account. The bulk of manuscripts are late, and not everything is represented equally. The Gospels have many manuscripts, while Revelation has very few. Within the textual tradition of each book, some, such as John, have a high percentage preserved early, while some, such as Mark, have considerably fewer in the earliest period. Furthermore, all of the above caveats are secondary to the quality of the manuscripts being counted. Not all of these manuscripts contain texts we would trust or that the scribe necessarily copied well.<\/p>\n
Figure 3.4. Percentage of Mark\u2019s Gospel preserved by chapter prior to the fourth century<\/p>\n
Figure 3.5. Percentage of John\u2019s Gospel preserved by chapter prior to the fourth century<\/p>\n
CONCLUSION<\/p>\n
In concluding this chapter, I want to offer a few points about how to approach and make use of the number of manuscripts in teaching, apologetics, or in any context dealing with nonspecialists. As an overview, the evidence should not be made to say more than it can, and it is crucial to be honest about problem areas in the evidence. If the internet age has taught us anything, it is that people can and will uncover gaps, manipulation, and dishonesty with the data if they are present. Christians, therefore, should be extra careful.
\nThe first point is that bigger numbers do not necessarily equate to better evidence. The reality is that many manuscripts have no noticeable impact on scholarship or critical editions, let alone on, say, English Bibles. It is also true that a very good and reliable text existed with Westcott and Hort\u2019s edition published more than a century ago, when there were far fewer manuscripts available.
\nDon\u2019t count noses. There are two ways to interpret the number of manuscripts responsibly when it comes to the reliability of the New Testament text. On the one hand, a large majority of manuscripts are text-critically unnecessary for establishing the original text, producing no more noticeable effect than a pebble dropped in the ocean. On the other hand, it is precisely this lack of effect that is important when judging reliability. If the bulk of the papyri discovered at the beginning of the twentieth century and all other manuscripts since then have not resulted in major revisions of our critical editions, then this attests to a remarkably stable text that can reliably be reconstructed even without them. The typical newly discovered manuscript is therefore likely to be both statistically insignificant and confirmatory of a reliable text. It is important to note that confirming a text as reliable is distinct from making it more reliable. The latter is certainly reliant on the manuscript evidence, but when manuscripts cease contributing new information, it is dependent wholly on method.
\nUse round numbers. The second point is to be content with round numbers. Using round numbers, such as fifty-three hundred, suggested in this chapter, is prudent for a couple of reasons. All of the complicating factors outlined above make attempts at precision a needlessly tedious task that is almost certain to end in error. Even fifty-three hundred is potentially generous. If the rate of error in cataloguing the majuscules is consistent for every category of manuscripts, then fifty-one hundred is a better approximation. The reason for bumping up the number to fifty-three hundred from fifty-one hundred is that minuscules and lectionaries are less affected than majuscules by the problem of double counting. This is because they tend to be less fragmentary, and therefore there is less likelihood that two or more pieces might be held in different libraries and be given different GA numbers. However, minuscules and lectionaries are still subject to fires and floods, being misplaced, being sold into obscurity, and being incorrectly added to the catalog. Returning to why using a round number is sufficient, it needs to be remembered that finding one more minuscule is not going to convince someone Christianity is true.
\nBe clear about manuscripts\u2019 (limited) apologetic value. This brings us to the final point, which is the importance of understanding what manuscripts can and cannot do. In the negative, manuscripts alone cannot prove the truth of Christianity. What manuscripts can do is provide evidence of a reliable text. A reliable text attested by thousands of manuscripts is just that: a reliable text. But a reliable text is not a guarantee of reliable content. Just as a reliable text of Thucydides\u2019s History of the Peloponnesian War still requires interpretation and verification to assess its historicity, so does the New Testament. Having a reliable enough text is undoubtedly important, because without it arguing for the accuracy of its material would be impossible. Yet providing arguments for the trustworthiness of a text\u2019s actual claims is not something with which textual criticism can help. Those types of important arguments must come from other fields of inquiry.
\nMy hope is that this chapter has been helpful for understanding the challenges faced when dealing with the number of New Testament manuscripts. Textual criticism appears to be more popular than ever and promises to stay at the forefront of many apologetics contexts. Accordingly, anyone working in this area must be aware of the issues and must deal with them honestly and with the required nuance.<\/p>\n
Key Takeaways
\n\u25ba The official catalogue of Greek New Testament Manuscripts is maintained by the INTF in M\u00fcnster. The total number of entries in the INTF catalogue is not necessarily the total number of Greek New Testament manuscripts.
\n\u25ba New manuscripts continue to be discovered, usually in existing libraries or collections. These, however, are by no means equal in size or significance.
\n\u25ba Most manuscripts of the New Testament are only manuscripts of part of the New Testament, and providing an exact count of them is a fool\u2019s errand. It is best to say that there are about fifty-three hundred Greek New Testament manuscripts in existence, although fifty-one hundred might be the safer estimate.<\/p>\n
CHAPTER FOUR<\/p>\n
MYTHS ABOUT CLASSICAL LITERATURE<\/p>\n
RESPONSIBLY COMPARING THE NEW TESTAMENT TO ANCIENT WORKS<\/p>\n
James B. Prothro<\/p>\n
This chapter takes up arguments that uphold the reliability of the New Testament by comparing the great wealth of manuscripts that attest to its text with the relatively slimmer manuscript attestation for works of the classical canon such as Homer or Suetonius. This argument is rhetorically effective for at least two reasons.
\nFirst, it is effective because of the numerical data. Many classical works survive in very few manuscripts, and many are dated to several centuries after the original work was composed. By contrast, the New Testament manuscripts number many\u2014often thousands\u2014more. There are, likewise, manuscripts\u2014fragmentary or otherwise\u2014from within a century after the New Testament documents are traditionally thought to have been composed. Likewise, the great number of manuscripts show substantial agreement across the board as to what the text of the New Testament is. These data engender confidence overall in our knowledge of the New Testament text and in our ability, with so many witnesses, to reconstruct it accurately.
\nSecond, the argument is effective because of the suggestive power of analogy. A skeptic who understands the process of textual criticism might simply ask, \u201cWhat if the New Testament, reconstructed to the best of our abilities, is still not what the original authors wrote?\u201d The comparative argument can then suggest that, if you trust that we have anything close to the original texts of authors such as Caesar or Thucydides (whose works often form a significant basis for history textbooks) despite their much poorer manuscript attestation, you should have at least as much credence in the New Testament.
\nThis chapter will not overturn the basic claims of the data or the practical goal of the comparative argument. Indeed, it is precisely because of the argument\u2019s value that I hope to clear up some myths and mistakes that make the argument vulnerable to critique or that cover over its value. The primary problem is with the numbers given for classical works, which are often inaccurate. A secondary difficulty is when the argument is presented to prove more than it legitimately can. In this chapter, I will try to illustrate the issues and, more importantly, will try to help future apologetics by demystifying some issues and suggesting some strategies for presenting the argument more strongly.<\/p>\n
THE DATA AND THE ARGUMENT<\/p>\n
First, we can get a sense of the argument by looking at some representative versions of it. Though it can be found at least as far back as Richard Bentley in the eighteenth century, perhaps the most influential scholar to employ the comparative argument is F. F. Bruce in his apologetic The New Testament Documents: Are They Reliable? (1st ed., 1943). Bruce also has the distinction of being a trained classicist himself, having taught classics in his first academic posts, which adds to his ethos as commentator on both fields. We will take Bruce as our major representative and then look at a few recent presentations.
\nBruce\u2019s book begins by addressing briefly what the New Testament documents are and their estimated dates. He then asks the question, \u201cWhat is the evidence for their early existence?\u201d He gives as a foil the theories of the old T\u00fcbingen school (F. C. Baur and company) that many New Testament books \u201cincluding the Gospels and the Acts\u201d were written at least a century after Jesus\u2019 death, and he counters by appealing to the \u201cgreater and more conclusive\u201d evidence available for their early existence. This is where the comparison with classics begins: \u201cThe evidence for our New Testament writings is ever so much greater than the evidence for many writings of classical authors, the authenticity of which no one dreams of questioning.\u201d He explains, there are \u201cover 5,000 Greek manuscripts of the New Testament in whole or in part,\u201d and the \u201cmost important\u201d ones (e.g., Vaticanus 03, Sinaiticus 01) date no later than the fourth century.
\nThat Bruce uses T\u00fcbingen as a foil illustrates that, at least in part, he wants to show that the New Testament documents were not written so long after the events they relate (note that he addresses the books that purport to relate events\u2014the Gospels and Acts\u2014especially). But he does not yet offer the reader the earlier papyri available to him, such as those in the Chester Beatty or Bodmer collections. Instead, he turns first to compare manuscript attestation for several classical histories:<\/p>\n
Perhaps we can appreciate how wealthy the New Testament is in manuscript attestation if we compare the textual material for other ancient historical works. For Caesar\u2019s Gallic War (composed between 58 and 50 BC) there are several extant MSS, but only nine or ten are good, and the oldest is some 900 years later than Caesar\u2019s day. Of the 142 books of the Roman History of Livy (59 BC\u2013AD 17) only thirty-five survive; these are known to us from not more than twenty MSS of any consequence, only one of which, and that containing fragments of Books iii\u2013vi, is as old as the fourth century.\u2026 The History of Thucydides (c. 460\u2013400 BC) is known to us from eight MSS, the earliest belonging to c. AD 900, and a few papyrus scraps, belonging to about the beginning of the Christian era. The same is true of the History of Herodotus (c. 488\u2013428 BC). Yet no classical scholar would listen to an argument that the authenticity of Herodotus or Thucydides is in doubt because the earliest MSS of their works which are of any use to us are over 1,300 years later than the originals.<\/p>\n
Bruce then contrasts the textual evidence for the New Testament. While any of these classical works may have only a handful of manuscripts, and most relatively very late, the New Testament\u2019s major witnesses were copied within only a few centuries after the originals were composed, and several papyri are earlier than that! If any modern reconstruction of an ancient work approximates the text of the original document, it is the New Testament.
\nOne might, of course, object that more manuscripts means more errors to deal with in the manuscript tradition. Here Bruce furthers the data he offered in the comparison. With the wealth of manuscripts, combined with the early date of many papyri and quotes from patristic authors, one comes to the conclusion that we could hardly be in a better situation when it comes to ascertaining the original text of the New Testament: \u201cIf the great number of MSS increases the number of scribal errors, it increases proportionately the means of correcting such errors, so that the margin of doubt left in the process of recovering the exact original wording is not so large as might be feared; it is in truth remarkably small.\u201d He quotes Sir Frederic Kenyon\u2019s statement that the chronological proximity of the extant New Testament documents to their times of composition demonstrates definitively \u201cthat the Scriptures have come down to us substantially as they were written.\u201d
\nThe comparison with classics is used to similar ends in recent apologists and scholars, though the data given sometimes diverge from Bruce (and sometimes do not). Craig Blomberg lays it out methodically: anyone asking whether what the New Testament authors wrote is true must first ask whether we even have what those authors actually wrote. His comparison with classics emerges several pages later as an illustration. For the very popular Iliad and Odyssey of Homer, he says, we have no more than twenty-five hundred manuscripts \u201cof those works put together.\u201d Herodotus is attested by only seventy-five, while there are \u201c20 copies of Thucydides, and 27 of the works of the Roman historian Livy.\u201d Blomberg also appeals to the relatively late date of classical manuscript attestation: \u201cAnd the oldest surviving manuscript for any of these authors dates from at least four centuries after the time it was first written, sometimes as many as nine centuries after, versus a gap of only one century, or less, for most of the New Testament books.\u201d Blomberg\u2019s numbers are not identical to Bruce\u2019s, and he adds Homer (whom Bruce does not mention), but the point is the same.
\nA more popular-level apologetic comes from Amy Orr-Ewing, who upholds the New Testament with \u201c24,000\u201d manuscripts (including versions and other witnesses) against the attestation for classical authors: Caesar with ten manuscripts beginning from the tenth century, Thucydides with eight manuscripts beginning from the tenth century, Plato with seven manuscripts from the tenth century, Tacitus with twenty manuscripts beginning from the twelfth century, Suetonius with eight manuscripts from the tenth century, and Homer\u2019s Iliad with 643 manuscripts from the fifth century BC. Orr-Ewing\u2019s figures are closer to Bruce\u2019s than to Blomberg\u2019s while including more classical authors than both. Nevertheless, the point is the same.
\nA final example comes not from an apologetics book but a primer for New Testament text criticism by Stanley Porter and Andrew Pitts. They cite Bruce (above) and provide essentially the same data as he on classical authors: Caesar\u2019s Gallic Wars, eight to nine manuscripts from the ninth century; Livy, twenty fragments (one from the fourth century); Thucydides, eight manuscripts from the tenth century; Herodotus, eight manuscripts from the tenth century. They then offer in comparison the New Testament, for which they claim \u201cover seventy-two hundred Greek manuscripts of various sizes and shapes representing different portions of the NT,\u201d not including versional evidence. From this they conclude, \u201cWhen compared with other works of antiquity, the New Testament has far greater (numerical) and earlier documentation than any other book. Most of the available works of antiquity have only a few manuscripts that attest to their existence, and these are typically much later than their original date of composition.\u201d
\nThese examples will suffice for now to show the argument\u2019s main contours and its use. In each presentation, the argument turns on two data points: (1) the relative wealth of manuscript attestation for the New Testament in comparison with classical authors and (2) the relatively early date of some of these manuscript witnesses. These two points, it should be said now, are indisputable overall. They are just true. However, there are some errors and inconsistencies in these (and other) presentations of the argument. Some are simply errors in numbers, some in the fairness of the numbers compared in the argument. In an argument about numbers, errors in numbers make the argument vulnerable to detraction. Before thinking about the argument overall, the following two sections will illustrate some of the errors and offer suggestions as to how to correct them and improve the argument.<\/p>\n
PROBLEMS WITH NUMBERS AND DATES<\/p>\n
One thing immediately noticeable from the few samples of the comparison above is a basic disagreement about data, particularly regarding the number and dates of manuscripts for the classics. This is true even for those works published within the last decade or so. Porter and Pitts (2015) have only eight manuscripts for Herodotus, but Blomberg (2014) has seventy-five. Orr-Ewing (2005) has eight manuscripts for Thucydides, while Blomberg has twenty. All of these numbers are, of course, lower than those for the New Testament, so the basic points of the comparison stand. But readers interested in the New Testament\u2019s reliability who read more than one book might ask themselves which numbers are correct, and apologists looking to make the argument anew might ask themselves the same question.
\nWhat this discrepancy introduces us to is, in fact, the deeper problem that many presentations of the comparative arguments are simply inaccurate when it comes to these numbers\u2014especially, though not only, when they are reproducing Bruce\u2019s figures. This has been pointed out sufficiently by Clay Jones, an apologist at Biola University. Jones devotes a whole article to updating the numbers in the Christian Research Journal. Jones rightly finds that classical works often have more manuscripts than are commonly given, and some of these are of considerably earlier dates than given.
\nRegarding the number of manuscripts themselves, a few examples will suffice. For Caesar\u2019s Gallic Wars, Jones counts not ten (so Bruce) but 251 manuscripts. Again, it is far less than the manuscript attestation for the Gospels, so the comparison\u2019s basic points remain, but 251 is a far cry from ten! Likewise, Orr-Ewing (and others) cite Plato as attested by only seven manuscripts, when there are again over two hundred. Perhaps the biggest gap in terms of numbers comes with Homer. Several writers state that the Iliad has only 643 manuscripts. Jones shows that internet searches (in 2012) reveal this to be a popular statistic on websites defending the New Testament\u2019s reliability (though notably not websites devoted to Homer). Where precisely authors have gotten this count is not always transparent; it may have come from Bruce M. Metzger, who in a 1963 essay cited numbers that add up to 643 manuscripts for the Iliad.
\nThough the source is credible and the argument is popular, Metzger\u2019s 1963 numbers are now simply inaccurate. Martin West, editor of the most recent Teubner edition of the Iliad, discusses 1,569 papyri in his monograph on that epic\u2019s text (published 2001). Indeed, since West\u2019s monograph, more papyri have been edited and published, some quite old (most dating to the first two centuries on either side of the common era). This, in addition to nearly two hundred nonpapyri (parchment and paper) manuscripts for the Iliad, increases the number of manuscripts by well over a thousand. Presumably, Blomberg\u2019s happily higher approximation for Homer (\u201cless than 2,500\u201d for the Iliad and Odyssey together) is an attempt to update these numbers.
\nIf these examples show that our manuscript numbers are often wildly off, other examples show that inaccurate counting also results in inaccurate figures for classical works\u2019 earliest manuscript attestation. For Herodotus, Bruce and others give about eight manuscripts; Blomberg gives 75. Jones\u2019s update instead gives 109, counting 49 papyri and nearly 60 nonpapyri. But even with these numbers, Jones too errs in reporting that the \u201cearliest surviving manuscript dates from the tenth century AD,\u201d over a millennium after Herodotus (d. 425 BC). Unless Jones means to exclude papyri from what he terms \u201cmanuscripts\u201d (which would contradict his count), we will have to roll this date back almost one thousand years. Papyri from the first or second century AD attest fragments from book 1 and a substantial part of book 2. Books 7 and 8 are attested in the second or third centuries AD. Book 9 was found in a Duke papyrus that was dated to the first or second century AD, but subsequent research suggests it comes from the same roll as another earlier papyrus\u2014putting this fragment of book 9 in the first or perhaps second century BC.
\nLikewise, Thucydides is given eight manuscripts, with dates beginning AD 900, by Bruce, Orr-Ewing, and Porter and Pitts, and given twenty manuscripts by Blomberg. However, even by 1913 (before Bruce), a whole book was filled with papyri of Thucydides. The 1938 Oxford edition of Thucydides speaks of some thirty papyri and parchments in evidence. Such figures, of course, would require updating in later decades, and recently papyri of Thucydides have been found dating even back to the third or mid-second century BC (long before 900).
\nThe initial point is this: numbers of known manuscripts can change, and relative dates for manuscript attestation can change drastically at the discovery of one papyrus. This is simply an issue of fact checking and updating, and it is expected in any discussion of antiquity. That numbers need updating\u2014especially those found in older books\u2014is not in itself a problem. It is a problem when modern apologists simply follow old figures without updating\u2014especially when those apologists are updating their New Testament numbers to include recent finds. It may not be intentional or underhanded\u2014and one should remember that proper apologists are required to be up on several fields (philosophy, archaeology, history, theology) and are stretched thin in their research time. But, again, in arguments about numbers, bad numbers weaken the argument.
\nThe primary set of \u201cmistakes\u201d that need to be dealt with have now been addressed. The rest of the essay now can be more constructive, to help us understand the number counts and consider the comparative argument\u2019s value and best use.<\/p>\n
UNDERSTANDING THE NUMBERS AND COMPARING FAIRLY<\/p>\n
It is important to say here that the New Testament apologist may be forgiven for not having absolute numbers. And one should not assume that apologists have been nefariously skewing data. But, given the above, the interested reader\u2014not least the apologist hoping to better the argument in the future\u2014might also wonder why such discrepancies in numbers and dates are found in the literature at all. If our numbers are wrong, where did they come from? And how can we avoid this in the future?
\nThese questions open up an important issue relating to the comparative argument, one that in part clarifies the discrepancies and also should inform apologists when making the argument: classicists and New Testament scholars do not always \u201ccount\u201d the same way. Especially at the time when many of our standard editions were written, the classical text critic is actually inclined not to count every possible witness. When working with a family tree or stemma of manuscripts for a textual tradition, witnesses that share the same distinctive errors or that are otherwise clearly copied from another known witness are actually removed from consideration when reconstructing the original text. This is because their testimony to the original text is considered merely derivative, contributing nothing distinctive. By this process, known as eliminatio (elimination), the critic\u2019s task can be significantly streamlined. Consider the instructions in Paul Maas\u2019s famous manual: \u201cIt will now be obvious that a witness is worthless (worthless, that is, qua witness) when it depends exclusively on a surviving exemplar or on an exemplar which can be reconstructed without its help. A witness thus shown to be worthless \u2026 must be eliminated (eliminatio codicum descriptorum).\u201d This results in a functional count, which can exclude many witnesses unnecessary for reconstruction.
\nNew Testament text criticism today, by contrast, is far more inclusive. This contrast in method is important. For one, it is itself a result of the larger and more varied textual basis for reconstructing the New Testament; our data set is too different to do it this traditional way. Likewise, because of contamination and scribes potentially checking more than one manuscript in making a copy, even a copy that is late or would otherwise appear derivative overall can preserve an original reading corrupted by earlier manuscripts. For the New Testament, at least ideally, no witness is \u201cworthless.\u201d This leads to an important point that the comparative argument can help bring out: New Testament textual critics are updating and improving their methods based on the nature of the data, discovering more appropriate and effective methods to reconstruct the text in light of what might in other methods be an unruly number of witnesses and variants.
\nThis difference in method should help clarify, in part, why the numbers are so divergent. Take the example of Thucydides. For Thucydides, Bruce claimed \u201ceight MSS, the earliest belonging to c. AD 900, and a few papyrus scraps, belonging to about the beginning of the Christian era.\u201d Now, as mentioned above, the Oxford text just a few years before spoke of some thirty papyri and parchments and several late manuscripts that aid in reconstructing one of the main witnesses. So on an inclusive count, Bruce is wrong, but on a functional count he is on the mark: there are about seven manuscripts from which editions of Thucydides are primarily derived, the others being eliminable or fragmentary (note that Bruce\u2019s number explicitly excludes \u201cpapyrus scraps\u201d).
\nSuch discrepancy in counting can even come up among classicists, with some giving total counts and others functional counts. Regarding Tacitus, F. Haverfield reported approximately twenty manuscripts for the second half of the Annals and Histories in 1916, and in 1941 critic C. W. Mendell reported \u201cthirty known manuscripts\u201d of Annals 11\u201321. Nevertheless, Jones can cite a book from 1999 stating, \u201cTacitus\u2019s historical works descend in two manuscripts, one for books 1\u20136, another for 11\u201316 and the surviving portions of the history.\u201d This is not because we lost a couple of dozen manuscripts but because of the difference in counting.
\nThis can help explain why some of our numbers look the way they do. The question for those making the comparative argument now, however, becomes a question of which way they will choose to count. Jones, for instance, finds the quote above about only two Tacitus manuscripts but finds another mentioning thirty-one derivative manuscripts. Jones commendably counts inclusively and gives the higher number. Porter and Pitts, however, citing the page where Bruce gives eight Thucydides manuscripts and mentions papyri outside his count, give readers the low number and neglect to mention the existence of the excluded papyri.
\nHere I would caution apologists to ensure that they are fair in their comparison. A fair comparison needs to compare like with like\u2014compare apples to apples, as the saying goes. In comparing the manuscript attestation for two different ancient corpora, we are comparing like with like. But fair comparison also demands a like treatment of both corpora. To further the metaphor, it is not simply enough to be comparing two sets of apples; we need to be sure that we keep our eyeglasses on when analyzing each rather than looking with blurry eyes on one apple and with untrammeled sight on another.
\nUnfortunately, apologists\u2019 numbers often reflect an inclusive count for the New Testament but a functional one for classical works. Porter and Pitts give Bruce\u2019s specific and low numbers for classics and then give in comparison \u201cover seventy-two hundred Greek manuscripts of various sizes and shapes\u201d (a gross overestimate on any count). Orr-Ewing gives the low numbers for classical texts, which we have seen usually does not include even all Greek or Latin manuscripts, and then gives a number for the New Testament (twenty-four thousand) that includes fragments and non-Greek versions. Not only is this unfair comparison, but it risks harming legitimate points of the comparative argument if the tables were turned. Indeed, even our traditional and more modest number of five thousand or more New Testament manuscripts includes all kinds of manuscripts that we hardly use to reconstruct the actual text, some fragmentary but valuable for their early dates. Note Bruce\u2019s qualifications to his numbers for classical works: he counts manuscripts that are deemed \u201cgood,\u201d \u201cof any consequence,\u201d or \u201cof any use to us\u201d and does not count early but fragmentary \u201cpapyrus scraps.\u201d But, we should ask, what if we counted the same for the New Testament? What if our many Byzantine-cluster manuscripts were excluded except for a few stemmatic representatives? What if the fragmentary papyri that are valuable for their early dates but useful to consult for only a handful of verses (or words!) of text were not in our count? I do not know exactly what the number would be, but I know it would be nowhere near five thousand.
\nNew Testament apologists (and critics) need to be sure they are comparing fairly, in view not only of the different numbers one can find but the different ways of counting numbers that might influence what one is reading. If we are going to count the New Testament evidence inclusively, we should do the same for the classical corpus we are comparing. What is so wonderful about the New Testament\u2019s textual attestation is that it is better in either case. Even when we are as inclusive as possible in our count of the classical works, the New Testament still has earlier relative attestation, more extensive attestation, and far fewer chronological gaps from the time of composition and the time of the printing press and modern editions. Remembering St. Ambrose of Milan\u2019s dictum \u201cNo man heals himself by wounding another,\u201d we should avoid treating this comparison as a competition. Happily, it is not one. To compare our wealth of data to the lowest possible numbers for classics (and consequently to use the late dates of only the \u201cgood\u201d manuscripts) is completely unnecessary. The health and stability of the New Testament\u2019s textual transmission are evident by the data alone. They are put into sharp relief when compared with the best possible numbers for classical works, but the comparison is not necessary to prove that our basis for the New Testament is very strong.<\/p>\n
THE VALUE OF THE ARGUMENT<\/p>\n
Even with updated numbers and clarification on the counts, the comparative argument still yields its two basic data points: (1) the wealth of manuscript evidence available for the New Testament\u2019s text and (2) the relatively early attestation we have for the New Testament in the manuscript record. But, after all this, one might ask whether these points are still worth making. Is the comparative argument a good argument? Yes, I think it is a good argument. But to make the argument well in the future, we need to consider what the argument is good for. This will help us know what burden we can and should make the argument bear (or not).
\nWe may say first that the comparison is excellent to illustrate the state of play in New Testament text criticism. When instructing people who are learning\u2014in a classroom or elsewhere\u2014about textual criticism for the first time, the comparison is excellent to illustrate the both happy and difficult task of the New Testament text critic. Helmut K\u00f6ster introduces the comparison by saying that it shows the New Testament\u2019s apparently much more favorable basis than most classical texts, while immediately also noting that the complexity of the textual tradition thereby requires more complex methods of editorial analysis. Indeed, it can illustrate why New Testament critics do not and cannot simply exclude some manuscripts from consideration, and an apologist noting the classical principle of eliminatio can use even such methodological differences between the disciplines to demonstrate the New Testament\u2019s very different textual basis and, importantly, introduce the methods we devise to deal carefully and reliably with such a base of evidence.
\nRelatedly, the comparison also helps people understand why we have so many variants, which is especially helpful for the apologist\u2019s task. People hear of something like four hundred thousand variant readings in the New Testament and easily become terrified that nothing we have is reliable. In addition to reminding people that most of these are inconsequential changes in orthography or word order, the data in the comparative argument can put this into perspective. Of course there are so many errors to speak of across a textual tradition that has so many handmade copies! Indeed, it should be a testament to the meticulous concern for detail and exactness that New Testament critics mark and consider all the small and unessential variations that they do. The great number of manuscripts, not to mention the general uniformity in most of the manuscripts, can help put an initially shocking number of variants into perspective.
\nIn addition to putting the problem of our many variants into perspective, the numbers given in the comparative argument also suggest that\u2014apart from having the originals themselves\u2014we have the best data imaginable to work with to reconstruct the originals accurately. Here, however, we need to remain circumspect about what the data can and cannot demonstrate. Some suggest simply from the numbers that we can trust that our reconstructed New Testament text represents the original documents. That New Testament scholars have a vast amount of evidence for the text is unquestionable. However, the wealth of textual data is not enough, on its own, to prove that our reconstructed text reliably matches the original compositions; to believe that we can reconstruct the original well with all the evidence depends largely on whether one trusts whether we have adequate methods for attaining this goal with the evidence we have. This is open to question, logically at least. For those such as atheist objector Robert Price, there will never be proof enough that anything we have corresponds to the original because we do not have the originals against which to check our reconstruction. The gap between the originals and our earliest copies still leaves the logical possibility that all the copies we\u2019re working with descend from already corrupted manuscripts.
\nBut if this counterclaim follows logically, the Christian apologist can easily respond that this logical objection is hardly reasonable. It is here that the comparative side of the argument adds particular value. The comparison with classics allows one to appeal to what skeptics think they know of ancient extrabiblical history and point to its much thinner textual basis. This is Bruce\u2019s move: \u201cThe evidence for our New Testament writings is ever so much greater than the evidence for many writings of classical authors, the authenticity of which no one dreams of questioning.\u201d In brief, the comparative appeal suggests that if you don\u2019t think you can trust the New Testament text, then you really can\u2019t trust any ancient text.
\nThis suggestion is, I think, fair game. In apologetics, one is trying to mount reasons for warranted belief as much as to prove facts. However, we should acknowledge that building warrant for trusting the New Testament on the basis of people\u2019s trust in classical texts comes with risks, because its effectiveness depends on the extent to which people trust the accuracy of the classical documents. Those who are unsure that we can know what historical classical authors wrote will hardly be moved by the comparison, even if they admit the New Testament\u2019s more reliable transmission. This argumentation also cannot ultimately address theories about postauthorial interpolations in the textual tradition or, as with Homer and others, questions about the notion of a single \u201cauthor\u201d in the first place (one suspects this is why Bruce compared only histories, not epics).
\nFinally, we must remember that the comparison cannot ultimately engender confidence in the historicity or theological truth of the New Testament (as happens often in more popular presentations of the argument). Even people who think Herodotus\u2019s text is sufficiently reliable and consult it as a historical source still do not think all his information (or his claims about where he got it) is true. At the end of the day, Christian apologists using this argument do not want people to believe classical texts the same way that they want people to believe the Gospels, and using this argument to suggest textual veracity (rather than textual reliability) can implicitly undermine this end by asking people to trust the New Testament in the same way that they trust classical texts. To say this is not to denigrate the value of the comparative argument\u2014it is very effective in illustrating the wealth of New Testament information and suggesting that we are better situated to reconstruct the original of the New Testament than that of any other ancient book\u2014but it is to say that further arguments about the truth of the texts will require more historical, philosophical, and theological tools than comparing manuscripts for the Gospels and Suetonius.<\/p>\n
CONCLUDING SUGGESTIONS FOR MAKING THE ARGUMENT WELL<\/p>\n
In assessing the comparative argument with classics, this essay has pointed out errors and attempted to contextualize and clarify some reasons why these errors prove to be so common. Still, apologists may come to this point and ask what they should do now. It is a different (and easier) thing to critique from the sidelines\u2014as I have done in this essay\u2014than it is to coach or quarterback! Where should an apologist\u2014who is already stretched for research time and forced to keep up with several fields\u2014go to find better numbers, and is there any way to shore up the comparative argument so that it is more effective or less vulnerable?
\nThe data that emerge in the comparative argument is impressive for the New Testament\u2019s textual basis. No matter how the numbers are updated or changed for classics, the New Testament still has more attestation, better attestation, and better early attestation. Indeed, noted classicist Giorgio Pasquali has said as much of the New Testament: \u201cNo other Greek text is handed down so richly and credibly.\u201d This cannot itself prove that we have exactly what the New Testament authors wrote down, but it does show that the New Testament scholar has better material to work with than scholars of standard classical works. If one has any confidence in our methods of analyzing the text of the New Testament in so many manuscripts (and I think one should!), the data in the comparative argument are an excellent step in engendering confidence in the New Testament. If this is the goal of the comparative argument, let me offer a few practical suggestions about how to make the argument well in the future.
\nKeep it relative. The comparative argument, by its nature, can only demonstrate that the New Testament has a better textual basis for scholars to work with than classical works, not that it has a perfect one. You can argue the latter point, but you\u2019ll need to introduce methodology into the mix. The good news about this is that for the comparison to do what it can do, all you need to show are the two basic data points mentioned at the beginning of this paper: the relatively greater number of New Testament manuscripts and the New Testament\u2019s relatively early manuscript attestation. What this means, thankfully, is that you only need to prove the relative point, not the exact numbers of manuscripts for classical authors.
\nDo not be so specific about numbers. It is unnecessary. The comparative argument\u2019s basic points are not at all compromised if you say something like \u201cfewer than fifty\u201d instead of \u201ceight.\u201d Even if you say something like, \u201cFew of the classical works have more than a couple hundred manuscripts, and most of them have far less,\u201d the point is made by the indisputably greater number of New Testament manuscripts. The more specific your number is, the more vulnerable it is to a skeptic who wants to weaken your credibility. Moreover, even if your numbers are immaculate today, new publications or new dates for previously published papyri might make your numbers inaccurate by the time people are engaging your argument. It would be a shame to allow the value of the comparative argument to be occluded in this way.
\nCheck for updated numbers. Of course, you may want some specific numbers to illustrate the comparison. Where should you go? The first and easiest places to check are recent authors who have researched the numbers more closely\u2014especially Clay Jones\u2019s article cited above. Another easy place to check is the Leuven Database of Ancient Books (LDAB), accessible at www.trismegistos.org\/ldab\/. Because of the nature of the site and the evidence (the list also includes quotations), simply counting entries down the page will not render an exact and definitive \u201ccount,\u201d but it will certainly allow one to check whether another scholar\u2019s estimate is close to the mark. Another benefit of the site is that it often links one to secondary literature by classical scholars. For instance, selecting the site\u2019s \u201cAdvanced Search\u201d option and searching \u201cHomer\u201d in the \u201cAncient Author\u201d box yields several pages of documents. When one clicks on any one document, details are provided such as estimated date, where it is housed, collections in which it is published, and secondary literature discussing it. This is an excellent place to find pieces discussing the state of the text of any author or work, which the apologist can then use to learn more for comparison.
\nCite a classicist. In fact, citing a classicist on the issue has the benefit of showing that you\u2019ve done your homework and allows you to give the year in which the data were published\u2014which allows you to make your point about the numbers but also allows some wiggle room due to updating over time. Nothing argumentatively is lost, and less is risked, if you say, \u201cIn 1971 x scholar counted only y manuscripts of z author; even if those numbers have changed slightly with more recent finds, they hardly compare to the New Testament,\u201d rather than making the absolute claim, \u201cThere are y manuscripts of z author.\u201d (This deferral in determining the number can also be achieved by saying, \u201cThe LDAB currently lists y manuscripts for z author.\u201d) Articles or books by classicists can be found through the references in the LDAB, by checking the introductions to critical editions of classical works (though many are still in Latin), or by standard research methods. I would advise checking more than one source and counting inclusively if you find a disagreement about numbers in the sources such as those shown above.
\nBe more circumspect. Using the LDAB and reading classicists can be time consuming, especially if you mean to give manuscript data on several classical authors. Another effective tack an apologist could take is to discuss only a few isolated examples. This allows you to research a little more and to summarize the difficulties classicists face in reconstructing these texts (in contrast to the situation with the New Testament). Peter Head\u2019s little book on the New Testament text does this well: rather than seeking to give a chart for several classical authors, he focuses only on three\u2014Josephus, Philo, and Tacitus. This technique allows him to summarize more accurately the situation and numbers for each, and it also gives the reader a more vivid contrast between what a textual critic has to work with in these authors versus the New Testament. He is able accurately to arrive at the basic data points of the comparative argument: the New Testament has \u201cmore manuscripts\u201d and \u201cmore early texts, including possibly half a dozen from the second century,\u201d with \u201cbasically no chronological gaps in the manuscript record.\u201d The point is accurate, made sufficiently and vividly, and with less vulnerability.
\nUse the data to introduce text-critical method. The comparative data can put the numerous New Testament variants into perspective and suggest that we are in the best position to understand and evaluate the New Testament text. But this also is an excellent occasion to introduce our methods for dealing with the many, many manuscripts. Show that New Testament scholars not only have better material to work with but that their tools for evaluating the material and reconstructing the text are excellent and constantly being tweaked for improvement. Use the other essays in this volume to show some of the developments. This is where the argument about the New Testament\u2019s relatively better manuscript attestation becomes an argument that what we have for the New Testament is not simply more reliable than what we have for Plato but that what we have is simply reliable.<\/p>\n
Key Takeaways
\n\u25ba Scholars and apologists often count all the manuscripts for the New Testament that exist (an inclusive count), whereas classicists generally only count the ones they need to use (a functional count). This needs to be considered when comparing numbers.
\n\u25ba Apologists\u2019 numbers too often reflect this inclusive count for the New Testament but a functional count for manuscripts of classical works and end up comparing apples and oranges. Whichever count is used, one should be consistent on both sides.
\n\u25ba When counting manuscripts and giving dates for comparison, scholars and apologists also often give numbers that reflect or exaggerate the most recent discoveries for New Testament manuscripts, but do not check for updated numbers or dates for classical manuscripts. Consistency in comparison should be clear here too.
\n\u25ba The comparative argument is valuable but limited; it can demonstrate only that the New Testament has a better textual basis than classical works, not that it has a perfect one. Text-critical methods are what give reliability to our use of the manuscripts, not the numbers alone.
\n\u25ba The more specific your number is, the more vulnerable it is to a skeptic who wants to weaken your credibility.
\n\u25ba Citing a classicist on the issue has the benefit of showing that you\u2019ve done your homework and allows you to give the year in which the data were published\u2014which allows you to make your point about the numbers but also allows some wiggle room due to updating over time.<\/p>\n
CHAPTER FIVE<\/p>\n
DATING MYTHS, PART ONE<\/p>\n
HOW WE DETERMINE THE AGES OF MANUSCRIPTS<\/p>\n
Elijah Hixson<\/p>\n
It is common when reading a primer on New Testament textual criticism, an apologetic defense of the Bible, or often a New Testament introduction to encounter a list of early manuscripts and their ages. For evangelicals, the attraction of early manuscripts is understandable. As Michael J. Kruger writes, \u201cThe less time 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.\u201d
\nThe date of a manuscript is indeed important. It can situate that manuscript in a proper historical context, or in some cases, it can require some readjustment to the understanding of some slice of history to account for the presence of a manuscript with that date. To take one common example, commentators often cite the date of P52 (P.Ryl. 457) as \u201caround AD 125\u201d to prove that John\u2019s Gospel was written in the first century rather than the second century, as argued by a number of nineteenth- and early twentieth-century scholars.<\/p>\n
Unfortunately\u2014and this seems to be especially the case for some evangelicals\u2014an attraction to early manuscripts can and does lead to a number of problems. The cavalier attitude toward the destruction of mummy masks\u2014artifacts of cultural heritage\u2014in the quest to find early manuscripts of the Bible has earned Scott Carroll and Josh McDowell a good deal of criticism from friend and foe alike. Appeals to P137, alleged to be a first-century fragment of Mark, became so common that when it was finally published, some were disappointed at the earliest fragment of Mark 1.
\nEarly manuscripts of the New Testament are not always good witnesses to the original text either. P72 (P.Bodm. VII\u2013VIII) is the earliest substantial manuscript of 1-2 Peter and Jude, but it seems to have been written by a rather careless scribe. A number of scholars even mention that P72 appears to have a noticeable tendency to change the text in order to emphasize the divinity of Jesus. James R. Royse writes of P72, \u201cWe seem to be compelled, therefore, to conclude that here too [at Jude 5] the scribe has deliberately introduced a reading in order to ascribe Deity to Christ.\u201d Bart Ehrman mentions the same reading in P72 as one of the \u201corthodox corruptions of Scripture\u201d he writes about. Even if the scribe was not actually responsible for the \u201corthodox corruptions,\u201d they are in the manuscript, and there are enough of them to form a pattern. It is therefore inadvisable to assume without qualification that earlier is always better, more accurate, or less likely to contain \u201ccorruptions\u201d when one of the earliest manuscripts of 1-2 Peter and Jude looks as though it was written by a copyist who changed the text in places to make a stronger case that Jesus is God.
\nWhen it comes to reporting the dates of manuscripts, some evangelical scholars and apologists often give unjustifiably early or narrow date ranges. By doing so, they leave the door open for criticism that they are misrepresenting the evidence. On early dates for P52, for example, Paul Foster writes, \u201cFor some people, having an early NT fragment is perhaps more of an ideological commitment than something that is established on the basis of close analysis.\u201d Such appeals to specific or early dates of manuscripts are often inappropriate in light of the uncertainties of paleographical dating, and as we will see, P52 does not automatically rule out a second-century composition of John\u2019s Gospel. This chapter seeks to explain the ways manuscripts are dated while explaining limitations of dating methods, to examine a case study of a commonly cited early manuscript whose \u201cearly\u201d date has been revised (P52), and to offer some solutions to nonspecialists who discuss early dated manuscripts in their work. The discussion emphasizes early dated manuscripts for two reasons. First, it is more common for such manuscripts to appear (often by name) in apologetic defenses of the New Testament. Second, there are fewer securely dated manuscripts in the earlier centuries, and consequently, paleographic dating is less reliable for early manuscripts than for later manuscripts.<\/p>\n
METHODS FOR DATING MANUSCRIPTS<\/p>\n
Occasionally a manuscript can be dated with certainty, and here we look at a few examples. Such instances include manuscripts whose scribes recorded the date of their production, with references to particular historical persons who can be identified elsewhere, or manuscripts with an archaeological context that settles the date of the manuscript. Even if a date cannot be assigned with complete confidence, historical or archaeological considerations can sometimes provide the earliest possible date, known as the terminus post quem (literally \u201climit after which\u201d), or the latest possible date known, as the terminus ante quem (literally \u201climit before which\u201d).
\nContent. Sometimes the scribe of a manuscript recorded when he or she produced it. One example is the scribe of GA 1415 (Athens, National Library of Greece, Manuscript 123).<\/p>\n
Figure 5.1. The colophon in GA 1415 (Athens, National Library of Greece, Ms. 123), f. 189r telling the precise date on which it was finished<\/p>\n
Beneath the decorative line, the scribe wrote \u03bc\u03b7\u03bd\u1f76 \u03c6\u03b5(\u03b2\u03c1\u03bf\u03c5\u03b1\u03c1\u03b9\u03c9) \u03ba\u0305\u03b1\u0305 (\u03b9\u03bd\u03b4\u03b9\u03ba\u03c4\u03b9\u03c9\u03bd\u03bf\u03c2) \u03b7\u0304 \u03b5\u03c4(\u03bf\u03c5\u03c2) \u03db\u0305\u03c7\u0305\u03bd\u0305\u03b3\u0305 (m\u0113ni phe[brouari\u014d] 21 [indikti\u014dnos] 8 et[ous] 6653), which means, \u201cMonth of February, the 21st, in the 8th indiction [i.e., the eighth year in a fifteen-year cycle]; year 6653.\u201d Dates were written relative to the creation of the world, which according to the Byzantine system was September 1, 5509 BC, so February 6653 translates to AD 1145.
\nEven without such a written date, manuscripts might still contain datable information. Sometimes a piece of papyrus was reused by writing another text on the back. If a piece of papyrus lacks a date on one side, a date or a reference to a datable person or event on the other side might still narrow the possible date range by providing a terminus post quem or terminus ante quem. Eric Turner undertook a study of reused papyri where both sides were dated and concluded that official documents such as contracts or deeds were generally reused within one hundred years, with \u201ca slight balance of probability in favour of re-use within 25 years.\u201d
\nManuscript content can also provide a terminus post quem. Turner mentions that P.Bodm. XX, a manuscript of a work describing the martyrdom of a bishop named Phileas, \u201cwhich on palaeographical grounds might plausibly be assigned to [the third century], cannot be earlier than c. A.D. 306, the year in which its hero-bishop faced his persecutor.\u201d Likewise, copies of a book obviously cannot be made before the author finishes writing it. P.Oxy. 3.405 was originally unidentified and dated \u201cnot later than the first half of the third century, and might be as old as the latter part of the second.\u201d It was later identified as a fragment of Irenaeus\u2019s Adversus haereses 3.9. Although the hand was judged to be from AD 150\u2013250, the copy could not have been made before Irenaeus wrote the work, around AD 180.
\nArchaeological indicators. Occasionally, the archaeological context can provide clues for dating. A volcanic eruption, for example, provides a terminus ante quem for the Herculaneum papyri. They obviously must have been written before Herculaneum and Pompeii were destroyed by the eruption of Mount Vesuvius in AD 79.
\nOne Greek New Testament papyrus that can be dated from its archaeological context is P10 (P.Oxy. 2.209), a writing exercise that contains Romans 1:1\u20137. In its editio princeps, Bernard Grenfell and Arthur Hunt note that it \u201cwas found tied up with a contract dated 316 A.D., and other documents of the same period.\u201d AnneMarie Luijendijk made the connection that the contract to which Grenfell and Hunt refer was one they had already published, P.Oxy. 1.103, a lease for a plot of land. More importantly, P.Oxy. 1.103 is part of a larger archive\u2014a collection of papyri compiled by a single individual in antiquity. This particular individual was one Leonides, son of Theon. Other dated manuscripts in his archive give it the range of AD 315\u2013334. Despite the fact that P10 is a part of Leonides\u2019s archive, it is unclear who exactly wrote it. Romans 1:1\u20137 is written in a different kind of script from the rest of the archive, so it is impossible to prove that the same person wrote both, simply because there is no point of comparison for the style of handwriting used for Romans. (Imagine going through a set of notes from someone that are written entirely in cursive and finding a single page of block-letter writing.) Whether Leonides himself wrote P10 or whether it is the fourth-century Egyptian equivalent of children\u2019s refrigerator art, it clearly does seem to be a writing exercise. In any case, we have a twenty-year window for its likely date because of the archive of which it is a part.
\nArchaeological evidence is not always so certain, however. The St. Cuthbert Gospel (London, British Library, Add Manuscript 89000) was first discovered in 1104, when St. Cuthbert\u2019s tomb at Durham Cathedral was opened. It was buried with him, so presumably it would have a terminus ante quem of the date of his burial\u2014he died on March 20, AD 687. However, Richard Gameson notes that the manuscript is too late to have been made during Cuthbert\u2019s lifetime. Based on several features, Gameson dates the manuscript to AD 710\u2013730. In this case, the discovery of the manuscript in the tomb of St. Cuthbert does not prove that it was written during Cuthbert\u2019s lifetime; it proves only that it was placed in his tomb at some point before the tomb was opened in 1104.
\nPaleographic dating. Though secure dates are ideal, in many cases, ancient manuscripts\u2014especially manuscripts of literary works such as the New Testament\u2014lack any securely datable features and must be dated paleographically. Paleographic dating relies on the appearance of the handwriting to assign a date range during which the manuscript was most likely produced. The general idea is that Greek (or Latin, Coptic, etc.) was typically written in one of a number of given styles, and handwriting within these styles changed over time. Particular styles of handwriting can be anchored here and there by securely dated manuscripts. For any given manuscript, a paleographer will first identify the style. Then he or she will decide where in the overall development of that style the hand falls\u2014within a particular style of handwriting, earlier manuscripts have earlier features, and later manuscripts have later features. A date is assigned based on where the hand sits on that early-to-late spectrum. The real situation is of course more complicated than I have presented here, but in general, this is the way dates have been traditionally assigned to manuscripts by paleographers.
\nAlthough the method is widely used, it is not perfect. Eric Turner is skeptical of putting too much weight on \u201cstyles\u201d of handwriting because \u201cthe term \u2018style of handwriting\u2019 suggests an external reality [that may not have existed], or at least an ideal present in the minds of the scribes who wrote it.\u201d Even if one assumes that the style theory works, manuscripts can have mixed features within a given style. Sometimes late manuscripts have \u201cearly\u201d features, or vice versa. Sometimes two manuscripts written in the same style of hand can be extraordinarily similar, despite having been written decades or even centuries apart. For this reason, responsible paleographers have always been careful with dates, often advocating broad ranges rather than narrow ones. According to INTF, for example, the hand of P94 can be dated to only a two-hundred-year window\u2014fifth or sixth century. Recent studies have placed increased emphasis on caution when assigning narrow date-ranges. Of the commonly given date of \u201caround 125\u201d for P52, Paul Foster writes, \u201cSuch a precise date is, of course, nonsense. Paleographical analysis does not allow such precision for undated texts.\u201d Turner warns that for \u201cliterary\u201d hands, \u201ca period of 50 years is the least acceptable spread of time.\u201d
\nA comparison of P.Ryl. 16 and scribe A of Codex Sinaiticus illustrates the problem that resulted in Turner\u2019s hesitancy to assign any date ranges shorter than fifty years on the basis of paleography alone. These two manuscripts are written in the same general style, called \u201cbiblical majuscule,\u201d but they were written forty to two hundred years apart from each other. The comparison aims to show why one should not be too eager to accept the earliest possible or even the latest possible dates.<\/p>\n
Figure 5.2. P.Ryl. 16 is a fragment of a comedy with a dated letter on the backside<\/p>\n
P.Ryl. 16 is a fragment of a comedy that was reused as a letter. The letter on the back can be dated to AD 256, giving this papyrus its terminus ante quem, but we do not know how long before AD 256 it was written. In 1967, Guglielmo Cavallo paleographically dated the manuscript to the narrow window of circa 220\u2013225 (!). The manuscript is part of an archive, however, which can shed some additional light on its date. In a study of reused papyri with dates on both sides, Turner suggested that P.Ryl. 16 might be as early as AD 150 because other reused papyri in the archive were about one hundred years old before reuse.
\nThe other hand of our comparison is scribe A of Codex Sinaiticus, which has a terminus post quem because it was produced with the Eusebian apparatus in the Gospels. It obviously could not have been made before Eusebius had developed this cross-reference system, which probably happened during the window of AD 290\u2013340. Codex Sinaiticus could be later but not earlier. On the basis of the handwriting, Cavallo dated Codex Sinaiticus to circa 360 \u201cor a few years later\u201d (!).
\nCavallo described the hand he calls \u201cbiblical majuscule\u201d as emerging from a \u201csober and undecorated script.\u201d \u201cTrue biblical majuscule\u201d has a \u201cvisible contrast between thin horizontal strokes and fatter vertical ones (particularly gamma, pi, tau), while oblique [i.e., diagonal] strokes appear in between (alpha, delta, lambda).\u201d Cavallo continued, \u201cAmong late examples \u2026 the script shows a stronger contrast between fat and thin strokes and decorative buttons at the extremities of the latter, in particular on the horizontal strokes of gamma, delta, epsilon, pi and tau.\u201d In other words, the hand that we are comparing should have thick vertical strokes and thin horizontal strokes. The later the hand is, the more the contrast there will be. Later examples of the hands tend to have decorative serifs on horizontal strokes, but earlier examples tend not to have them.
\nIn the above comparison, we see handwriting from two manuscripts that Cavallo dated 135 to 140 years apart, and to be fair, P.Ryl. 16 is earlier than Codex Sinaiticus. There are differences between the two hands, admittedly. The epsilon (\u03b5) and sigma (\u03f2) are not the same\u2014they are slightly more decorative in Codex Sinaiticus. The phi (\u03c6) in Codex Sinaiticus is more angled than round. Codex Sinaiticus has generally more difference between the thick vertical strokes of nu (\u039d) and the thinner diagonal stroke (but not always; see the \u03bd in \u03bc\u03b5\u03bd\u03b5\u03b9\u03bd [menein]). Still, the two hands bear a remarkable similarity. Additionally, tau (\u03a4) in P.Ryl. 16 has more prominent serifs, or \u201chooks,\u201d on the ends of the horizontal stroke than tau in Codex Sinaiticus\u2014a feature Cavallo considered to be more characteristic of later examples of \u201cbiblical majuscule.\u201d Cavallo himself dated the manuscripts 135 years apart, but this comparison is even more problematic if P.Ryl. 16 is a few decades earlier, as Turner argued.<\/p>\n
Figure 5.3. A comparison of P.Ryl. 16 (left) and scribe A of Codex Sinaiticus (right), showing their similar hands<\/p>\n
The point of the comparison is to illustrate how difficult it can be to assign a date paleographically. Yes, something is different between the two hands, but can we accurately measure it to be 135 years\u2019 worth of development? Furthermore, \u201cdevelopment\u201d assumes that the styles changed in linear fashion, as Turner noted. But, he said, \u201cIf [a particular style] was written in several centres it is likely that cross-influences will have affected this style, as they did other styles.\u201d To state it alternatively, do we really see 135 years\u2019 worth of linear development between the hands of P.Ryl. 16 and scribe A of Codex Sinaiticus, or are the two similar because the scribes had similar influences? If the latter, we cannot really be confident enough to assign such narrow dates as Cavallo does.
\nAdditionally, it is possible for scribes contemporary with each other to have conflicting features with regard to date. P. J. Parsons mentioned features of Codices Sinaiticus, Vaticanus (03), and Alexandrinus (02), in which a scribe or scribes exhibit later features of handwriting than other scribes working at the same time on the same manuscript. Even a near-perfect match with a securely dated manuscript is not enough to assign a secure date paleographically.
\nOther methods. Other aspects might occasionally provide more information on the date of a manuscript. For example, Turner suggests that the format and dimensions of a codex might shed light on its date. Illuminated manuscripts have an additional layer of available information in their art. If a manuscript has been found with an intact cover, the cover might provide a terminus post quem. A deed dated October 7, AD 348, for example, was reused in the cover of Nag Hammadi Codex VII, indicating that the codex had to have been made after that date and suggesting a similar date range for the other Nag Hammadi codices.
\nScientific testing could be of some use as well. The York Gospels (York, Minster Library, Add Manuscript 1) have a terminus ante quem of around AD 1020 because the manuscript came to York through Wulfstan, the archbishop of York, who died in AD 1023, but it was unclear how long before AD 1020 it was written. DNA analysis revealed that an unusually high percentage of the parchment sampled came from female calves. One would expect male calves to be used for parchment because males are less valuable for maintaining and growing a herd. Cattle numbers in the British Isles suffered in the late tenth century because of a cattle plague that swept through between AD 986 and 988, and female cattle would have been especially valuable to repopulate herds after the plague. The best explanation for the unusually high presence of parchment made from female calves in the York Gospels is that its pages were made from the skins of female calves that died in that plague, which, if correct, suggests a strong possibility for a composition date of the York Gospels around the year AD 990.
\nRadiocarbon dating might be a possibility, though it requires the destruction of a small piece of the item dated, and most conservators would probably forbid such destruction. Additionally, scientific testing is often not able to settle historians\u2019 questions. R. E. Taylor and Ofer Bar-Yosef write, \u201cRadiocarbon \u2018warps\u2019 create periods throughout the 14C time scale where there are inherent systemic limitations in the precision with which a 14C-based time segment can be expressed.\u201d Like paleography, radiocarbon dating can give only a range of most probable dates, but because of these \u201cwarps,\u201d there are periods within the overall timeline where radiocarbon dating results are less useful than others. For example, as Josephine K. Dru shows, \u201c14C science can often distinguish the 130s from the 40s CE. But it cannot distinguish the 130s from the 220s CE\u2014though the time difference is the same!\u201d These specifics mean that, if P52 were to be radiocarbon dated, the best we could hope for (if the tests achieved state-of-the-art accuracy and precision) is a better sense of whether its papyrus (the material on which the text was written, not the writing itself) originated before 130, between circa 135\u2013225, or later. Carbon-14 analysis could not show whether 135, 150, 175, 200, or 225 is more probable. Is that worth the cost of a slightly smaller P52?
\nIn summary, the process of dating a manuscript often involves a combination of methods and\/or sources, and even still, assigned dates must be taken with caution. Turner writes, \u201cPalaeography is neither a science nor an art, but works through a continual interaction of the methods appropriate to both approaches. And in the last resort a judgment has to be made\u2014and judgment [sic] is fallible.\u201d<\/p>\n
P52: DOES IT DATE TO AROUND 125?<\/p>\n
C. H. Roberts and the earliest manuscript. In light of the difficulties of paleographic dating and the fallibility of judgments, we turn to the fragment that holds the distinction of being the most commonly cited early manuscript: P52 (P.Ryl. 457). It is a fragment of John\u2019s Gospel about three-and-a-half by two-and-a-half inches in size to which academics and apologists alike frequently appeal as the oldest extant manuscript of the New Testament. When C. H. Roberts first published P52 in 1935, he assigned it a date on the basis of its handwriting: \u201cOn the whole we may accept with some confidence the first half of the second century as the period in which P.Ryl. Gr. 457 was most probably written.\u201d Roberts arrived at that date by comparing the handwriting of P52 to that of other known papyri. It is important to note that Roberts did not say that P52 was written between AD 100 and 150 but that it was most probably written then.
\nA few considerations provide reason for revising Roberts\u2019s early date for P52 (\u201cmost probably\u201d AD 100\u2013150). First, Roberts\u2019s two closest matches to the hand of P52 were not themselves securely dated. Second, the securely dated specimens in general were not close matches. Third, there are now many more published manuscripts with which to compare P52 than when Roberts first published it in 1935, such that consensus regarding the paleographic dates can change. In the case of one of the two \u201cclose matches\u201d\u2014P.Egerton 2\u2014it did. Roberts compared P52 to an early dated manuscript that is no longer considered to be so early. A recent redating of P.Egerton 2 concluded that it dates to circa AD 150\u2013250 and that \u201cit is not impossible that [P.Egerton 2] was produced sometime at the turn of the third century.\u201d In summary, Roberts assigned the early date to P52 primarily on the basis of problematic comparisons or manuscripts without secure dates, one of which has recently been dated much later. In light of the many papyri published since 1935, we must now consider whether P52 really is as old as is commonly claimed.
\nRecent rejection of an early date. In 1975, Eric Turner accepted Roberts\u2019s date of circa AD 100\u2013150, but with reservation. Turner mentioned P.Amh. II 78 as a manuscript with \u201csimilarities\u201d\u2014it is a petition firmly dated to AD 184. In 1989, Andreas Schmidt identified two other manuscripts\u2014still only paleographically dated\u2014that exhibit close similarities to P52. Schmidt then suggested that P52 dates to circa AD 170, plus or minus twenty-five years. Sixteen years after Schmidt\u2019s study, Brent Nongbri made an even stronger case for adopting a wider date range for P52.
\nNongbri began his important article thus: \u201cThe thesis of this paper is simple: we as critical readers of the New Testament often use John Rylands Greek Papyrus 3.457, also known as P52, in inappropriate ways, and we should stop doing so.\u201d To be perfectly honest, he isn\u2019t completely wrong. His work demonstrated that most of Roberts\u2019s comparanda (manuscripts with comparable handwriting) are not close matches to P52, and he also introduced several additional securely dated comparanda. Nongbri concluded, \u201cAny serious consideration of the window of possible dates for P52 must include dates in the later second and early third centuries.\u201d It is important to note that Nongbri is not suggesting that we abandon the possibility of an early date of P52. Rather, he is arguing that the early date range for P52 should be extended to include later dates.
\nMore recently, other manuscript specialists have rejected the AD 100\u2013150 date for P52. Don Barker, a papyrologist at Macquarie University, writes, \u201cIt is difficult to place [P52] in a very narrow time period,\u201d and he assigns P52 to anywhere in the second or third centuries. Barker continues, \u201cThis may be unsatisfactory for those who would like to locate [P52] in a narrower time frame but the palaeographical evidence will not allow it.\u201d Christian Askeland cites Nongbri and Barker\u2019s work with approval, condemning \u201cindefensible arguments for ridiculously early dates of various New Testament papyri\u201d in his own article on the difficulties of paleographic dating. Whether or not Nongbri and Barker are correct that P52 could be as late as the third century, they are absolutely correct that dates of circa AD 125 or AD 100\u2013150 are too early.
\nHow useful is P52 to apologetics? First, P52 does not and cannot offer definitive proof that John\u2019s Gospel is a first-century composition by an eyewitness. Even if P52 were written in the afternoon of April 26, AD 125 (it wasn\u2019t), it would prove only that sections from John 18 were in Egypt by AD 125. Technically, such a date does not prove that John\u2019s Gospel was in its \u201cfinal\u201d (canonical) form by then, nor does it prove that the text it contains is any more than a few months old. An early date of P52 might render these possibilities unlikely\u2014even extremely unlikely\u2014but it cannot disprove them. Two examples from redaction-critical commentaries demonstrate this point. First, Rudolf Bultmann accepted a date of P52 in the period of AD 100\u2013150 and still argued that as much as forty years could have passed between the original writing of John\u2019s Gospel and a final redaction that left it in the canonical form we have today. Second, Walter Schmithals was well aware of the existence of P52, but he still dated a final redaction of John\u2019s Gospel to around AD 160\u2013180. Given the uncertain nature of paleographical dating and the fact that P52 has not deterred source-critical scholars from adopting second-century dates of a final redaction to John\u2019s Gospel, we quote again Paul Foster\u2019s remarks about the usefulness of P52: \u201cWas John\u2019s Gospel written before the end of the first century? Yes, probably. Does P52 prove this to be the case? No, probably not.\u201d
\nSecond, the occasional absence of P52 in critical editions of the Greek New Testament shows why it is not necessary for reconstructing the original text. The UBS (published in 1975) makes no mention of P52. It is included in the index of manuscripts to the UBS and UBS, but neither edition cites P52 for or against any variant in the text. The NA contains eleven variation units for John 18:31\u201333, 37\u201338 (the verses for which P52 is extant for at least part), but because the papyrus is so fragmentary, the NA cites P52 only once. The one instance is a variant more or less with respect to word order at John 18:33. The reading of P52 was adopted as the text of John\u2019s Gospel against the reading of the majority of manuscripts, but Westcott and Hort had already adopted this reading as the text of John decades before P52 was published. Likewise, P52 makes no appearance in the apparatus of the recent Greek New Testament Produced at Tyndale House Cambridge, an edition that emphasizes its dependence and use of early manuscripts and scribal habits not only for the text but even the spelling and paragraphing of the New Testament.
\nThird, some scholars who are neither trained papyrologists nor paleographers have proposed unusually early or narrow dates for P52, and these dates should not be accepted. Karl Jaro\u0161 (AD 80\u2013125), Philip Comfort (AD 110\u2013125), and Carsten Peter Thiede (AD 80\u2013130) are each controversial for their early dates, which have failed to gain scholarly acceptance. Pasquale Orsini and Willy Clarysse criticize Jaro\u0161 and Comfort (specifically, the edition of New Testament papyri he cowrote with David Barrett) for their early dates, and they dismiss Thiede altogether. With respect to the date of P52, Orsini and Clarysse note that the comparanda used by Comfort-Barrett and Jaro\u0161 are inappropriate because they are not even the same \u201cstyle\u201d of handwriting as P52. They show that the early dates proposed by Jaro\u0161 and Comfort-Barrett are methodologically unsound, concluding, \u201cBiblical scholars should realise that some of the dates proposed by some of their colleagues are not acceptable to Greek palaeographers and papyrologists.\u201d Christian Askeland rightly rejects the conclusions of Jaro\u0161, Comfort, and Thiede, describing their respective works under the heading \u201cPaleography Gone Wrong.\u201d Craig A. Evans describes \u201cassertions of very early dates for some papyri\u201d by a few scholars including Thiede and Comfort as \u201cespecially problematic.\u201d Many recent scholars who mention the date of P52 give it a mid-second century or late-second century date.
\nAs it happens, P52 is not the only New Testament manuscript to receive such controversial treatment from these authors and others. In his Commentary on the Text and Manuscripts of the New Testament, Comfort cites \u201cmanuscripts with certain dating\u201d in defense of his circa 200 date for P45, but none of the manuscripts he cites are dated securely. Two are reused rolls with dates on one side (but not the side relevant to P45), and the other is a paleographic date of a cursive hand\u2014admittedly easier to date than a literary hand but by no means certain. Comfort also mentions E. C. Colwell\u2019s famous study of scribal habits, including Colwell\u2019s conclusion that the scribe of P75 copied \u201cletter by letter,\u201d but he seems unaware of Klaus Junack\u2019s study from 1981 criticizing Colwell\u2019s conclusions about \u201cletter by letter\u201d copying or Dirk Jongkind\u2019s more recent article that touches on the same issue. Comfort cites H. J. M. Milne and T. C. Skeat for having \u201cdemonstrated that Scribe A of Codex Vaticanus was likely the same scribe as Scribe D of Codex Sinaiticus,\u201d but this statement is simply untrue. Milne and Skeat came to the opposite conclusion. After a discussion of the similarities between the hands, they write, \u201cIt would be hazardous to argue identity of the two hands.\u201d Comfort even appeals to Guglielmo Cavallo for a mid-second-century date of P66. Cavallo did accept that date in 1967, but he changed his mind in favor of a later date by 1975, so Comfort\u2019s appeal to Cavallo in the present tense is misleading. In general, one gets the impression that Comfort gives emphasis to references that could be used to support his controversial conclusions, and even then he does not always represent them accurately.<\/p>\n
CONCLUDING SUGGESTIONS<\/p>\n
We do have early manuscripts of the New Testament, and apologists are right to appeal to them. Even if our extant witnesses are not quite as early as we once thought, the number of early manuscripts of Christian Scriptures is a testimony to their importance to early Christians. That we can even identify tiny fragments as New Testament manuscripts by the text they contain is a testimony to the macrostability of the New Testament text. Below are a few practical suggestions for identifying and reporting the dates of early New Testament manuscripts responsibly.<\/p>\n
\u2022 Always use a full date range for a manuscript rather than the midpoint or early end of a date range. A date of \u201ccirca AD 200\u201d might imply either a range of 175\u2013225 or a range of 150\u2013250, and it does not convey the fact that a manuscript dated to 150\u2013250 is just as likely to be written in 248 as it is likely to be written in 156, or anywhere in between. We simply cannot be more precise.
\n\u2022 Avoid sensational dates and excessively early or narrow date ranges. If it sounds too good to be true, assume that it is. Remember Eric Turner\u2019s assessment that for literary hands (which applies to almost all early New Testament manuscripts), fifty years is the smallest acceptable window for a date range assigned on paleographic grounds. A date in the first century for any New Testament manuscript is an immediate red flag\u2014especially for an unpublished manuscript.
\n\u2022 Accept the full range of the date given by INTF. This information can be found in the back of NA (pp. 792\u2013819) or via the electronic version of their official register of Greek New Testament manuscripts. The electronic availability of the Liste ensures that their information is freely accessible with nothing more than an internet connection, and as they are the keepers of the official register of New Testament manuscripts, no one can fault a nonspecialist for trusting their judgment. INTF dates P52 in the second century. The online Liste assigns P52 to the mid-second century (125\u2013175), but NA gives the full second century as the possible date. I suggest adopting the broader range. For P66 and P75, the online Liste has dated them to the early third century, with dates of AD 200\u2013225. NA reports \u201cc. 200\u201d for each manuscript. Thus, in these instances, I suggest accepting the broader range of AD 200\u2013225. Turner\u2019s objections to a range of fewer than fifty years notwithstanding, again\u2014no one can fault a nonspecialist for accepting the date range adopted by INTF.
\n\u2022 Rather than appealing to P52 to rule out a second-century composition of John, appeal to other factors such as its reception in the second century to argue for its authenticity. This approach is admittedly more complex, but Charles E. Hill has written a thorough discussion of its second-century reception that would be helpful in formulating such an argument.<\/p>\n
Finally, remember that Christianity is a faith that stood unshaken for centuries while the earliest copies of its texts lay quietly buried in the sands of Egypt. People became Christians long before P52 was discovered, and they will continue to do so even if it is not quite as ancient as was previously thought.<\/p>\n
Key Takeaways
\n\u25ba Often a manuscript can be dated only by paleography, which is a difficult and imprecise way of assigning the date by means of an assessment of the handwriting.
\n\u25ba It is almost always unwise to assign a date range of fewer than fifty years on the basis of paleography; a range of seventy-five to one hundred years is typically more preferable.
\n\u25ba The middle year of a date range is no more likely (but also no less likely) to be the \u201cactual date\u201d of a manuscript than any other date in that range. Always try to give the full range; do not assume the earliest date is the right date.
\n\u25ba The responsible date range of P52, probably our earliest New Testament manuscript, is AD 100\u2013200, and a few scholars even extend this range into the 200s.<\/p>\n
CHAPTER SIX<\/p>\n
DATING MYTHS, PART TWO<\/p>\n
HOW LATER MANUSCRIPTS CAN BE BETTER MANUSCRIPTS<\/p>\n
Gregory R. Lanier<\/p>\n
Decades ago the introduction to the NA declared that we have moved out of an age of majuscules (nineteenth century) and papyri (early twentieth) and into an \u201cage of the minuscules.\u201d Yet for the pastor, student, commentary writer, and even scholar, the temptation remains to focus on the usual suspects (say, 01, 02, 03, 04, 05, P75), for with the discovery and accessibility of an embarrassment of riches in terms of post-ninth-century minuscule manuscripts has come a bewildering sense of staring into the abyss of manuscripts known primarily by their sterile numerical labels. Indeed, one of the key challenges of recent work on the text and transmission of the Greek New Testament is this: What on earth do we make of this \u201cfrighteningly large number of manuscripts\u201d?
\nOne common response\u2014in practice, if not also in theory\u2014is to treat these later manuscripts as a black box: that is, an indistinguishable mass of corrupt and secondary manuscripts that, due to their chronological distance from the original\/initial text, contribute nothing toward recovering it. The default setting of this \u201cdating myth\u201d runs thus: a later manuscript is worse because of its later date\u2014that is, that the length of time permits more stages of copying and corruption\u2014and an earlier manuscript is better because of its earlier date. If so, the thousands of later manuscripts en masse are \u201ccorrupt\u201d (seemingly the most common epithet) and useless, and can conveniently be ignored. A nearly opposite response, voiced by a small minority of scholars, is to treat most of these later manuscripts\u2014for a variety of reasons\u2014not as worse but rather better than the oft-touted but \u201ccorrupt\u201d majuscules and papyri.
\nThis apparent stalemate presents a conundrum for our text-critical apologetic, because both responses unintentionally undermine our confidence in the fidelity of the New Testament text and its transmission. If the former response is right, then can we really hang our hat on an argument from quantity (e.g., compared to the Iliad), since vast numbers of manuscripts that juice the statistics are deemed corrupt and useless? How do we explain the implications of the wildly corrupt state of the text that prevailed throughout most eras of the Christian world? Alternatively, if the latter response is right, then how do we account for the apparent early and widespread corruption of these ancient artifacts that, at least according to the Tischendorf\/Westcott-Hort tradition, get us closer in time to the era of the authors?
\nThis essay will attempt to put our text-critical apologetic on firmer footing by modifying the default settings in this way: sometimes later manuscripts are better manuscripts, though not always. To begin, we must clarify what it means to be \u201cbetter.\u201d<\/p>\n
DEFINING BETTER: WHAT ARE WE REALLY AFTER IN TERMS OF MANUSCRIPT VALUE<\/p>\n
When discussing the textual integrity of the Greek New Testament and the value of the more than five thousand Greek manuscripts (not to mention Latin and others)\u2014none of which perfectly contain the earliest form\u2014there is a need to be clear about what exactly we are aiming for. Put succinctly, what needs to be demonstrated in order to have a high degree of confidence that the wording of the New Testament has been faithfully transmitted from its point of origin throughout the world over the past millennia? Three central considerations come to mind.
\n(1) Tradition: Is the whole textual tradition, including its \u201clater\u201d stages, essentially stable? Is it even right (let alone helpful) to say that one branch of the tradition is hopelessly \u201ccorrupt\u201d and essentially useless compared to another? (2) Text: To what degree of confidence can we approach the reconstruction of the earliest text? Or, more specifically, how and when do later manuscripts provide us with readings that are earlier and more likely to be original, and how do they help us piece together the history of the text? (3) Quality controls: How do scribal features of later manuscripts shed light on the quality of the process as it unfolded over an incredibly long period of time? Are we dealing with scribes who were sloppy or, worse, mindlessly conforming to some standard imposed on them? Or are there reasons to believe that at least some later scribes were able \u201ctextual critics\u201d?
\nThe term better, then, is not merely a function of whether later manuscripts sometimes help in variant decisions\u2014though that is a big part of it\u2014but also a measure of the overall integrity of the entire process throughout the period of transmission. When positioned this way, later manuscripts often have much to contribute. We will thus proceed by examining the value of later manuscripts in forming an overall view on tradition, text, and quality controls.<\/p>\n
REFINING OUR APPROACH TO LATER MANUSCRIPTS<\/p>\n
Tradition: Appreciating the Byzantine witnesses. The most concrete manifestation of the later-is-worse \u201cdating myth\u201d is the long-standing divide over the Byzantine or Majority tradition. Broadly speaking, this is a form of the text found in the lion\u2019s share\u2014though not all\u2014of later manuscripts throughout the Christian East and West, a specific iteration of which attained near-universal status in the printing-press era in the form of the textus receptus. Since the overthrow of its hegemony in the 1800s (by Lachmann, Tischendorf, Westcott, Hort, and others), the Byzantine tradition has often been pronounced a \u201csecondary kind of text inferior in its value.\u201d Whether admitting it or not, most eclectic textual scholars who tend to value earlier manuscripts (and\/or internal considerations) over later manuscripts give \u201cno real role to the Byzantine text.\u201d
\nSpace does not permit a detailed analysis of the arguments against the Byzantine text-form, nor the recent resurgence of scholarly attempts to reassert a robust, principled defense of it (and shed prior inferior arguments). Rather, we will make four observations that bring better balance to the discussion and, for the purposes of text-critical apologetics, actually strengthen the entire enterprise regardless of one\u2019s overall text-critical method.
\nThe Byzantine tradition is not the same thing as the textus receptus. It is surprising how tenacious this conflation has been, even among scholars who should know better. Yes, the manuscripts used by Erasmus are \u201clate\u201d and broadly from the reservoir of witnesses labeled (in hindsight) \u201cByzantine.\u201d But today\u2019s reconstructed Byzantine text differs in nearly two thousand places from the textus receptus, which largely underlies the KJV. The problems behind the selection and editing of manuscripts underlying Erasmus\u2019s initial work and the ultimate textus receptus are well known, but the guilt by association that spills over into the a priori exclusion of the Byzantine text is unwarranted.
\nThe Byzantine text is not monolithic and thus cannot be so easily thrown out en masse. The very fact that there is a debate about the Byzantine text at all presupposes that there is something out there to which we can point and say, this is Byzantine and that is not. In other words, though Hermann von Soden showed long ago that there is diversity even in \u201cthe Byzantine text,\u201d there is a latent impression among scholars of all stripes that whatever we call the Byzantine text is essentially uniform or monolithic. This text is then regarded with suspicion (i.e., the product of ecclesiastical pressure to conform to a uniform and deficient standard text) or confidence (i.e., the product of near-perfect divine preservation of the text). In this, much of the debate pushes in the direction of rejecting it entirely or accepting it wholesale. However, in recent scholarship, including work done by the more rigorous proponents of Byzantine priority, there has been a renewed appreciation for the internal diversity within the Byzantine tradition itself.
\nFor instance, four of the key Byzantine witnesses for Acts and the Catholic Epistles (minuscules 18, 35, 319, 617) vary among themselves\u2014that is, contain more than one reading for a given unit of text\u20148 percent of the time, representing nearly one thousand textual units; for comparison, 01, 02, and 03 (key majuscules for these New Testament writings) vary among themselves 13 percent of the time. No one would, of course, suggest the latter group is monolithic and undifferentiated, so why the former when the overall variability profile is, all things considered, not that different? Moreover, scholars have long pointed out (though they perhaps have been ignored) that the scribes of manuscripts typically deemed \u201cByzantine\u201d in text-type were by no means uniform in procedure. This lack of homogeneity forces us, then, to reconsider whether its seat at the text-critical table should be revisited.
\nThe entire textual stream\u2014including the Byzantine tradition\u2014is far more stable than typically admitted. As mentioned above, certain scholars tend to privilege the earlier majuscules and papyri (or the so-called Alexandrian text-type) and pit the Byzantine tradition against it as largely corrupt and secondary; the reverse holds among Byzantine proponents. Only recently, however, have we been able to quantify rigorously the overall stability of the textual stream. For Acts and the Catholic Epistles (one-fifth of the Greek New Testament), 15 percent of the text is completely nonvariant among the hundreds of collated Greek witnesses, including the Byzantines; an additional 54 percent is likewise nonvariant among the most frequently cited sixteen majuscules and minuscules (including four Byzantines) spanning over one thousand years. Furthermore, the undivided Byzantine text fully agrees with the new Editio Critica Maior (ECM) text of Acts and the Catholics\u2014that is, the form of the text with which it should, as the theory goes, be diametrically opposed\u201494 percent of the time. For the Gospels, the rate of agreement between the NA and select majority readings is at least 86 percent based on sampling, though the number may adjust upward as fuller collation data becomes available. Put differently, the core tradition remains remarkably stable over time, in that the difference between the two texts usually thought to be most polarized is actually fairly small.
\nDistinctly Byzantine readings often have ancient roots. For Byzantine proponents, this is a tautology since Byzantine tradition is the most ancient one. But even among their opponents, research in the past several decades has demonstrated that hundreds of specific readings that have been classified at times as distinctly Byzantine are not, say, secondary conflations or corruptions arising from some later recension, but in fact are already attested by witnesses that are often a millennium older. For instance, P45, P46, and P66 share over one hundred readings with the Byzantine tradition against the early majuscules, and other chronologically earlier witnesses such as 02, 032, and some versions regularly contain what later become classified as Byzantine variants. In fact, the recent ECM editions for Acts and the Catholic Epistles accept Byzantine readings against the four major majuscules (01, 02, 03, 04) ten times. More tellingly, thirty-six of the fifty-two recent modifications to Acts in the ECM were specifically in the Byzantine direction (even if not exclusively Byzantine).
\nSuch findings have led several scholars to argue that what later becomes identifiable as \u201cByzantine\u201d developed progressively over time; that is, a multitude of Byzantine readings go back as far as the 200s, though the mature Byzantine tradition or text-form did not clearly solidify until the 900s. From a text-critical perspective, then, the fact that \u201cthe Byzantine [tradition] has preserved second-century tradition not preserved by the other text-types\u201d indicates that it should, at least sometimes, be treated as on par with other witnesses and not discarded to the \u201clate\u201d heap. Indeed, a growing chorus of scholars not otherwise part of the Byzantine-priority camp admit that \u201ctenacious negative bias against the Byzantine majority text\u201d is in need of a \u201creevaluation,\u201d such that the Byzantine tradition \u201cis an important witness to the early text.\u201d
\nIn sum, when one takes into consideration the entirety of the textual stream and the indications of its stability, one realizes that the long-standing competition between earlier and later manuscripts and the text-forms they represent is overblown: for the vast majority of the Greek New Testament\u2019s textual history, \u201cthe text is Alexandrian and Byzantine and every other text-type.\u201d In other words, the core textual tradition encompassed by any text-type is both very large and very stable. Granted, scholars will continue debating the major differences represented in the Byzantine tradition, such as the endings of Mark (Mk 16:9\u201320), the pericope adulterae (Jn 7:53\u20138:11), and miscellaneous readings of some length (Acts 8:37 and the like), but on the whole, such differences are the exception, not the rule. If nothing else, scholars on all sides should admit that \u201crejection en bloc of the \u2018Byzantine text\u2019 \u2026 tends to rob us of a most helpful instrument\u201d and that its uncritical replacement with a new textus receptus based on an earlier-is-better-and-later-is-worse \u201cdating myth\u201d needs to be reexamined. Klaus Wachtel sums up well: \u201cThe high agreement rates connecting these witnesses demonstrates that a large body of text was safely transmitted from the very beginning of its transmission through the Byzantine period to today.\u201d<\/p>\n
TEXT: APPRECIATING THE LATER CHANNELS OF AN ANCIENT STREAM<\/p>\n
The prior discussion on textual traditions more broadly leads to the important question: How can \u201clater\u201d manuscripts contain earlier and often high-quality texts if they are so far removed in time from the earliest sources? The common misconception is that manuscript copying always degrades over time\u2014much like the children\u2019s game of Telephone\u2014whereby later manuscripts arise from scribes who copy contemporary ones, which themselves are likewise late and corrupt, and make them worse. But is that the case? Here we will cast our net more broadly and consider how minuscules, including non-Byzantine ones, give us insight into how textual preservation can actually improve over time. We will work through two major considerations on this front and then draw some important conclusions about how \u201clate\u201d manuscripts can and should shore up our text-critical apologetic in terms of reconstructing the original\/initial text.
\nLater scribes on occasion self-consciously copied much earlier manuscripts. For many of the thousands of minuscules, we simply know very little about their origins, scribes, exemplars (= manuscript being copied), and so on. But in some cases, we know a bit more, and one thing we find is that, with reasonable frequency, a manuscript produced at a later date has been directly copied, either entirely or in portions (resulting in \u201cblock mixture\u201d), from a much older manuscript (which may now be lost) or an intermediate exemplar that itself derives from a much older manuscript. Let us consider a few examples.
\nThe famous minuscule 1739, which is one of the most frequently cited for Paul and the Catholic Epistles, was produced by a scribe named Ephraim in the mid-950s. After more than a century of research on 1739 and its scribe, the general consensus is the following. Ephraim was a quite good scribe whose \u201cvigilant insistence on exact reproduction\u201d extended even to reproducing the column formatting and holes or gaps of his exemplar. Further, the exemplar he copied was produced by a talented scribe or compiler who himself produced a \u201ctext-critical production\u201d that, as indicated in marginal notes, was based on two or more manuscripts that date at least to the 400s (maybe earlier), originate most likely in the library at Caesarea, and (at least for portions) reflect the text and\/or commentary of Origen. What makes this further significant is that Ephraim appears to have worked in Constantinople in a scriptorium that, as we know from other minuscules copied there, was producing largely \u201cByzantine\u201d output; Ephraim stands out, then, as one who cuts against the grain and conveys a text that is often dubbed \u201cAlexandrian\u201d despite its provenance in the heart of Byzantium. Furthermore, other minuscules share a similar text with 1739 and form a \u201cfamily\u201d (primarily 945 and 1891 but also possibly 206, 322, 323, 429, 453, 522, 630, 1704, and 2200), though a precise relationship between Ephraim\u2019s work and that of these other scribes is undetermined. In short, 1739 is a tenth-century manuscript that conveys a high-quality text found in manuscript(s) at least five centuries older and that, in turn, makes its way downstream to other minuscules:<\/p>\n
Figure 6.1. The genealogy of minuscule 1739<\/p>\n
Family 1 offers another example for the Gospels. Family 1 (f1 in most apparatuses) is an oft-studied group of minuscules that, at latest tally, includes at least 1, 22, 118, 131, 205, 209, 872, 1192, 1210, 1278, 1582, 2193, 2542. Of these minuscules, 1 (ca. 1100s) and 1582 (ca. 900s) are the most prominent, and several independent studies have shown persuasively that they are independent copies\u2014through one or more intermediate steps\u2014of a much older shared exemplar (often dubbed \u201cA-1\u201d) that is no longer extant. While the text contained in these manuscripts (and Family 1 as a whole) has much in common with other fully Byzantine witnesses, it also shares numerous affinities with various non-Byzantine forms of the text as well as church fathers, thus giving this family high visibility in textual criticism. Moreover, most likely the scribe of 1582 is the same Ephraim discussed above for 1739, and the ancestor of \u201cA-1\u201d goes back to at least the 500s, yielding the following picture:<\/p>\n
Figure 6.2. The genealogy of 1582<\/p>\n
The early-600s translation of the Greek New Testament known as the Harklean Syriac (syrh) also aptly illustrates the notion of a later textual tradition with much earlier roots. In 615\/6, Thomas of Harqel, the bishop of Mabbug and a participant in the Miaphysite controversy, completed a \u201cscholarly revision of the lost Philoxenian version\u201d of the Syriac New Testament. This Philoxenian version (syrph)\u2014completed by Polycarp under the commissioning of Philoxenius in 507\/8\u2014itself has earlier roots in both the late fourth-\/early fifth-century Syriac Peshitta (syrp) and the third-century Old Syriac (syrs or syrc). But Thomas was less interested in the Syriac tradition itself and more interested in repristinating it toward the Greek. In his Syriac colophons and various marginal notes, most of which have been preserved fairly well in downstream copies, Thomas indicates that he revised the Philoxenian by comparing it to what he considered \u201caccurate\u201d Greek manuscripts to which he had access\u2014two or three for the Gospels, two for Paul, one for Acts\/Catholics, and two for Revelation\u2014presumably from at least the 500s. Due to the pronounced way in which Thomas sought to conform the Syriac to his Greek exemplars, we can with some confidence back-translate the Harklean Syriac text (and marginal glosses) into Greek. From these reconstructions we know that the Greek text consulted by Thomas in the 600s is extremely close to that contained in minuscule 2138 (and its \u201cfamily,\u201d including at least ten other minuscules dating from the tenth to thirteenth centuries), which was produced by a scribe named Michael in 1072, likely at a scriptorium in Constantinople. In other words, the medieval scribes who produced the \u201cfamily 2138\u201d manuscripts had access to a form of the text that is accurately transmitted from manuscripts at least five centuries older and used by Thomas (who, in turn, was interacting with\u2014even going against\u2014Syriac texts with roots yet another three to four centuries earlier). The complex picture shows, in a pronounced way, how later manuscripts can convey earlier texts:<\/p>\n
Figure 6.3. The genealogy of 2138 and its close relatives<\/p>\n
A final illuminating but more tentative example is Family 13 (f13, also known as the Ferrar Group; nearly twenty minuscules and growing) and particularly one of its best representatives, minuscule 69. The Family 13 text has long been considered important in textual criticism, and while pinning down precise origins is difficult, the current hypothesis is that these minuscules share some sort of genealogical relationship with 038\/\u0398 (Koridethi, 800s), possibly 032\/W (Freer Gospel, 400s\u2013500s), and some form of the Greek text available to Origen. What is more certain is that the Family 13 minuscules at least share a pre-900s archetype (now lost); that the subsequent minuscules over the next several centuries show increasing assimilation toward a Byzantine form of text; but that the latest member of the family, 69 (Leicester Codex), has the most pristine form of text, deriving directly from the much earlier archetype rather than through intermediate copies. Most Family 13 members were copied in southern Italy, but 69 was copied by a very productive scribe named Emmanuel in the late 1400s in England, where he apparently had access to and directly used a much older manuscript. Though the picture is fuzzier than the other examples described above, it still contributes to our broader point:<\/p>\n
Figure 6.4. The genealogy of Family 13<\/p>\n
Later scribes on occasion correct their \u201clater\u201d exemplars using \u201cearlier\u201d manuscripts. As a general rule, \u201cscribes reproduced the manuscripts that were available to them\u201d; that is, they were primarily copyists seeking faithfully to render what they had in front of them, whether a contemporaneous exemplar or (as in the cases outlined above) an older one. However, even \u201clate\u201d copyists sometimes used other manuscripts to make revisions to their base exemplar either in the text itself or in the margins, and, in some cases, these other manuscripts are demonstrably much older. For instance, minuscule 424 (ca. 1000s) contains a text in the Epistles that has been corrected from a Byzantine form toward something close to 1739, whose compiler, in turn, not only worked from a much older exemplar (as argued above) but also corrects it against readings found centuries earlier in Irenaeus, Clement, Origen, Eusebius, and Basil. Minuscule 700 (ca. 1000s) shows the influence of numerous earlier forms of the text as well as an extremely rare variant of the Lukan Lord\u2019s Prayer that may be as early as the time of Marcion. Minuscule 2464 was produced in the same scriptorium as the Byzantine \u201cUspenski Gospels\u201d (minuscule 461), but it deviates so markedly at points from its peer that it is likely the scribe was also comparing to a different, and likely much older, exemplar. We could add more, but the point is this: while some scribes evince a tendency to work within a contemporary bubble, so to speak, some go the other way and edit a late exemplar by comparing it to, and integrating readings from, earlier manuscripts. This does not necessarily mean that the result is any closer to the original text, but it does illustrate another mechanism by which later manuscripts may contain earlier forms of text.
\nIn light of the above phenomena, we must maintain a distinction between the age of a manuscript (the specific artifact produced on papyrus or parchment by a scribe at a particular time and place) and the age of the text it contains. This distinction between an \u201cinternal age\u201d (text itself) and \u201cexternal age\u201d (physical artifact) has been established from at least the time of J. S. Semler (1765). The notion that \u201cthe more recent need not be the worse\u201d (recentiores non deteriores, formulated by Pasquali in the 1950s) is not new news, especially for those who subscribe to a Byzantine priority position for which this principle is utterly foundational. But among those who otherwise hold a form of the \u201clater is worse\u201d myth of manuscript dating, this principle is often eclipsed by the radiant light of the majuscules and papyri. An important corrective, then, is this: the text of a late manuscript\u2014which, when understood this way, is the true \u201cwitness\u201d for textual criticism\u2014\u201cmay be considerably older than the manuscript itself.\u201d In other words, if we think of the early stages of textual transmission as a river of witnesses in which the original\/initial text was carried along (with variation, of course), we clearly see signs of quite pristine channels at much later stages.
\nWhat value, then, do \u201clater\u201d channels offer to the task of reconstructing the original\/initial text? Several minuscules, at least for certain books of the New Testament, contain \u201ca valuable early text which can compete with even the best of the uncials,\u201d often providing significant later (downstream) corroboration of readings already known from earlier manuscripts. But even among lesser-quality minuscules, there is the possibility of finding a \u201cprecious stone\u201d hidden among the lesser pebbles of competing variants\u2014that is, a rare variant reading that deserves attention either for reconstructing the original\/initial or for shining light on interesting historical and theological factors (even if nonoriginal). Finally, minuscules contribute to fleshing out more fully the overall history of the transmission of the text. All these factors indicate that our text-critical apologetic should be careful not to be dismissive toward (or ignorant of) later minuscules just because they are \u201clate.\u201d<\/p>\n
QUALITY CONTROLS: APPRECIATING THE WORK OF THE LATER SCRIBES<\/p>\n
While the scribal habits and process controls of earlier scribes have been well documented, there have been few systematic analyses of later scribes, no doubt in part due to the sheer number of manuscripts. Filling this void are two competing and equally vague impressions: the later, medieval scribes were either incompetent and thoughtless in their work, or they were subject to a forced march to conform to the Byzantine standard. In either case the dating myth is perpetuated, as the later manuscripts produced by such later scribes are deemed all the worse. But are such impressions well grounded?
\nOne interesting way to get a read on this is to examine some of the most hotly debated \u201clater\u201d passages (particularly between eclectic critics and Byzantine advocates), where we might expect blind conformity or even sloppiness to arise. Is there any evidence that later scribes were exercising careful text-critical judgment even in these passages? Let us take a look at three examples.
\nThe pericope of the adulterous woman (Jn 7:53\u20138:11). We cannot, of course, get into the innumerable complexities here, but observe the following notes from \u201clater\u201d scribes:<\/p>\n
\u2022 565 (800s), at the end of John 21: \u201cThe chapter about the adulteress, not being present in the current copies, was omitted; it was located right after \u2018does not arise.\u2019 \u201d
\n\u2022 1 (1100s), at the end of John 21 (where it inserts the pericope adulterae): \u201cThe chapter about the adulteress: in the Gospel according to John, this does not appear in the majority of copies; nor is it commented upon by the divine fathers whose interpretations have been preserved\u2014specifically, by John Chrysostom and Cyril of Alexandria; nor is it taken up by Theodore of Mopuestia and the others. For this reason, it was not kept in the place where it is found in a few copies, at the beginning of the 86th chapter [according to Eusebius], following, \u2018Search and see that a prophet does not arise out of Galilee.\u2019 \u201d
\n\u2022 Mingana Syriac 480 (1700s): \u201cThis story is not found in all manuscripts. But Abba Mar Paule found it in one of the Alexandrian [NB: the city] manuscripts and translated it into Syriac as written here from the Gospel of John.\u201d
\n\u2022 Family 13: Several members of this minuscule family transfer the pericope adulterae to Luke 21:38, and in doing so they typically make some sort of scribal annotation or brief comment, some even showing awareness of the fact that other witnesses include it in John.<\/p>\n
Whether such scribal phenomena have a bearing on establishing the authenticity of pericope adulterae is a separate question. For our purposes the key point is this: some later scribes knew of the status of this pericope in other Greek manuscripts as well as church fathers and took such information into account in their work.
\nThe ending(s) of Mark (Mk 16:9\u201320). We also see similar scribal annotations at the ending of Mark, often immediately prior to this disputed passage.<\/p>\n
\u2022 1, 205, 209, 1582 (900s\u20131400s): \u201cIn some of the copies the Evangelist is completed to this point, as far as which Eusebius Pamphili also made his canons. But in many these are also present \u2026\u201d (or something similar)
\n\u2022 15, 22, 1110, 1192, 1210 (900s\u20131100s): \u201cIn some of the copies the Evangelist is completed to this point. But in many these are also present \u2026\u201d (or something similar)
\n\u2022 20, 215, 300 (1000s): \u201cFrom here to the end does not occur in some of the copies, but in the ancient copies it all occurs in full\u201d
\n\u2022 199 (1100s): \u201cIn some of the copies, this does not occur, but it stops here\u201d<\/p>\n
What we see here\u2014regardless of how much weight such comments should have in deciding on the original\/initial ending\u2014is a clear awareness by these scribes (or those of their archetypes) concerning competing readings in other manuscripts, which factored into their decision on how to handle the Markan ending(s).
\nSweat like drops of blood (Lk 22:43\u201344). Several minuscules of Family 13 treat this famous passage in quite an interesting way: some scribes omit it from Luke entirely, and nearly all move it to Matthew 26:39; but in either case, they always give some sort of scribal indication of the linkage between the two locations. This suggests the scribes in this family were aware of competing options among manuscripts available to them and were, for reasons unknown, making a \u201cconscious transfer of the verses from Luke to Matthew.\u201d They may have been wrong for doing so, but it was by no means random or accidental.
\nNo doubt some scribes were clumsier than others, and there was certainly a tendency toward increasing textual conformity in the medieval era. But these examples\u2014and we could include numerous others for lesscontentious passages\u2014indicate that at least some later scribes were by no means lazy, untrustworthy, or mindless. Rather, they appear far more aware as text critics than usually conceded, allowing their awareness of readings from \u201colder\u201d manuscripts, patristic readings, and even competing \u201clater\u201d readings from their own contemporaries to serve as process controls for their copying work. If so, then even if their final output contains nonoriginal\/noninitial textual readings, we can nevertheless have a higher degree of confidence in the essential conscientiousness of their efforts.<\/p>\n
A PRIMER ON \u201cBETTER\u201d MINUSCULES<\/p>\n
What remains is to provide some guidance on a select number of key \u201clater\u201d minuscules of above-average value that should always be a factor in studying the text of the New Testament. Before providing a brief primer, we must offer the caveat that, due in part to some of the scribal processes described above whereby a scribe might consult or correct against other manuscripts than their immediate exemplar, essentially every manuscript (or its ancestor, if it is a near identical copy) is to some lesser or greater degree \u201cmixed.\u201d Moreover, this mixture is not necessarily uniform across a manuscript, as a scribe might switch exemplars or consult a different manuscript for one epistle versus another, or even switch halfway through a Gospel, and so on. Due to such vicissitudes, some sections\u2014macro collections such as Gospels, or specific books\/epistles, or even subsets of chapters within books, and the like\u2014are \u201cbetter\u201d than others. Such block mixture means that the value of a given minuscule may vary based on where one is within it.
\nIn an effort to quantify this in a helpful way, we will describe the general tendency of each minuscule listed below using as shorthand the two opposed perspectives that dominate modern textual criticism, introduced above. Namely, does the text of the minuscule lean in the direction of the eclectic NA text that, on balance, reflects earlier majuscules and papyri ()? Does it lean in the direction of the Byzantine tradition ()? Or is it somewhere in between ()? In this way the table can serve everyone, regardless of one\u2019s macro philosophy of the text. We will break down the analysis for each major section of the New Testament, acknowledging the challenge of block mixture. Representative comments on each minuscule will round out this primer.<\/p>\n
MS
\nCentury
\nGospels
\nActs
\nCatholics
\nPaul
\nComments
\n1 (f)
\n12th
\no
\no
\no
\n\u201cClosely knit textual family \u2026 [texts] share a unique profile of Non-Majority text readings,\u201d including numerous singular readings
\n13 (f)
\n13th
\nLaunched the debates about a \u201cCaesarean\u201d text-type
\n33
\n9th
\n\u201cQueen of the cursives\u201d
\n69
\n15th
\nFamous for idiosyncratic handwriting; one of few minuscules used in S. P. Tregelles\u2019s edition
\n81
\n11th
\n\u201cOne of the most important of all minuscule manuscripts\u201d
\n579
\n13th
\nMost valuable in Mark\/Luke
\n700
\n11th
\n\u201cDeeply interesting manuscript,\u201d with a peculiar mix of variants
\n892
\n9th
\nArguably the best minuscule of the Gospels; first Greek-only manuscript with pericope adulterae
\n1175
\n11th
\nSame caliber as 33 and 1739 in portions
\n1241
\n12th
\n\u201cInteresting text\u201d but all over the board in terms of mixture
\n1739
\n10th
\nIn Paul, it is \u201ccomparable, in age and quality, to P46\u201d
\n1881
\n14th
\nOne of the most consistently cited minuscules in NA28
\n2344
\n11th
\nClosely related to 33 and 69, but varies per book
\n2464
\n9th
\nLeast valuable in Romans, but valuable elsewhere<\/p>\n
Table 6.1. A primer on the textual tendencies of important minuscule manuscripts<\/p>\n
CONCLUSION<\/p>\n
We began by articulating how the \u201cdating myth\u201d that assumes a \u201clater\u201d manuscript is worse because it is later\u2014and the opposite view that privileges later Byzantine manuscripts\u2014both introduce problems for our text-critical apologetic. In response, we have attempted to clarify the value of later manuscripts by examining the fidelity of the entire textual tradition (namely, restoring some appreciation for the Byzantine tradition while nevertheless acknowledging its shortcomings); illustrating how \u201clater\u201d manuscripts can have \u201cearlier\u201d texts via copying or correcting against older exemplars; and shoring up confidence in the scribal quality controls even in the medieval era. In short, with a well-defined meaning of better in place, we can say with confidence that later manuscripts can be better ones, though not always, of course.
\nThe net effect is that the unparalleled mass of minuscules\u2014though intimidating indeed\u2014deserves its day in court: \u201cExcept in von Soden\u2019s inaccurate and unused pages, the minuscules have never been allowed to speak. Once heard, they may well be found wanting, but at least their case will have been presented.\u201d Fortunately, the use of computers, indefatigable efforts by text critics around the world to digitize and collate minuscules, and the application of new approaches to the data that do not, at least in principle, prioritize external age over internal age (namely, the coherence-based genealogical method\/ECM efforts), are ensuring that \u201clater\u201d minuscules, even the inferior ones, are heard.<\/p>\n
Key Takeaways
\n\u25ba Although it is often thought that the later a manuscript is, the worse its textual quality will be, this is not always true. Sometimes later manuscripts can have better readings than some earlier manuscripts. Even our latest manuscript can preserve very early readings in isolation.
\n\u25ba The core textual tradition of the New Testament remains remarkably stable over time. The difference between the two texts usually thought to be most polarized is actually fairly small.
\n\u25ba Later scribes show evidence of conscientious work, seen in their choice of early manuscripts and in their choice to correct their manuscripts.<\/p>\n
CHAPTER SEVEN<\/p>\n
MYTHS ABOUT COPYISTS<\/p>\n
THE SCRIBES WHO COPIED OUR EARLIEST MANUSCRIPTS<\/p>\n
Zachary J. Cole<\/p>\n
One of the trickiest questions about the transmission of the New Testament text is this: in the earliest period, before the time of the Roman emperor Constantine (r. AD 306\u2013337), who exactly was doing the copying? Many books about textual criticism offer an answer that goes something like this:<\/p>\n
The copying practices we have considered thus far have been principally those of the first three centuries of Christianity, when most of the copyists of the Christian texts were not professionals trained for the job but simply literate Christians of this or that congregation, able to read and write and so called upon to reproduce the texts of the community in their spare time. Because they were not highly trained to perform this kind of work, they were more prone to make mistakes than professional scribes would have been.<\/p>\n
This quotation is from Bart Ehrman\u2019s now-famous Misquoting Jesus. Later on, he reiterates, \u201cIn the early Christian centuries, scribes were amateurs and as such were more inclined to alter the texts they copied\u2014or more prone to alter them accidentally\u2014than were scribes in the later periods who, starting in the fourth century, began to be professionals.\u201d Forget the familiar image of medieval monks hunched over writing desks. In the early period, we are told, Christian manuscripts were copied by zealous but passably literate amateurs rather than \u201cprofessionals.\u201d Ehrman is not the only one to hold this view. In fact, many introductions to the subject seem to agree that the first copyists of the New Testament were insufficiently untrained novices who corrupted the text through their incompetent work and outright tampering with the text. This understanding bears serious implications about the reliability of the text. If the earliest New Testament copyists were energetic but inexperienced, how do we know that the text was not corrupted beyond recovery in this formative stage? This is a sobering thought.
\nBut, strangely enough, other voices sing a different tune. Some scholars assert the exact opposite about early scribes. The impression one gets from reading these scholars is that most early Christian copyists were\u2014on the contrary\u2014actually professional, careful, scrupulous, and well trained, and they studiously avoided making any changes to the text. Much like their Jewish counterparts who painstakingly copied the Hebrew text of the Old Testament with reverence and attention to the minutest detail, so also many early Christian copyists transmitted the text believing it to be the very word of God. After all, the earliest Christians were themselves Jewish, and it would be natural for them to copy New Testament books with the same care as they would Old Testament books.
\nSo which is it? Of these two different portraits of early Christian copyists, which is the right one? Were the early New Testament copyists mostly zealous amateurs, or were they careful professionals? Or were they somewhere in between? In this chapter we will attempt to dispel a few myths about the early scribes, their training, and their handling of the New Testament text. We will take as our focus the first three centuries of Christianity, the era prior to the emperor Constantine, in order to evaluate what the evidence tells us about the early copyists. To do this we will first examine the competency of early scribes: What do we know about their training or lack thereof? Second, we will examine their attitudes toward the text: Did they faithfully copy it, or did they rewrite it whenever and however they wanted?<\/p>\n
ZEALOUS AMATEURS? THE TRAINING OF EARLY CHRISTIAN SCRIBES<\/p>\n
Professionals or not? We begin with the argument that the earliest scribes were zealous but incompetent. Why is this idea popular? One reason goes back to the work of a well-known and important scholar of early Christianity, Colin H. Roberts. In a series of lectures delivered in 1977, Roberts argued that there is a noticeable similarity between early Christian manuscripts and another body of ancient texts known as \u201cdocumentary papyri.\u201d Broadly speaking, documentary papyri (often called just \u201cdocuments\u201d) are the mundane texts left behind by ancient people: personal letters, tax receipts, contracts, grocery lists, business memoranda, and other kinds of functional pieces of writing. Because these were practical documents, the scribes who produced them usually used handwriting that was unpretentious, often quickly written, and sometimes plainly unattractive. This writing is often \u201ccursive\u201d in the sense that letters are joined together for the sake of speed and convenience. The documentary style of writing stands in contrast with the slowly written, carefully executed letters often used in the copying of books of literature, namely, \u201cliterary hands.\u201d Expensive and luxurious copies of classics such as Homer\u2019s Iliad and Odyssey were often written in the elegant and consistent calligraphy of a literary hand rather than the more ordinary documentary hand. The way that I might scribble down a grocery list on my way to the store will no doubt pale in comparison to the fancy calligraphy found on an opulent wedding invitation.
\nThere can be quite a bit of overlap between literary and documentary handwriting, as they were not rigid and hermetically sealed styles. Some ancient books were written in informal, unpretentious hands, and at the same time some documents were written carefully and slowly. In any case, what Roberts argues is that the earliest manuscripts of the New Testament have more in common with documentary papyri\u2014the practical, everyday texts\u2014rather than literary papyri. In particular, Roberts observes that early Christian manuscripts are more often in codex form rather than roll form, they use abbreviations in the text, their scripts are unpretentious, and they divide the text with punctuation and paragraphing\u2014all of which are features also found in documentary texts. His conclusion: the earliest Christian copyists were trained to make documents, not literature.
\nHere is Roberts himself on the matter:<\/p>\n
What I think they all [early Christian papyri], in varying degrees, have in common is that, though the writing is far from unskilled, they are the work of men not trained in calligraphy and so not accustomed to writing books, though they were familiar with them; they employ what is basically a documentary hand but at the same time they are aware that it is a book, not a document on which they are engaged. They are not personal or private hands; in most a degree of regularity and of clarity is aimed at and achieved.\u2026 In none can be traced the work of the professional calligrapher or the rapid, informal hand of the private scholar.<\/p>\n
So far, so good. For the most part, scholars today are in agreement with Roberts\u2019s assessment. Unfortunately, however, many today misunderstand his actual argument. It is often assumed that the unpretentious scripts of early New Testament papyri indicate an unprofessional standard of copying. That is, unattractive and unpretentious writing must mean \u201cuntrained\u201d scribes and poor standards of transcription. But this is not actually what Roberts argues, nor is it true.
\nIt is simply a mistake to assume that unattractive and workaday handwriting entails a lack of training or a poor standard of copying. Roberts does not claim this, nor is it demonstrable from any other study. Style of script is one thing; accuracy in transcription is another. In fact, there is evidence that suggests attractive, literary handwriting often entails inaccurate copying. For example, in describing a \u201cdeluxe\u201d copy of Virgil, Roberts himself comments, \u201cThe text itself was both carelessly written and (as is not infrequently the case with \u00e9ditions de luxe [deluxe editions]) of poor quality.\u201d Carefully note what Roberts says here: the pretty letters and de luxe presentation of this book did not guarantee a high level of care in transcription\u2014and in fact this \u201cis not infrequently the case.\u201d Roberts is not alone here; we can see from the comments of many ancient authors that book shops and other professional copying services were frequently unreliable because of their poor standard of copying.
\nIn reality, there is no evidence to assume a necessary connection between calligraphy and care in transcription\u2014these are separate issues. The textual quality of a manuscript must be determined on its own terms, not on the erroneous assumption that the beauty of its script is the most relevant factor. To return to Roberts\u2019s argument about Christians and documentary papyri, he simply makes the point that early Christian copyists appear to have been trained to copy documents rather than works of literature\u2014yet trained they were! Scribes who copied documentary papyri were indeed still scribes, and they needed requisite training, competency, and ability to do so. Documentary papyri are not inherently substandard; they serve a different purpose from copies of literature. We should not make the mistake of thinking that works of literature somehow required a higher level of accuracy than did documents. Just ask the business owner whether she needs accurately copied contracts, or the merchant whether he wants accurate receipts, or indeed the government whether it requires accurate tax records.
\nTo put it bluntly, Roberts never argues that early Christian scribes were unprofessional in the sense of untrained or inexperienced. For this reason, I think the term professional can be anachronistic and ultimately misleading. When discussions of early scribes rely on this term, they tend to conjure modern understandings of occupation and vocation. But these modern ideas are likely to distort our perception of ancient copying. Rather than asking whether Christian scribes were \u201cprofessionals,\u201d I prefer to ask whether they were competent and trained.
\nCompetent and trained? Recent studies. Regardless of the attractiveness of early Christian manuscripts, what does the evidence suggest about the training and competency of the scribes themselves? Fortunately, this question can be answered with a great deal of certainty. This is because one simply needs to examine the earliest extant manuscripts to observe the quality of their work. In fact, many scholars have recently done this kind of analysis with a view to discovering the level of training and ability of early Christian scribes. Put simply, what they have found is that most of the early Christian manuscripts are clearly the products of trained and competent copyists, not zealous amateurs.
\nTwo such studies are worth highlighting here. The first is the work of Kim Haines-Eitzen. In her monograph Guardians of Letters, Haines-Eitzen asks, \u201cWho were the scribes who copied early Christian literature during the second and third centuries?\u201d To answer this question, she examines the evidence of early Christian scribes found in ancient literature, archaeology, and the papyri themselves. One of her observations is key:<\/p>\n
What is striking about our earliest Christian papyri is that they all exhibit the influences of literary and documentary styles, and they all seem to be located in the middle of the spectrum of experience and level of skill. The scribes who produced these copies fit well into the portrait of multifunctional scribes\u2014both professional and nonprofessional\u2014whose education entailed learning how to write a semicursive style.<\/p>\n
Note carefully what she argues here. All of the earliest Christian papyri evidence a mix of documentary and literary influence. This means that early Christian papyri, far from looking like the work of \u201camateurs,\u201d actually bear the characteristics of both documentary papyri and literary manuscripts and thus can be located in \u201cthe middle of the spectrum of experience and level of skill.\u201d Finally, these papyri exhibit the work of \u201cmultifunctional\u201d scribes. Haines-Eitzen uses the term multifunctional to describe copyists who were able to produce both documentary texts and works of literature, not simply one or the other.
\nOnce again, we meet the terminology of professional versus nonprofessional, and this can lead to confusion. So we should be clear. For Haines-Eitzen the term professional simply indicates someone who was by trade a scribe or held a scribal title (such as the administrative scribes in Egypt). Importantly, the term is not synonymous with \u201ccompetent\u201d (in fact, we have evidence of professional yet clearly incompetent scribes!). Many copyists in the ancient world were slaves, servants, or freedpersons who were trained and fully able to copy texts accurately, even though they may not have been paid for it, and it may not have been their primary task. These individuals often filled the roles of \u201csecretaries, clerks, stenographers, and record keepers\u201d in large households, but they would not necessarily be scribes by trade or by title, or considered \u201cprofessional\u201d in the modern sense. Thus, one of Haines-Eitzen\u2019s key findings is that early Christian literature was circulated and disseminated through informal channels formed on social networks, or \u201cin-house.\u201d When seeking to obtain a copy of a New Testament book, early Christians probably did not commission the \u201cprofessional\u201d scribes of the book shops for their services; rather, they more likely enlisted the help of individuals within their own social circles. But this does not in any way imply that these individuals were insufficiently trained or that they were unskilled amateurs. In fact, the opposite appears to have been the case, and she states this quite clearly:<\/p>\n
The fact that Christian papyri (as well as many classical papyri more generally) all exhibit the influences of documentary and literary styles indicates scribes who were either comfortable with and experienced in both styles or trained in more general styles of writing that could be adapted in rather simple ways to different tasks; it seems to me that the latter scenario is more likely since had these scribes had extensive training in literary book hand, their hands would have manifested this training.<\/p>\n
In other words, while it is not so clear that many early Christian copyists were professional scribes by trade, there is no doubt that most were competent and experienced.
\nThere is another, more recent study that comes to a similar conclusion. In 2016, Alan Mugridge published an exhaustive analysis of the physical form of early Christian papyri, observing features such as scribal hand, writing material, dimensions, letters per line, lines per page, punctuation, critical signs, and so on. He too comes to the conclusion that most early Christian transcribers were able copyists: \u201cIt seems clear that the vast majority of the Christian papyri were copied by trained scribes.\u201d He goes on:<\/p>\n
All of this should cast doubt on the view that on the whole Christian manuscripts were copied by unskilled writers during the early centuries, and also suggests that we need to re-examine any implications drawn from this view that the transmission of Christian texts was quite inaccurate. It is true that, as Roberts proposed, many of the papyri show the hand of scribes accustomed to producing documents. Nevertheless, they still exhibit the skill of the trained scribes who produced such documents, including a number of papyri copied to a calligraphic standard, even though we do not know whether they were paid for their efforts or did their work voluntarily.<\/p>\n
Here Mugridge generally confirms the argument of Haines-Eitzen and shows that the evidence of the manuscripts themselves exposes a misperception about the early scribes. While it remains true that, in many ways, Christian papyri resemble documentary papyri (as they use the codex book form and abbreviations), it is important to stress that this does not equate to \u201cunprofessional\u201d work in the sense of untrained or inexperienced. As a group, early New Testament manuscripts show the same levels of care, experience, and accuracy that one could reasonably expect of any ancient text. So, to describe the earliest Christian copyists as \u201cnonprofessionals\u201d is misleading because it conjures an impression of the evidence that is warped by modern ideas. While it may well be true that they were not necessarily scribes by trade or by title (although some probably were), it is quite clear that they were nonetheless capable of high-quality work.
\nSome examples. At this point, it is worth considering a few specific examples. We can examine, for instance, the manuscript called P45\u2014a papyrus copy of the canonical Gospels and Acts that dates to the third century AD. Thirty folios (or sixty pages, counting front and back) remain of an original 220 folios or so (or 440 pages). The scholar who first edited the manuscript describes its handwriting in this way: \u201cThe codex is written throughout in a small and very clear hand.\u2026 The writing is very correct, and though without calligraphic pretensions, is the work of a competent scribe.\u201d G\u00fcnther Zuntz published one folio from this same codex and describes it in similar terms:<\/p>\n
His hand is on the whole amazingly even, and his practice with regard to orthography, punctuation and the use of nomina sacra astonishingly consistent. The beginnings of lines are placed one below the other with great regularity so as to effect, from top to bottom, an almost faultless, straight line running parallel with the fibres of the papyrus.\u2026 The scribe aims at neatness rather than rigid uniformity.<\/p>\n
Far from the scribblings of an amateur, P45 preserves the work of a consistent and practiced scribe. Another example is P46, which is a collection of Paul\u2019s letters that dates to the third century AD. Eighty-six folios (or 172 pages) remain of this manuscript. Its editor compares its handwriting to that of P45 (mentioned above):<\/p>\n
The script of the papyrus is in marked contrast with that of the Chester Beatty papyrus of the Gospels and Acts [P45]. It is far more calligraphic in character, a rather large, free, and flowing hand with some pretensions to style and elegance. It is upright and square in formation, and well spaced out both between the letters and between the lines.\u2026 In general it may be said that the letters are rather early in style and of good Roman formation.<\/p>\n
Around the same time, Henry Sanders agreed with this assessment, stating, \u201cThe writing is of the book hand type and the letters are carefully formed and well spaced.\u201d This characterization can hardly describe the work of an unskilled transcriber.
\nA third and final example is P75, a substantial copy of Luke and John that dates to the third century AD. Fifty-one folios (or 102 pages) now remain of an original seventy-two. Like the other texts we have seen, it was written in a \u201cclear and generally carefully executed uncial.\u201d More recently, Sarah Edwards comments, \u201cThe manuscript was written on papyrus of such fine quality that the verso is generally as smooth as the recto. In texture it resembles hand-woven linen.\u2026 The codex is far more beautiful than the photographs reveal. If I were to describe it with one word, that word would be \u2018elegant!\u2019 \u201d Scholars widely recognize the high-quality work found in P75, and there is not the slightest suspicion it is the work of a zealous but untrained Christian.
\nMany more examples could be discussed here, but the point should be clear. When we look at the early manuscripts themselves, what we find is that they clearly are not the work of incompetent fanatics. Rather, most of them bear the marks of trained and capable scribes who blended techniques of documentary and literary writing styles. So much for \u201camateurs.\u201d
\nBut we also do not want to overstate the case. Not all of the early Christian papyri we have today can be described in such positive terms. Some of them are indeed sloppily written, carelessly produced, and evidently written by untrained individuals. One good example of this is P72, a third-century copy of 1-2 Peter and Jude. The handwriting of this manuscript is noticeably rough and unclear. Scholars believe that this manuscript was intended for private use (rather than for congregational worship). Another example is P47, a third-century papyrus copy of Revelation. Although its handwriting is readable, the script is somewhat sloppy and inconsistent with things such as letter size. The most recent study of this manuscript puts it bluntly: \u201cWe are not dealing with a highly skilled calligraphic hand.\u201d
\nWe need to be wary of painting with brushstrokes that are too broad. Just as it is inaccurate to say that all early copyists were untrained amateurs, it is likewise inaccurate to say that they were all experts. Rather, among the early manuscripts we find a wide range of scribal skills and abilities, but still a significant majority appear to be competent transcribers.
\nThe length of New Testament manuscripts. A final point is worth thinking about before we move on. One of the ways that \u201camateur\u201d copyists are identified is by their inability to maintain a consistent script for extended periods of time. That is, someone who is not trained or sufficiently accustomed to the task of copying a lengthy text will gradually lose the ability to write legibly and consistently. In a fascinating study, Roger Bagnall and Raffaella Cribiore examine more than three hundred ancient letters that were written by women. Their purpose is to learn as much as they can about ancient women from the physical remains of these letters. Interestingly, they note that fatigue is one of the clear indications of a \u201cpersonal hand;\u201d that is, the person making the document was untrained in transcription: \u201cIt happens very often, however, that personal hands produced acceptable results at first but are then betrayed by the length of a text. In contrast to scribal hands, which seem to become more impatient and faster toward the end, personal hands are often overcome by the effort, just as are school hands.\u201d
\nWe have all seen this in our own experience: try as we might to maintain some consistency, the fact is that legibility usually lessens as the amount of writing increases. This was evidently true with ancient writers as well. If we consider the length of early Christian manuscripts, it is remarkable how consistent the scripts are. As noted above, many of our early manuscripts are extremely lengthy: recall that P45 originally contained around 220 folios, P46 originally contained around 104 folios, and P75 originally contained around 72 folios. What we see in the surviving pages of these manuscripts is a consistent clarity of script. This is a testament to the capability and skill of our earliest transcribers. Moreover, this fact is all the more significant when we consider how short most ancient letters were compared to New Testament books: the average ancient letter was less than one page in length. The prolonged consistency of the hands found in most New Testament manuscripts reflects the work of scribes who were able to maintain legible and consistent scripts for impressive stretches. What this fact demonstrates is experience in transcription.
\nThus far we have made some important clarifications about early Christian scribes: they were not all \u201camateurs\u201d and \u201cnonprofessionals.\u201d In fact, the evidence we have suggests that most of the early copyists were competent, trained individuals who were more than capable of doing the work of transcription\u2014whether or not they were scribes by trade or by title. It is easy to assume that, prior to Constantine, Christians must have been deprived, short-staffed, and without access to capable scribes. This idea seems to fit with our understanding of the fledgling church in that era. Nevertheless, an investigation of the actual manuscripts reveals a more nuanced picture that involves scribes of a variety of skills and training, but definitely including many who were capable. When we return once again to Roberts and his original comments and set aside the term professional, the point can be seen quite clearly: \u201cThe Christian manuscripts of the second century, although not reaching a high standard of calligraphy, generally exhibit a competent style of writing which has been called \u2018reformed documentary\u2019 and which is likely to be the work of experienced scribes, whether Christian or not.\u201d
\nA note on evidence. Before moving on, it is worth admitting what we do not know. Specifically, we do not have the same amount of evidence for the earliest period of transmission as we would like. Our material evidence reaches back well into the third century, but for the second century it is extremely fragmentary, and for the first century it is (at present) nonexistent. We cannot make firm judgments about the training of scribes whose work no longer remains. This fact is worth bearing in mind. Even still, there is little reason to suggest that what we would find in the late first\/early second centuries would differ greatly from what we see in the late second\/early third centuries.<\/p>\n
WILD AND FREE? EARLY SCRIBAL FIDELITY<\/p>\n
Now we must turn to our final question. Experienced or not, were the earliest scribes in the habit of changing the text whenever they wanted? Even if we were to establish the competency of early scribes, this does not mean that they always copied with care. Here we must be careful to distinguish ability and attitude. A scribe might have been capable of producing the most luxurious and accurate of manuscripts yet at the same time display a casual or lax attitude toward the text. Such a situation was in fact hinted at above when we noted Roberts\u2019s comments about Greco-Roman \u00e9ditions de luxe\u2014they often are pretty but penned by inattentive scribes. So even if we know that most early Christian scribes were trained and capable, as I have argued above, we still need to consider whether they have been faithful to the text they received.
\nTo anticipate my conclusion, I argue that it is very unlikely that the text has been corrupted beyond retrieval by freewheeling scribes. While there may be instances of some \u201cfree\u201d and perhaps wild copying in the early period, we have good reason to believe that original words of the apostles have been faithfully preserved. I offer three points that support this view. First, due to the wealth of our evidence, we have a paradoxical safety in numbers. Second, we have strong evidence that the text has been carefully transmitted in at least one stream of manuscripts. Third, when we look at the nature of our textual variants, they do not appear to be the kind of variants that would arise if scribes were in the habit of writing whatever they wanted.
\nThe wealth of manuscripts. The first point to consider is the sheer mass of manuscripts we have. Having thousands of New Testament manuscripts helps to mitigate\u2014at least to a certain degree\u2014the effects of those who might tamper with the text. If we had only a handful of New Testament manuscripts, it would be extremely challenging to identify places where the text had been corrupted. But this is not the case; thousands of New Testament manuscripts have come down to us, and they are continually being discovered. This embarrassment of riches poses some challenges for us, but one of its benefits is that errors in transmission can be more easily identified and removed.
\nDavid Instone-Brewer describes this idea as a \u201cparadoxical safety in numbers.\u201d What he means is that the immense volume of material evidence for the New Testament text entails two important facts. On the one hand, having more manuscripts means having more textual variants; this is simple mathematics. All handwritten documents will have errors in them. On the other hand, more manuscripts at the same time permits more textual stability in the long run. As Instone-Brewer describes it,<\/p>\n
Imagine that two pupils wrote down their teacher\u2019s dictation; we could find both contained some errors but it would often be impossible to decide which of the two versions was correct. Now imagine that a class of thirty pupils wrote down the same dictation; there would be many more errors in total, but the correct version would be much easier to work out because only one or two would make the same error. With so many New Testament manuscripts we therefore have a strong foundation for discovering the original wording.<\/p>\n
This is a helpful analogy. We should not be too concerned about the danger that one sleepy or perhaps heretical scribe has hopelessly corrupted the text for all who followed. Our knowledge of the text of the New Testament is not based on the witness of one or two manuscripts but rather a great many that come from different times, different regions, and even different languages. To be sure, we must keep in mind that it would be misleading to suggest we have thousands of manuscripts that date to the first few centuries of Christianity; in fact, we do not. Rather, from this early period we have about 125 Greek manuscripts of the New Testament. While this may not sound impressive at first, comparison with any other ancient book will demonstrate how this constitutes a richness of evidence.
\nAn example might be beneficial here. Many English Bibles skip John 5:4, going straight from John 5:3 to John 5:5. This is more than simply a mistake in numbering; it signals a text-critical problem. What is missing in the elusive John 5:4? The story is about Jesus healing the lame man at the pool of Bethesda, and John 5:4 explains why the sick and lame were at the pool in the first place: \u201cFor an angel of the Lord went down at certain seasons into the pool, and stirred up the water; whoever stepped in first after the stirring of the water was made well from whatever disease that person had\u201d (NRSV). What is significant about this variant is that our knowledge of early manuscripts helps us to see how the verse came to be added later. The earliest Greek manuscript to contain John 5:4 is 02 (Codex Alexandrinus), an important manuscript that dates to the fifth century AD. However, the discovery of several earlier manuscripts has shown that the text of John in the third and fourth centuries did not have this verse. Four manuscripts in particular show no sign of this verse: P66, P75, 01, and 03. These manuscripts go straight from John 5:3 to John 5:5. With this early textual evidence, we can see that John 5:4 appears to be an explanatory gloss by a later scribe who tried to add some detail that John appeared to leave out. The point is this: scholars of the New Testament are enormously privileged to have not just one or two but several dozens of manuscripts that date to the earliest centuries of Christianity. This wealth of evidence significantly increases the likelihood that textual aberrations will be identified and removed.
\nThe evidence of careful transmission. The second point to consider relates to evidence that the text has been copied carefully. There is no disputing the fact that many New Testament manuscripts contain corruptions and rewritings. This is simply a fact of all handwritten texts. But it is nonsense to suggest that every manuscript is hopelessly corrupt or that the original text is beyond hope of recovery. Text critics widely recognize that many New Testament manuscripts evidence a conscious effort to produce a reliable, accurate text. Some of the best-known manuscripts fit this description, for example, Codex Sinaiticus (01), Codex Vaticanus (03), and P75, but there are many others. Scholars have long recognized that such manuscripts contain high-quality texts and are witnesses of careful, accurate copying of the New Testament. Historically, scholars have seen these manuscripts as representatives of an identifiable textual stream flowing from the ancient Egyptian city of Alexandria, which contrasts with the \u201cWestern\u201d tradition, thought to be a separate stream of manuscripts characterized by a freer attitude toward the text. In this regard, we can allow Bruce Metzger and Bart Ehrman to explain:<\/p>\n
It would be a mistake to think that the uncontrolled copying practices that led to the formation of the Western textual tradition were followed everywhere that texts were produced in the Roman Empire. In particular, there is solid evidence that in at least one major see of early Christendom, the city of Alexandria, there was conscious and conscientious control exercised in the copying of the books of the New Testament.<\/p>\n
Scholars now tend to avoid speaking of textual \u201ctraditions\u201d and \u201ctext-types\u201d as such, but this quotation nevertheless captures an important aspect of the transmission of the text. It is not simply evangelicals who recognize 01, 03, and P75 to be accurately copied texts; this is the majority view among text critics across the theological spectrum. This fact confirms for us that while there may have been early copyists who took liberties with the text that they received, the same cannot be said of all or even most of these early transcribers. It is furthermore significant that many of our earliest witnesses illustrate this tendency toward conscientious transcription. For example, many early papyri, such as P1, P4, P64+67, and P77+P103, show strong textual affinities with Sinaiticus and Vaticanus. This fact is significant for several reasons, but one worth noting here is that we are able to demonstrate a practice of careful copying over time. That is, scholars are able to track a consistent line of careful copying practices from the second century, with P4 and P64+67, to the third century, with P66 and P75, all the way to the fourth century, with 01 and 03, and beyond, all of which cohere closely in their wording. When these early witnesses all agree in their wording, this allows us to have a high degree of confidence about the state of the text in even earlier stages of transmission.
\nThe nature of our textual variants. The third consideration to make relates to the nature of the textual variants we see in the New Testament. If we ask what sorts of changes and alterations we find among our manuscripts of the New Testament, the answer is illuminating. Supposing that scribes were in the habit of rewriting verses, adding stories, deleting verses, and writing whatever they wanted, we should expect to see a certain kind of variation in our manuscripts that demonstrates such was the case. We might expect to see, for instance, added paragraphs or chapters and multiple versions of books. But what is striking is that these sorts of changes are noticeably absent from the manuscript record. That is, the kind of textual variation we see in manuscripts affects units of text that are the length of a verse and smaller.
\nThere are, of course, some exceptions to this rule. For example, two units of twelve verses each, in Mark 16:9\u201320 and John 7:53\u20138:11, are in doubt. Also, there is a \u201cWestern\u201d text of Acts that is a bit longer than the received form, but not because of an added scene or story but because of minor additions within existing verses made throughout. The Gospels and Acts in particular, therefore, demonstrate that while scribes did change the texts they copied, the nature of those changes is significant. As Michael W. Holmes describes,<\/p>\n
In short, a very high percentage of the variation evident in the text of the Four Gospels and Acts affects a verse or less of the text. On this level, the fluidity of wording within a verse, sentence, or paragraph is sometimes remarkable. At the same time, however, in terms of overall structure, arrangement, and content, these five documents are remarkably stable. They display simultaneously, in other words, what one may term microlevel fluidity and macrolevel stability.<\/p>\n
This is a useful framework with which to work: microlevel fluidity and macrolevel stability. Using this terminology helps to situate the particular task of the New Testament textual critic. For the most part, our concern is not with major portions of text\u2014their (in) authenticity, rewriting, and\/or rearrangement\u2014but rather at the level of the verse and smaller.
\nHowever, this stands in contrast with the task of many who study other ancient texts, as these are often confronted with comparatively major differences between witnesses. For example, Homer\u2019s Iliad and Odyssey are well-attested works in terms of surviving manuscripts (especially the Iliad). However, both texts contain many lengthy passages that are textually uncertain, ranging from single lines to whole paragraphs and to the entirety of a book. A similar example is the early Christian text the Shepherd of Hermas. This work appears to have been written and circulated as two separate books, which were later combined and supplemented with other passages. For some other ancient books, many portions are simply lost to us. For example, first-century Jewish historian Josephus wrote a number of works that have been preserved, but only imperfectly. Peter Head notes that we have 134 manuscripts with portions of the Greek text of Josephus\u2019s works, but in many places these are incomplete, leaving several passages with no extant text at all (though some can be supplemented by surviving Latin translations).
\nSo if indeed the early scribes of the New Testament were in the habit of taking liberties with the text, such activity has been restricted mostly to the level of the verse and smaller. This fact does not by any means solve the many problems that textual critics face in reconstructing the original wording, but it does simplify the task enormously. Rather than worrying about major structural differences and missing portions of text, New Testament scholars work with a body of literature that is stable in its macrostructure and more fluid in its details.<\/p>\n
CONCLUSION<\/p>\n
In this chapter we have addressed several myths about early scribes and their work. First, it is a myth that the early Christian copyists were all untrained amateurs. Second, it is also a myth that early Christian copyists were all rigorously schooled in the tradition of Jewish scribes. The reality is more complicated: from the evidence of their existing work (the manuscripts themselves), we see that the majority of early Christian copyists were competent and experienced scribes who were more than capable of doing the job of transcription, even if they were not producing luxurious copies of literature. Third, it is a myth that Christian scribes changed the text whenever they saw fit. The kinds of variants that we see in the textual record do not reflect this tendency, and the high number of reliable manuscripts available to us leads most scholars to agree that we have a very good idea of what the apostles originally wrote.<\/p>\n
Key Takeaways
\n\u25ba The earliest copyists of New Testament manuscripts were neither careless amateurs nor professionals with Xerox-machine accuracy. From the quality of the handwriting, we find a wide range of scribal skills and abilities among the early manuscripts, but a majority appear to be competent transcribers.
\n\u25ba As a group, early New Testament manuscripts show the same levels of care, experience, and accuracy that one could reasonably expect of any ancient text. When compared with carefully copied later manuscripts, the earliest scribes do not appear overly careless, as they are often described.
\n\u25ba The macrostructure of the New Testament is remarkably stable, especially in comparison to other ancient works. Though textual variation exists, it is usually at the micro rather than macro level.<\/p>\n
CHAPTER EIGHT<\/p>\n
MYTHS ABOUT COPYING<\/p>\n
THE MISTAKES AND CORRECTIONS SCRIBES MADE<\/p>\n
Peter Malik<\/p>\n
The manual copying of texts was a laborious, multifaceted task that involved several simultaneous mental and mechanical processes. The basic process of transmitting the wording from the exemplar onto the empty leaf of papyrus, parchment, or paper involved much more than merely \u201creading\u201d and \u201cwriting.\u201d The tedious process of copying entailed many concomitant tasks. To name but a few, scribes had to heed the text layout on every page and determine the proper word division; they had to check for and correct possible errors in the exemplar as well as resolve what to do with corrections or marginal notes that might have been present in the text. At a material level, copyists had to make sure their writing materials were in order, which, among other things, included periodically replenishing their pen with fresh ink and resharpening it to keep the strokes even. It has been argued, and not without reason, that such processes provided scribes at various levels with plenty of opportunities for error. In other words, scribes were busy people, and copying was a demanding job, which inevitably resulted in inaccuracies and blunders of all kinds.
\nHowever, the changes that are sometimes thought to be of particular historical significance, because of what they say about either early Christian beliefs, the origin of the Gospels, or the theological motives of scribes, often turn out to be rather more mundane than it seems at first sight. In this, we see that scribes\u2019 main goal was not to innovate, and when they did it was often accidental. While many such inaccuracies are still often to be seen in manuscripts, the scribes themselves were not ignorant of their shortcomings. Indeed, we may often observe that they strove to improve and revise their work before they handed it over to posterity. Scribal corrections are the clearest token of that effort.<\/p>\n
COPYING AND CORRECTING: EXAMPLES FROM THE MANUSCRIPT TRADITION<\/p>\n
The evidence of corrections is not always easy to interpret. Unlike such phenomena as nomina sacra (\u201csacred names\u201d), numerical abbreviations, readers\u2019 notes, or paragraphing, which are typically not transferred to modern editions, corrections are in varying degrees represented in critical apparatuses, hence their textual content can be accessed by such means. Yet corrections are much more than textual changes: their mode, timing, author, and source are important subjects in their own right, subjects that are in turn of vital importance for the accurate appraisal of corrections encountered in an apparatus. Corrections may also disclose scribes\u2019 attitudes toward their task and texts they were copying, thus being particularly valuable in the study of ancient book production. Importantly, how corrections are interpreted may significantly alter one\u2019s perception of scribes\u2019 overall performance. Understandably, then, corrections constitute an important piece in the puzzle of transmission history and as such are well worth closer attention.
\nAt a general level, there is no all-encompassing method for the study of manuscript corrections, as their nature varies from scribe to scribe, from manuscript to manuscript. Hence, the enterprise of this sort is inherently inductive and so requires patient acquaintance with the corrections themselves as well as with the broader tendencies of the manuscript where they occur. Not a few early manuscripts contain very sparse corrections or even none at all; in some de luxe codices, however, we have centuries of extensive correcting activity by different hands (e.g., Codex Sinaiticus). In some cases, the scribe corrected errors only in the process of copying, never checking his work thereafter; at other times, numerous reviews were conducted, sometimes even by means of two different exemplars. Rather than providing a vague theoretical framework for the study of corrections, then, it might be more useful to work through two specific examples of corrected manuscripts and (hopefully) learn something of the problem\u2019s complexity in the process.
\nP66: The more corrections, the merrier (or messier) the text? There are six early papyrus manuscripts, which may be regarded as \u201cextensive\u201d by the virtue of their substantial state of preservation. The first three of these are P45 (P.Beatty I; LDAB 2980), P46 (P.Beatty II; LDAB 3011), and P47 (P.Beatty III; LDAB 2778). Being part of the famous Chester Beatty biblical papyri, they were acquired by Sir Alfred Chester Beatty in the early 1930s and were published by Sir Frederic G. Kenyon within the following decade. The discovery was nothing short of sensational, for the Beatty manuscripts were in fact the first find of substantially preserved biblical papyri to be published. Thus, these artifacts presented scholars with a unique opportunity to study, on a much larger scale, the remnants of early Christian textual and material culture from Roman Egypt. Some two decades later, Martin Bodmer acquired a sizable lot of extensive papyri with more diverse contents, written in Greek and Coptic. The Bodmer papyri have brought to light many intriguing texts, but for our present purposes they are, besides the Beatty materials, the only other papyrus collection to include extensive early papyri with New Testament texts\u2014namely, P66 (P.Bodm. II; LDAB 2777), P72 (P.Bodm. VII\u2013VIII; LDAB 2565), and P75 (P.Bodm. XIV\u2013XV; LDAB 2895).
\nThe first published Bodmer New Testament papyrus is P66, a (probably) third-century codex containing sizable portions of John\u2019s Gospel. Initially, the discovery of P66 was rather quickly overshadowed by the publication of P75, a two-Gospel codex containing parts of Luke and John, owing to the latter\u2019s remarkably close affinity with Codex Vaticanus. However, despite the fact that the text of P66 was deemed more \u201cfree\u201d and further removed from the earliest recoverable text of John than that of P75, one particular aspect of this papyrus caught scholarly attention very early on. Gordon D. Fee, followed by Erroll F. Rhodes, brought to attention more than four hundred (!) instances of correction in the manuscript\u2014the vast majority penned by the same scribe. Arguably, such a high density of correcting by the original scribe is unparalleled in the New Testament manuscript tradition, and it raises a number of intriguing issues.
\nIn his recent study of scribal habits in P66, James R. Royse presents what is hitherto the most thorough investigation of corrections in this papyrus. Royse counts 465 such places and makes a number of distinctions among them. First, he identifies forty-nine \u201ccorrections in scribendo\u201d\u2014that is, made by the scribe in the very process of copying. At each such place, the scribe would correct his error in the making before proceeding further. Second, there are 164 \u201ccorrections of slips,\u201d where the scribe would, again, correct an obvious error, but at some later point rather than in the process of creating it\u2014whether upon completing the error or during a later check of his work. Third, Royse has a group of 126 corrections he refers to as \u201csignificant.\u201d These are places where the initial reading is not an obvious nonsense or orthographical slip but is still regarded as a scribal error.
\nNow, without these corrections (excepting the final group), the scribe\u2019s performance in his basic task of copying the text of his exemplar could have been regarded as extremely sloppy. Indeed, such is the assessment given by Ernest C. Colwell: \u201cIt is hard to believe that [producing a good copy] was the intention of P66.\u201d He characterizes the scribe\u2019s copying as \u201ccareless and ineffective\u201d and even marked by \u201cwildness.\u201d But the basis on which Colwell reached this conclusion is the text of P66 before correction. How should we then regard the text as the scribe left it after all these hundreds of revisions? Royse\u2019s appraisal, which follows the latter approach, is rather more positive: except for P75, the error rate reached by the scribe of P66 is \u201cmuch lower than that of any other papyri\u201d Royse considered (i.e., P45, P46, P47, P72). We see, then, that despite the scribe\u2019s initial lack of skill and care in reproducing his exemplar, he was greatly concerned for the accuracy of his copy, and the hundreds of corrections of his initial errors epitomize that concern.
\nThe final level of complexity in P66\u2019s corrections is reached in a group of (according to Royse) 126 readings where the scribe probably used another exemplar, one with different textual affinities. The practice of using another exemplar was not unheard of in antiquity and was at times followed in book shops that produced copies of literary papyri. Most likely, such usage was not reflective of \u201ctext-critical\u201d motivations on the scribe\u2019s part per se; rather, consulting a second exemplar was probably meant to facilitate a further check on the work one had already completed. But such intention behind the use of another exemplar need not mean that the intended result was actually achieved. Introducing readings from a different exemplar that exhibited different textual affinities occasions and exacerbates the problem of \u201ccontamination\u201d of the text, which is so typical of the New Testament tradition. At places with such corrections, the subsequent scribe using P66 as an exemplar for a further copy of John would have faced the same dilemma as its current readers: Should I follow the reading before or after correction\u2014or both? Be that as it may, P66 offers scholars a rich pool of information for further scrutiny at many different levels.
\nCodex Montfortianus (GA 61): Traces of Greek philology in the sixteenth century. The early age and the concomitant high sociohistorical value of the papyri often overshadow the manuscripts that constitute the vast majority of the Greek New Testament tradition, namely minuscules. These codices, made of parchment or paper, were written in a cursive Greek script that became widely used chiefly from the tenth century onward. Scholars tend to refer to this mass of materials by the shorthand \u201cByzantine manuscripts,\u201d which might imply a cohesive, stable whole. This, in general, is not entirely incorrect, given the greater similarity of such manuscripts in comparison with the papyri and early majuscules. It is far from true, however, at a level of individual manuscripts, which also exhibit their own peculiarities as well as unique scribal characteristics and textual features.
\nOne such a late treasure trove is GA 61, also known as Codex Montfortianus (Trinity College Dublin, Manuscript 30). Probably produced early on in the sixteenth century, this minuscule rose to fame primarily as the first Greek manuscript known to attest the so-called Comma Johanneum, a brief trinitarian interpolation at 1 John 5:7\u20138. It was from this manuscript, in fact, that Desiderius Erasmus introduced the Comma into the third edition of his Greek New Testament, thus becoming the standard feature of the textus receptus (and so also the Bible translations based on it). The feature for which GA 61 is little known, however, is an abundance of marginal notes, added by two different hands subsequent to its production.
\nOf particular interest are the corrections made in the book of Revelation. The principal text of Revelation in GA 61 reflects what is commonly known as the \u201cKoine\u201d strand of the Byzantine tradition. Virtually all of some sixty marginalia added by another hand in Revelation 1\u20134, however, exhibit a shift toward a form of text with much higher proportion of readings of the \u201cAndreas\u201d strand, as well as introducing some other, unique variants. Each reading is accompanied by a siglum denoting the Greek word \u03b3\u03c1\u03ac\u03c6\u03b5\u03c4\u03b1\u03b9 (graphetai, \u201cit is written\u201d), which is often used to mark off alternative readings in Byzantine manuscripts. Remarkably, one recent study has shown that these marginal corrections are likely to have been made on the basis of one of the first two editions of Erasmus\u2019s Greek New Testament. These marginalia, then, are not so much \u201ccorrections\u201d meant to remedy prior scribal errors but rather are expressions of philological activity on the part of a careful early modern reader of Revelation. In effect, what we have here are text-critical notations that alert the users of the codex to competing textual traditions at various points in the Greek text. The person who added these marginal notes shows awareness of the numerous differences between this particular copy of the Greek New Testament and another exemplar\u2014in this instance possibly a printed edition\u2014that may have been regarded as equally authoritative.
\nIn the case of GA 61, then, the 61c readings found in the apparatuses of our editions are not to be taken as corrections of scribal slips. Rather, they constitute something of a separate witness reflecting an alternative text form attested in Erasmus\u2019s edition. These editorial interventions are indicative of a critically acute reading and philological sensibility, as well as an interest in and awareness of the variant readings on the part of one of its sixteenth-century users. Given that these readings were made after Erasmus consulted the manuscript for his own edition, we may conjecture that the person responsible for them was a scholar resident in the British Isles rather than a native Greek speaker of the Eastern Orthodox background. As such, then, these marginal glosses are tangible traces of careful engagement with Greek Scriptures in the early modern biblical scholarly tradition. Indeed, this type of activity epitomizes the emerging marriage of the humanistic philological approaches and the increasing use of the original languages in Western theological circles. All in all, then, the case of GA 61 is a telling example of how corrections (depending, of course, on their age and nature) can offer useful information concerning the subsequent use of\u2014as well as the sociohistorical issues pertaining to\u2014the manuscript in which they appear.<\/p>\n
INTERPRETING THE EVIDENCE OF CORRECTIONS: THREE CASE STUDIES<\/p>\n
In discussing the above examples, we have encountered two very different manuscripts that display two very different types of correcting activity. Those two cases helped us observe corrections more broadly as a type of scribal activity with its own set of general tendencies, peculiarities, and significance. Having illustrated the problem on a broader scale, we now turn to a handful of readings where the evidence of corrections has considerable bearing on one\u2019s analysis of the text. We begin again with P66, the muchcorrected papyrus of John.<\/p>\n
Figure 8.1. Revelation 1 in Codex Montfortianus (Trinity College Dublin MS 30) showing corrections based on Erasmus\u2019s printed editions<\/p>\n
Jesus, the really thirsty, not-a-mere-man kind of God: Ehrman and P66. Few people who have heard about textual criticism at all have not heard the name of Bart D. Ehrman. Although Ehrman is widely known particularly for his New York Times bestseller Misquoting Jesus, a provocative book meant to introduce textual criticism to the broader public, perhaps the most famous of his scholarly works is The Orthodox Corruption of Scripture. In this (no less provocative) work, Ehrman embarks on a quest to unearth what he considers to be one very important, yet hitherto neglected, influence behind textual variation\u2014namely, the various christological controversies in the second and third centuries of the Common Era. This now-classic book has received many a review and critical response\u2014positive and otherwise\u2014and it is not within the scope of this discussion to respond to it in broad terms. For our purposes, however, it might be useful to recall one salient detail raised by Fee in what has been perhaps the most critical of Ehrman\u2019s reviews.
\nEhrman discusses three readings of the aforementioned P66 (Jn 10:30; 19:5, 28), where the variation involves scribal correction. At the first of these, Ehrman notes that the addition of the article before \u201cGod\u201d in P66 enhances \u201cthe force of this pronouncement [\u2018I and the Father are one,\u2019 Jn 10:30] and the clarity of the Jews\u2019 understanding of it.\u201d Now, this is a rather complex argument, not least because it extrapolates a strong semantic link from quite meager variation. What is more, however, this is not so much a reading of \u201cP66\u201d but rather of P66*\u2014that is, a reading before the scribal correction. In his work on scribal habits, Royse plausibly suggests that the initial reading arose by simple erroneous \u201cdoubling\u201d (dittography) of a syllable at the end of the Greek word yourself (\u03c3\u03b5\u03b1\u03c5\u03c4\u1f78\u03bd \u03c4\u03cc\u03bd, seauton ton), which the scribe then corrected against the same exemplar.
\nEven more intriguing, perhaps, is John 19:5, Ehrman\u2019s second example. Here he cites P66 as omitting the sentence \u201cBehold the man.\u201d He further asserts that the support furnished by the \u201cotherwise unrelated\u201d Old Latin and Lycopolitan (a dialect of Coptic) versions \u201cshould alert us to the possibility of a deliberate modification of the text.\u201d Following this line of reasoning, Ehrman then interprets the omission as reflecting the \u201ctroubling\u201d implications of the sentence: Jesus cannot have been a \u201cmere mortal.\u201d Again, it is uncertain whether the scribe of P66 would have been so troubled by the utterance of Pilate, who was in the end responsible for Jesus\u2019 execution. From a text-critical point of view, however, Ehrman\u2019s argument is considerably weakened by the fact that the omission seems to have been corrected: although the page in question is partly defective, the place of omission involves correction marks, and one can possibly even make out some traces of the correction in the lower margin area. This, of course, does not explain the presence of the variant in the aforementioned versions, but if an error like this could happen in the Greek, there is no a priori reason why it cannot have occurred independently in Latin and\/or Coptic. More troubling still is the fact that Ehrman never seems to show awareness that the reading of P66 was corrected by the same scribe, in each of the above cases against the very exemplar he used to copy the text in the first place.
\nBoth of Ehrman\u2019s abovementioned examples are meant to lend support for his argument concerning \u201canti-adoptionistic corruption\u201d in the New Testament textual transmission. The third instance is of \u201canti-docetic\u201d type and concerns variation at John 19:28. In this case, the scribe of P66 initially omitted the phrase \u201cin order that the Scripture might be fulfilled.\u201d Again, Ehrman argues that the reading does not appear to have originated by accident, since it does not involve an omission due to similar endings (homoioteleuton) or beginnings (homoioarcton) of the words and since P66 is not, in his view, \u201cparticularly prone to omissions of entire phrases or clauses.\u201d In accounting for the omission\u2019s origin, then, Ehrman warns us, \u201cWe would be remiss in this case to overlook the theological possibilities of the corruption.\u201d In particular, he posits that the presence of this clause may have struck the scribe as implying that Jesus was not genuinely thirsty: he said it only in order to fulfill the scriptural mandate. Whether Ehrman\u2019s explanation squares with the scribe\u2019s exegetical sensibilities is a moot point, however. For, as above, the omission was rectified by means of scribal correction. That the scribe would do away with such a conscious intervention into the text as Ehrman envisages seems most unlikely. Rather, we have an error\u2014possibly an omission by skipping over a line in the exemplar or simply by oversight\u2014that at some point the scribe noticed and corrected so as to restore the wording of the exemplar.
\nAs noted above, it was Fee who first noticed that Ehrman failed to distinguish between corrected and uncorrected readings in P66. His response to Ehrman\u2019s aforementioned interpretations is worth quoting in full:<\/p>\n
The deliberate \u201ccorruption,\u201d therefore, does not exist at all, since the correction in each case, which aligns the text with the rest of the MS tradition, was made by the original scribe himself (among hundreds of such). This scribe\u2019s \u201ccorrections\u201d are what are clearly deliberate\u2014and these show no interest in christology.\u2026 Significantly, this scribe stands squarely in the middle (ca. 200 CE) of the two centuries of Ehrman\u2019s interest. If Ehrman\u2019s case for \u201cchristological corruption\u201d so clearly fails in our one certain piece of evidence for deliberate variation, then one might rightly question the degree of deliberation in a large number of other variations as well, which seem to have equally good, if not better, explanations of other kinds for their existence.<\/p>\n
Whatever one makes of Fee\u2019s critique, it seems clear that the way in which the critic evaluates (or fails to evaluate) the evidence of these corrections has noticeable impact on her analysis of variants, including their reception-historical significance.
\nWho was in the beginning (of Mark\u2019s Gospel in Codex Sinaiticus)? Our next example comes from the famous Codex Sinaiticus (01; LDAB 3478), without doubt one of the most splendid expressions of Christian literary culture. One of the distinguishing features of Sinaiticus is the vast number of corrections made at various points of its initial production as well as its subsequent use. The first layer of corrections was made in the scriptorium by the very scribe who penned the codex, and the Codex Sinaiticus website designates these with the \u201cS1\u201d label. There are a good number of such cases (varying in extent from book to book), but not nearly as many as those made by the so-called C group, some two centuries later. Besides reflecting some measure of production care on the scribes\u2019 part, these corrections spark special interest because, in some biblical books, they have been shown to reflect the use of another exemplar. At such places, the manuscript in fact becomes a textual witness to both readings in question. Yet it is not always easy to distinguish a correction made against another exemplar from a mere correction of prior error, as we shall see in the following (somewhat notorious) example.<\/p>\n
Figure 8.2. The correction involving \u201cSon of God\u201d in Mark 1:1 in Codex Sinaiticus<\/p>\n
The abrupt opening of the Gospel of Mark has long fascinated its interpreters, not least thanks to a pressing textual problem in its very first verse: Did the author intend to open his narrative with \u201cthe beginning of the gospel of Jesus Christ\u201d or with \u201cthe beginning of Jesus Christ, son of God\u201d? A brief look at the critical apparatus in NA reveals that Sinaiticus is cited for both the omission and inclusion of the term \u201cson of God\u201d (\u03c5\u1f31\u03bf\u1fe6 \u03b8\u03b5\u03bf\u1fe6, huiou theou). For omission, the apparatus cites \u05d0*, a reading of the manuscript before the correction, whereas for the inclusion the siglum used is \u05d01, the earliest layer of corrections. Both the initial reading and the subsequent correction event have been interpreted variously. On the one hand, the scribe may have omitted the words \u201cson of God\u201d by homoioteleuton, which seems plausible given that each of the consecutive words would have had the same ending and given that such omissions are ubiquitous in the manuscript tradition. On that interpretation, the correction may have been affected by the same scribe, who simply restored the reading of his exemplar. In that case Sinaiticus\u2019s support for the shorter reading is considerably weakened.
\nOn the other hand, it is not impossible that, at this point, the scribe transmitted the reading of his exemplar accurately, such that the alteration was made by means of another exemplar at a later review in the scriptorium. That the correction was not made in the process of copying seems clear from a different, lighter-colored ink as well as from thinner pen strokes, which are characteristic of subsequent scriptorium corrections. Furthermore, the letter formation is reminiscent of corrections made by the scribe\u2019s \u201csenior\u201d colleague (termed \u201cscribe D\u201d). Interestingly, in several other places in Mark where scribe D can be identified as the author of a correction, he seems to have consulted a different exemplar. As noted above, similar evidence may be found in other New Testament books. If this analysis holds, then Sinaiticus provides genetic support for both the omission and the inclusion of the title \u201cson of God\u201d in the beginning of Mark\u2019s Gospel. Deciding on which of the two readings is to be preferred, however, is not the matter for our discussion, and it does not hinge solely on one\u2019s assessment of this variation in Sinaiticus.
\nThe Egerton Gospel: \u201cOur fathers\u201d are \u201cyour fathers.\u201d So far, all of our examples have concerned manuscripts with texts from the New Testament canon. In order to illustrate that such issues are not constrained to \u201cbiblical\u201d texts, we shall take a look at P.Egerton 2 + P.K\u00f6ln VI 255 (LDAB 4736), a manuscript with noncanonical contents. Named by its principal editors as an \u201cunknown gospel\u201d and now widely known as the Egerton Gospel (GEgerton), this text has attracted considerable scholarly attention, not least because of some conspicuous similarities with Jesus traditions found in the New Testament, especially in the Fourth Gospel. Upon its publication, most scholars argued variously for GEgerton\u2019s familiarity with or even literary dependence on the relevant New Testament texts, with only a few occasional dissenting voices.
\nThe most recently published portion of GEgerton, housed in the Papyrus Collection of the University of K\u00f6ln, contains a saying that bears close similarity to John 5:39: \u201cFor if you believed Moses, you would believe me; for concerning me he wrote to your fathers\u201d (frag. 1v, lines 20\u201323). Despite rather close correspondence, GEgerton does deviate from the canonical account by adding the phrase \u201cto your fathers,\u201d which too may well reflect Johannine influence (see Jn 6:49). Although typically papyrological details are not at the forefront of debates concerning literary dependence, Francis Watson has recently used just this datum as the point of departure for his argument for GEgerton\u2019s being something of a \u201cJohannine source.\u201d
\nBut let us return to the reading of our papyrus. It appears that the scribe first began to write \u201cour fathers\u201d (\u03c0\u03b1\u03c4\u03c1\u03ac\u03c3\u03b9\u03bd \u1f21\u03bc\u1ff6\u03bd, patrasin h\u0113m\u014dn) but then changed the reading to \u201cyour fathers\u201d (\u03c0\u03b1\u03c4\u03c1\u03ac\u03c3\u03b9\u03bd \u1f51\u03bc\u1ff6\u03bd, patrasin hym\u014dn), practically by altering just one letter. Watson criticizes Michael Gronewald, the fragment\u2019s editor, for preferring the corrected reading in the main text of his edition. According to Watson, once this \u201cdubious editorial decision\u201d has been abandoned, the Egerton saying may be seen as reflecting a \u201cMosaic stratum,\u201d that is, a tradition reflective of a Jewish-Christian setting before the expulsion of Christians from the synagogue. Watson, however, does not present clear papyrological reasons for either preferring the reading prior to correction or judging Gronewald\u2019s editorial decision as \u201cdubious.\u201d In effect, he commits the same fallacy.
\nWhile there are several ways in which this reading may be interpreted, a few conclusions can be made with a good measure of confidence. For one thing, it seems clear that the correction was made by the scribe himself and that there are no indications in GEgerton of a secondary correction by any other scribal hand(s). Moreover, neither the extent of the corrections in the papyrus nor the text\u2019s apparent lack of wider circulation suggests the possibility that the initial scribe would have used another exemplar with variant readings, with which he would have altered the \u201ccorrect\u201d reading to an alternative, competing one. Indeed, all the corrections in our papyrus are just that, corrections of prior scribal errors; and in each case, the correction was quite possibly made shortly after the initial error.<\/p>\n
Figure 8.3. P.K\u00f6ln VI 255; PA fragment of the Egerton Gospel, showing the correction in line 5 from \u201cour fathers\u201d (\u03c0\u03b1\u03c4\u03c1\u03ac\u03c3\u03b9\u03bd \u1f21\u03bc\u1ff6\u03bd, patrasin h\u0113m\u014dn) to \u201cyour fathers\u201d (\u03c0\u03b1\u03c4\u03c1\u03ac\u03c3\u03b9\u03bd \u1f51\u03bc\u1ff6\u03bd, patrasin hym\u014dn)<\/p>\n
In view of the foregoing, we can conclude that, even in such a short text without wide distribution or a history of authoritative use as GEgerton, the scribe took care to transmit the wording of his text with a good measure of accuracy. We have also seen, however, that interpretation of a corrected reading must rest on careful evaluation of evidence rather than on unfounded assumptions in either direction\u2014not to mention one\u2019s exegetical preferences. Rather, GEgerton seems to show awareness of the Johannine account at this point yet still further reinforcing Jesus\u2019 distancing from his Jewish interlocutors (called \u201clawyers\u201d in GEgerton; compare \u201cJews\u201d in Jn 6:41, 52). Even though we might not have a definitive confirmation of GEgerton\u2019s dependence on John here, this reading creates doubt about Watson\u2019s argument to the contrary.<\/p>\n
CONCLUSIONS<\/p>\n
Biblical manuscripts have a lot to offer if we are willing to pay close attention to their details and interact with them on their own terms. Nowhere has this principle been expressed clearer than in F. J. A. Hort\u2019s famous dictum: \u201cKnowledge of documents should precede final judgement upon readings.\u201d Knowledge of documents entails knowledge of their general tendencies as well as their particular details. One such detail (namely, scribal corrections) occupied us in the preceding discussion.
\nWe have seen that the scribal task was demanding, and so is the task for those of us who have undertaken to make sense of its results. The fact that the scribes were often willing (and at times eager) to improve on the results of their\u2014occasionally sloppy\u2014end products makes our task even more challenging but also all the more intriguing. It is hoped that a closer acquaintance with this particular aspect of the manuscript tradition will help us to be more alert to intricacies involved within and behind the variants we might encounter.<\/p>\n
Key Takeaways
\n\u25ba Corrections in manuscripts show that scribes strove to improve and revise their work before they handed it to posterity. The corrections in P66, for instance, show that despite its scribe\u2019s initial lack of skill and care in reproducing his exemplar, he was greatly concerned for the accuracy of his copy.
\n\u25ba Corrections added later in the life of a manuscript can show how its readers interacted thoughtfully with its text. They can even shed light on whether a scribe made any intentional, theologically motivated changes.
\n\u25ba Careful attention to corrections plays a role in debates about the origin of the Gospels. In the case of the so-called Egerton Gospel, a reading thought to support its use in writing John\u2019s Gospel turns out to be unlikely.<\/p>\n
CHAPTER NINE<\/p>\n
MYTHS ABOUT TRANSMISSION<\/p>\n
THE TEXT OF PHILEMON FROM BEGINNING TO END<\/p>\n
S. Matthew Solomon<\/p>\n
For too long, textual scholars have had to study the transmission of the Greek text of the New Testament with incomplete data. More than twenty years ago, Thomas Geer pleaded, \u201cFor too long in our discipline, too much has been based on too little. The time has come for full collations of manuscripts to enable us to write confidently about the history of the New Testament text.\u201d The primary goal of this chapter is to describe and analyze the text of Philemon as found in the entire Greek manuscript tradition. One of the outcomes of this chapter is to demonstrate that more work is needed, showing why work to establish the text of the New Testament continues. The eighteenth- and nineteenth-century pioneers in New Testament textual criticism would have taken full advantage of the access to images currently enjoyed by contemporary scholars, had images been available. They were hindered, however, by the cost of travel and the time it took to examine manuscripts in their individual settings\u2014usually libraries and monasteries. Contemporary scholars are now able to use critical apparatuses as jumpingoff points for further research because they can view textual variants in images of manuscripts made widely available.
\nA legitimate question at this point is, If we basically have the original text of the New Testament, why should we investigate more manuscripts in more detail? The best answer is that the story of textual variants in the text of the New Testament is only beginning to be told. While major variation units are covered in the Greek New Testaments that most scholars, seminarians, pastors, and laypeople use, exponentially more textual variants exist. To demonstrate this point, it is important to note that a comprehensive comparison of all Greek manuscripts of Philemon has revealed at least 330 new textual variants (i.e., new readings, not necessarily new variation units) compared to major editions dating back to the 1800s. For a letter that is only 335 words long in NA\/UBS, this number demonstrates that there is much more going on in the manuscript tradition than appears. The UBS apparatus presents only four variation units, and the NA apparatus presents only twenty-three. The reality, however, is that hundreds of places of textual variation exist for the small letter of Philemon.
\nI hope that by the end of this chapter, you will see how important complete examinations of all New Testament manuscripts are to the continuing study of New Testament textual criticism and how the text was transmitted throughout the years. We must remember that every single handwritten copy of the Bible was someone\u2019s Bible\u2014whether personal or church copies (more likely). Since this is the case, we can look at every single one of these copies in a quest both for the original reading and for how churches and individuals have interpreted the text over the years. Sometimes, textual variation can act as commentary on the text of the New Testament, a fact to which we will return below.<\/p>\n
COLLATING PHILEMON<\/p>\n
Textual scholars read and compare manuscripts so they can piece together the \u201coriginal text\u201d and understand better the people who copied those texts. The job of the New Testament textual critic is incredibly difficult, however, because of the vast wealth of data available (fifty-three-hundred-plus manuscripts). One tool that textual critics across the centuries have used is called collation, wherein two or more texts are compared to each other to reveal and catalog the differences between them.
\nTo date, a large number of manuscripts of the New Testament have not been examined in detail. While the oldest witnesses to the text of the New Testament (as well as the most important newer ones) have been studied, much work remains to be done. Never in the history of the world have scholars had access to more data than is currently available. In the blossoming world of digital humanities, people involved in New Testament textual studies have experienced unprecedented access to manuscript images. Libraries and museums around the world are digitizing their collections, allowing access to manuscripts that was previously costly to the viewer in both time and money. The INTF is in the process of making available online images of New Testament manuscripts that were previously available only in microfilm format on site. The CSNTM has photographed manuscripts in high-resolution digital format. Previous generations of scholars worked well with the information that was available, often using collations compiled by others who studied particular manuscripts in person. Access to the manuscripts themselves was limited. Thankfully, the world is now ready for comprehensive study of individual books and manuscripts of the New Testament.
\nFor my dissertation, I collated every available Greek manuscript of Philemon against the modern text of UBS\/NA. After the collations, I input all the differences into a textual apparatus, where all textual variants for the entire Greek tradition of the text of Philemon are presented. But, before we go any further, I will explain the basics of what it is textual critics actually do in collation.
\nPaper collations. Traditionally, the process of collating manuscripts included comparing a particular manuscript to a standardized base text. For many, this standard was the textus receptus because it largely represented the text of the New Testament in the majority of manuscripts, leaving fewer differences to record. The collator would compare the text of a manuscript with the text of the textus receptus, and when the two texts disagreed, the collator would record the differences.
\nLet us look at an example. At Philemon 1 in the textus receptus, Paul identifies himself as \u03b4\u03ad\u03c3\u03bc\u03b9\u03bf\u03c2 (desmios, \u201cprisoner\u201d). In GA manuscript 945, Paul identifies himself as \u03b4\u03bf\u1fe6\u03bb\u03bf\u03c2 (doulos, \u201cslave\u201d). So, when a collator came across the reading in 945, he or she would record as follows:<\/p>\n
Verse: base text reading] manuscript reading
\nPhilm 1: \u03b4\u03b5\u03c3\u03bc\u03b9\u03bf\u03c2] \u03b4\u03bf\u03c5\u03bb\u03bf\u03c2<\/p>\n
In this way, the textual critic is recording only the differences between that particular manuscript and the textus receptus. Where there are no differences between the two, the text of the manuscript can be assumed to contain the text of the textus receptus. In other words, the collation represents the entire text of the manuscript in an abbreviated space that allows scholars to access data easily.
\nAnother way to produce a paper collation is to print the base text and then write in the differences between the manuscript and the base text. In this way, one step is removed: the collator does not need to record the base text reading every time there is a difference between the two. The collator produces what amounts to a paper transcription of the manuscript, producing the text exactly as it appears. The Center for New Testament Textual Studies produced its electronic critical apparatus using the latter paper transcription method described above, the results of which were then manually entered into an electronic document.
\nThe work of the collator is laborious. The collator records every single difference between the manuscript and the base text, a process that requires many hours of work. For several reasons, collation is especially difficult. First, the goal of producing a collation is to be perfect. For this reason, at least two collations should be completed. The difficulty with achieving perfection should be apparent. When comparing two documents and recording the differences, the collator could easily miss differences between the texts in the act of looking back and forth between the two documents. So the collator in most instances should read letter by letter in order to avoid errors. A further challenge to the collator is avoiding fatigue. The collator must take breaks regularly to avoid undue eye strain.
\nElectronic transcriptions. Despite the advent of computers, textual critics should still master the ability to complete a paper collation, as some libraries and other settings may not allow for the use of electronics when inspecting a manuscript. Contemporary textual critics, though, prefer to transcribe manuscripts electronically, which has advantages over paper transcriptions. These electronic transcriptions are made by changing a digital base text to match as closely as possible what is displayed in a manuscript. Electronic transcriptions provide data that can be used in a variety of ways by anyone. A transcriber can also add paratextual data in the form of markup languages such as XML tagging. What this means is that not only is the manuscript text itself recorded but many of the codicological features as well. These paratextual features include paragraph markers, punctuation, column and page breaks, illuminations, decorations, titles, and postscripts, among other things. A comprehensive electronic transcription allows for a manuscript to be published as it appears but in electronic format.
\nThe INTF, the Institute for Textual Scholarship and Electronic Editing, the Museum of the Bible Scholars Initiative, the Center for New Testament Textual Studies, and the International Greek New Testament Project now transcribe texts of the New Testament electronically, which allows for unprecedented levels of collaboration. Collators compare images of manuscripts to digital base texts, making changes to the electronic text. The result is an exact digital transcription of the text of the manuscript, complete with tagging of paratextual features.
\nPhilemon collation process. The process of comparing every single manuscript of Philemon to one another began with several decisions. First, I decided to use the same process of collating as the Center for New Testament Textual Studies (as described above). Also, I decided that I would record every difference found in the manuscripts, including the minor differences discussed above. That way, once a paper transcription was completed, I would not have to consult the manuscript for the text (although some paratextual features might need to be revisited).
\nA practical example can illustrate why it is important to record seemingly meaningless differences. One of the major textual variants in the New Testament can be found in Romans 5:1. Here, two different readings are found in the manuscripts: \u1f14\u03c7\u03bf\u03bc\u03b5\u03bd (echomen, \u201cwe have peace\u201d) and \u1f14\u03c7\u03c9\u03bc\u03b5\u03bd (ech\u014dmen, \u201clet us have peace\u201d), where the difference between the two types of verbs is one single vowel. Omega and omicron sound very similar when read out loud. Normally, this type of difference would be overlooked, except that in this instance the difference in vowels changes the mood of the verb from a concrete thought (indicative) to a more contingent thought (subjunctive) in a moment when Paul is discussing justification by faith. Potential theological ramifications might be found in the textual variant found in Romans 5:1. One of the best witnesses to the text of Paul\u2019s letters is 33, which reads \u1f14\u03c7\u03c9\u03bc\u03b5\u03bd (ech\u014dmen, \u201clet us have\u201d). This manuscript, though, repeatedly changes between omicron and omega throughout its text of the New Testament. Given this tendency, is 33 a witness to \u201clet us have\u201d or to \u201cwe have\u201d? Although 33 has an omega in Romans 5:1, the answer is ambiguous. When viewing the NA apparatus, one would not be immediately aware that vowel interchange occurs frequently in 33.
\nThe research process was laborious. Not only did I have to sift through the 750-plus Pauline manuscripts to determine which of them contained Philemon, I had to arrange access to those manuscripts, whether through images online, through the microfilms and photographs stored at research centers such as the Center for New Testament Textual Studies or INTF, or by visiting libraries to see the manuscripts in person. Only then could I begin collating. The entire process lasted one and a half years. Since accuracy is valued over a completed product, the work of a textual critic moves slowly. The smallest details, the smudging of ink, a hole in the manuscript\u2014everything is closely examined to determine what text the manuscript contains. This focus on 100 percent accuracy explains why the entire New Testament has not been analyzed in this comprehensive way. Philemon is 335 words long in NA\/UBS, making it one of the smallest books in the New Testament. Many other New Testament books are much longer and therefore require more time and energy.<\/p>\n
THE TEXT OF PHILEMON<\/p>\n
The process of transcribing and collating manuscripts allows scholars to track how ancient texts morphed and changed throughout the copying process over approximately fifteen hundred years. In this section, I will provide a general overview of the text of Philemon in the Greek manuscript tradition. Also, I will consider what the text of Philemon would look like if we did not have the earliest manuscripts that we do.
\nThe text of Philemon before AD 900. Paul\u2019s letter to Philemon can be found in more than 570 Greek manuscripts. This number can be deceiving, though. Only twenty-three out of the 570-plus manuscripts of Philemon. date before AD 900. The earliest witnesses include P61 (ca. 700), P87 (third century), P139 (fourth century), 01 (fourth century), 02 (fifth century), 04 (fifth century), 06 (sixth century), 016 (fifth century), and 048 (fifth century). The earliest Greek witness to Philemon, P87, is a small scrap containing little text, fewer than one hundred letters. P139, too, is a small scrap containing portions of a handful of verses. A later papyrus manuscript of Philemon, P61, also is fragmentary and contains very little text. Manuscript 04 is lacking the first two verses; 016 is fragmentary, having suffered burns; and 048 is a palimpsest in which the undertext has been written over and is very difficult to see, leading to deficient portions for Philemon. In other words, of the nine manuscripts containing Philemon before circa AD 700, only three contain the entire text of Philemon: 01, 02, and 06. Fourteen extant manuscripts of Philemon date to the ninth century AD: 010, 012, 018, 020, 025, 044, 0150, 0278, 0319, 33, 1424, 1841, 1862, and 1900. These manuscripts represent different textual clusters and signify a period of some textual diversity. In addition, thirty-six extant manuscripts of Philemon date to the tenth century, and these contain texts very similar to the Majority Text. So, while we have many manuscripts of Philemon, it is important to note that the majority of those manuscripts represent a later text. Only about 4 percent of the manuscripts we have for Philemon date to before the year 900, and that number climbs to only about 10 percent when including manuscripts from the 900s.
\nThe text of Philemon in later manuscripts. While scholars try to use representatives of all manuscripts and principles of textual criticism to reconstruct the initial text, many manuscripts from the second millennium AD are not consulted because their texts are so similar. We need to consult all the manuscripts in order to determine which, if any, manuscripts are representative of the majority of later, similar manuscripts (i.e., the majority text). This may also be the case with Philemon, minus a few instances. So how would the text of Philemon read if we did not have manuscripts that dated before the year 900? In this section we will examine the differences between the majority text of Philemon and the initial text as represented in the UBS\/NA texts.
\nIf the earliest witnesses to Philemon did not exist, a couple of familiar readings might be different in modern Greek New Testaments and English translations. One of the differences found between the NA\/UBS text and the majority text can be found in Philemon 2, where the letter is addressed to Apphia as \u1f00\u03b4\u03b5\u03bb\u03c6\u1fc7 (adelph\u0113, \u201csister\u201d) in the NA\/UBS and \u1f00\u03b3\u03b1\u03c0\u03b7\u03c4\u1fc7 (agap\u0113t\u0113, \u201cbeloved\u201d) in the majority text. The vast majority of later manuscripts and early printed critical editions of the Greek New Testament do not contain what is believed to be the initial text of Philemon in Philemon 2. This is not to say that no one would have had access to this reading in the time period after the year 900. On the contrary, \u201csister\u201d can be found in twenty manuscripts after AD 900. Without the earlier witnesses from a key period in the transmission of Philemon, we might adopt the wrong reading as original. In other words, if we did not possess the eleven other manuscripts before the year 900 that have the reading \u201csister,\u201d we would more than likely not consider it a viable option for the initial text.
\nAnother reading that would be different without the earliest Greek manuscripts of Philemon can be found in Philemon 7, where Paul states that he prays \u201cbecause I have great joy.\u201d In the majority of Greek manuscripts, especially those after 900, Paul states that it is \u201cbecause we are very thankful.\u201d If the earliest manuscripts did not exist, then the reading \u201cI have great joy\u201d would probably be seen as secondary because only about 10 percent of the manuscripts that date from the tenth to the sixteenth centuries contain that reading. The rest read \u201cwe are very thankful.\u201d
\nThe prime example of how different Philemon might be without the earliest Greek witnesses comes in Philemon 12, where the initial text reads, \u201cwhom I sent to you, him, this one is my heart.\u201d In this syntactically difficult sentence, the initial text states that Paul is sending Onesimus, who is very dear to him, back to Philemon. Out of the entire Greek manuscript tradition, only five manuscripts contain this initial text: 01*, 02, 010, 012, and 33. The remaining Greek witnesses to the text of Philemon add the verb receive, so the verse reads \u201cwhom I have sent to you, him, this one is my heart, receive him,\u201d looking forward, perhaps, to the idea of Paul\u2019s command to receive Onesimus as himself in Philemon 17. Scribes added the verb in Philemon 12 in multiple places, further suggesting it is not part of the initial text of Paul\u2019s letter. The problem with adding this warm welcome prematurely is that it goes against Paul\u2019s logic in this part of the letter. Paul describes his feelings solely for Onesimus and does not want to give Philemon hope that Onesimus will return for good.<\/p>\n
TYPES OF VARIANTS IN PHILEMON<\/p>\n
Whereas some places in the New Testament manuscript tradition seem to contain textual variants that are conscious changes to the text made by either copyists or readers, most of the variants present in Philemon seem to be introduced simply to make the text clearer. In the ancient world, most reading was done aloud in public settings, so many variants no doubt crept into the textual tradition to make the text communicate clearly. In this section I will present examples of these types of variants that occur frequently throughout the text of Philemon. Examples of variants that seem to be conscious changes to the meaning of the text will be handled in the next section.
\nBefore turning attention to specific examples, a brief introduction to the types of textual variants found in manuscripts is needed. Textual critics have observed both major and minor variants over the years. Major variants include additions or omissions, replacements, and differences in word order. These are the types of variants that appear in the apparatuses of critical editions.<\/p>\n
Minor variants, on the other hand, include spelling differences that do not change the meaning of words, movable nu endings, obvious errors, differences in how proper names are spelled, errors in sight where text is added or removed due to a slip of the eye, and sometimes different types of abbreviations. These minor variations are prolific in the manuscript tradition, wherein almost every word of the text of the New Testament can be found in a variation unit of some sort (whether major or minor). The idea that the UBS\/NA editions represent basically the entire manuscript tradition minus the variants listed in their apparatuses is one that needs to be eliminated.
\nAdditions and omissions. Let us now consider a few examples from Philemon. The first category is additions and omissions. Sometimes copyists or readers added text that made the meaning clearer. Examples of this can be observed in Philemon 10 and Philemon 25. In Philemon 8, Paul begins his appeal on behalf of Onesimus, and in Philemon 9 he introduces the main verb \u03c0\u03b1\u03c1\u03b1\u03ba\u03b1\u03bb\u1ff6 (parakal\u014d, \u201cI appeal\u201d). In Philemon 10, this same verb is used, and some manuscripts insert different coordinating conjunctions in different positions. For example, 075 along with thirty-four other manuscripts adds \u03bf\u1f50\u03bd (oun, \u201ctherefore\u201d), connecting the second instance of \u201cI appeal\u201d back to the first. A handful of other manuscripts at this point add either \u03ba\u03b1\u03af (kai, \u201cand\u201d) or \u03b4\u03ad (de, \u201ceven\u201d) to make the connection. These conjunctions were added to aid both readers and hearers in the public reading of the text. In most cases, there are no major exegetical implications and the change in meaning is insignificant, especially since most of these types of readings are secondary. Another example in Philemon 10 is the addition of the pronoun \u03bc\u03bf\u03c5 (mou, \u201cmy\u201d) after the prepositional phrase \u1f10\u03bd \u03c4\u03bf\u1fd6\u03c2 \u03b4\u03b5\u03c3\u03bc\u03bf\u1fd6\u03c2 (en tois desmois, \u201cin bonds\u201d). This is an unnecessary clarifying feature of the majority of manuscripts of Philemon, seeming to be a move on the part of scribes to ensure that the reader understood that Paul is the one imprisoned, not Onesimus.
\nAnother type of addition may be liturgical in nature. In Philemon 25, which is a benediction of sorts, reading \u201cmay the grace of the Lord Jesus Christ be with your spirits,\u201d two examples can be observed. First, most manuscripts of Philemon have added the pronoun \u1f21\u03bc\u1ff6\u03bd (h\u0113m\u014dn, \u201cour\u201d) to the phrase \u03c4\u03bf\u1fe6 \u03ba\u03c5\u03c1\u03af\u03bf\u03c5 (tou kyriou, \u201cof the Lord\u201d). Most manuscripts have also added the familiar \u1f00\u03bc\u03ae\u03bd (am\u0113n, \u201camen\u201d) to the end of the phrase. These additions do not affect meaning in any significant way but may have arisen due to the liturgical nature of Philemon 25 and the letter\u2019s use in worship.
\nReplacements. In the second category, copyists and\/or readers sometimes replaced or substituted words in the text with little to no effect on meaning, usually making the text slightly clearer. Examples of this concept can be observed in Philemon 5 and Philemon 20. In Philemon 5, Paul is fleshing out why he is thankful for Philemon, because he has heard of his love and faith that he has \u03c0\u03c1\u03cc\u03c2 (pros, \u201ctoward\u201d) the Lord Jesus and \u03b5\u1f30\u03c2 (eis, \u201cfor\u201d) all the saints. The language here can be seen as awkward for Paul, as the use of \u03c0\u03c1\u03cc\u03c2 with \u201cfaith\u201d in reference to \u201cLord Jesus\u201d is not found elsewhere in the New Testament. Some copyists, then, have changed the prepositions to correspond to normal usage, with some changing \u03c0\u03c1\u03cc\u03c2 to \u03b5\u1f30\u03c2 and\/or \u03b5\u1f30\u03c2 to \u03c0\u03c1\u03cc\u03c2 in the verse. More than likely, these decisions stem from a desire to normalize the syntax in this verse with the rest of Paul\u2019s writings.
\nAnother example of a replacement is found in Philemon 20, where we can observe where some manuscripts read \u1f10\u03bd \u03c7\u03c1\u03b9\u03c3\u03c4\u1ff7 (en christ\u014d, \u201cin Christ\u201d), while most read \u1f10\u03bd \u03ba\u03c5\u03c1\u03af\u1ff3 (en kuri\u014d, \u201cin the Lord\u201d), and a few omit the phrase altogether. This variation unit is unique in that the difference between the two main readings is a difference of one very similar-looking letter. In Christian manuscripts, copyists used a convention scholars call nomina sacra or \u201csacred names,\u201d wherein they abbreviated certain divine and otherwise significant words, adding a line above the abbreviation. The difference between these two readings is a single letter: \u03c7 (chi) versus \u03ba (kappa). These two letters are easily confused, and the difference in meaning is not very significant in the context of Philemon 20. Although this counts as a major variation unit, the difference in meaning is not significant, and the rise of different readings is mechanically explainable.
\nTranspositions. In the third category, copyists and\/or readers sometimes changed the word order in the text with little to no effect on meaning. Examples of this can be observed in Philemon 1 and Philemon 13. In Philemon 1, Paul includes his typical letter greeting, including the phrase \u201ca prisoner of Christ Jesus.\u201d The textual tradition is split almost fifty-fifty here between manuscripts that read \u201cChrist Jesus\u201d and manuscripts that read \u201cJesus Christ.\u201d This transposition is curious considering that Romans, 1 and 2 Corinthians, Ephesians, Philippians, Colossians, and 1 and 2 Timothy all read \u201cChrist Jesus,\u201d while Galatians, 1 and 2 Thessalonians, and Titus read \u201cJesus Christ.\u201d Even though this counts as a major variation, there is no major difference in meaning. Similarly, in Philemon 13, the majority of manuscripts reverse the order of \u03bc\u03bf\u03b9 \u03b4\u03b9\u03b1\u03ba\u03bf\u03bd\u1fc7 (moi diakon\u0113, \u201che might serve me\u201d). This change does not affect the meaning of Philemon 13, but the move was more than likely done to aid in reading. In its original word order, two pronouns are read one after the other: \u03c3\u03bf\u1fe6 (sou, \u201cof you\u201d), then \u03bc\u03bf\u03b9 (moi, \u201cto me\u201d). It appears that many readers of Greek manuscripts preferred not to read these pronouns in sequence, breaking them up with a change in word order that does not affect meaning at all due to the inflectional nature of the Greek language, wherein word order is not the most important factor in the grammatical function of words.
\nWhile viewing the apparatus of NA or UBS, one might come to the conclusion that the text of Philemon contains few textual variants. This is certainly not the case. Major variants, such as the examples just discussed, and minor variants, explained above, can be found throughout the text of Philemon. In fact, almost every word in Philemon is included in a variation unit wherein at least one manuscript contains a textual variant. Most of the variants do not greatly affect meaning, nor do they tell the story of the history of the textual transmission. For this reason, most textual variants are left out of apparatuses. An examination of all Greek witnesses to the text of Philemon has revealed that the textual tradition contains more variants than some might expect, even if most have no serious merit.<\/p>\n
TEXTUAL VARIANTS AS COMMENTARY<\/p>\n
If most textual variants do not affect meaning in any significant way, should we still study them? At some point, every ancient and medieval Greek manuscript was used in a real church in a real community. Each of these manuscripts was someone\u2019s Bible. Over the years, though, textual variants arose through the copying process. Regardless of how these variants arose, churches that read from these manuscripts may not have known of any different readings. Our question, then, becomes: Once the task of establishing the original text is satisfactorily complete, what are we to do with the leftover variants? Additionally, can textual variants shed light on how an ancient community understood the text of the New Testament? In the same way, can textual variants help in modern interpretation of the text of the New Testament? This section will explore a selection of exegetically significant variants from Philemon that can shed light on meaning and interpretation throughout the centuries.
\nFrom \u201cparticipation\u201d to \u201cministry.\u201d Philemon 6 is by far the most difficult verse in the book to understand in terms of vocabulary and syntax. As Paul continues his thanksgiving section, he provides a purpose clause that gives an object to his ambiguous prayer from the second half of Philemon 4, stating, \u201cso that the participation of your faith may become effective in the knowledge of every good thing which is in us in Christ.\u201d Most manuscripts in the Greek tradition read \u03ba\u03bf\u03b9\u03bd\u03c9\u03bd\u03af\u03b1 (koin\u014dnia, \u201cparticipation\u201d). Some manuscripts (01c, 0150, 1874, 1881, and others), though, read \u03b4\u03b9\u03b1\u03ba\u03bf\u03bd\u03af\u03b1 (diakonia, \u201cministry\u201d) in place of \u201cfellowship\u201d or \u201cparticipation.\u201d The overwhelming support of early and reliable witnesses gives precedent to \u201cparticipation\u201d over \u201cministry,\u201d making the second reading the focus of our discussion here.
\nScribal considerations must be examined as well. The specific meaning of \u03ba\u03bf\u03b9\u03bd\u03c9\u03bd\u03af\u03b1 here has been the subject of some debate. Modern scholars have provided a range of meanings for the term: \u201cassociation,\u201d \u201ccommunion,\u201d \u201csharing,\u201d \u201cparticipation,\u201d \u201cpartnership,\u201d \u201ccontributions,\u201d \u201cfellowship,\u201d and others. Modern difficulties in understanding the meaning of \u03ba\u03bf\u03b9\u03bd\u03c9\u03bd\u03af\u03b1 may shed light on why \u03b4\u03b9\u03b1\u03ba\u03bf\u03bd\u03af\u03b1 can be found in some manuscripts. Perhaps ancient and medieval readers and scribes shared the same difficulty and sought to clarify the verse by changing \u03ba\u03bf\u03b9\u03bd\u03c9\u03bd\u03af\u03b1 to \u03b4\u03b9\u03b1\u03ba\u03bf\u03bd\u03af\u03b1 to avoid ambiguity.
\nAnother important aspect to consider is strictly a letter issue: the two terms share several common letters in Greek. A careless scribe initially could have made the change from participation to ministry unintentionally. In this case, ministry could have stayed in the tradition, as the noun makes sense in the flow of Paul\u2019s letter. Whichever way the variant arose, the presence of ministry in the textual tradition points to difficulties in interpretation throughout the centuries. Perhaps this variant reading can help us in interpretation. In this particular instance, past readers struggled with the meaning of the word participation. Some readers, though, were comfortable with the term ministry or service of faith here. This variant reading helps us to understand that this activity of faith that Paul was praying would become effective was certainly connected to the church and involved an act of service or ministry, not simply a passive type of fellowship.
\nThe addition of work. Continuing in Philemon 6, Paul prays that the participation of Philemon\u2019s faith will be effective \u201cin the knowledge of every good thing.\u201d The ambiguous use of \u1f00\u03b3\u03b1\u03b8\u03bf\u1fe6 (agathou, \u201cgood thing\u201d) here led to the addition of a clarifying noun, \u1f14\u03c1\u03b3\u03bf\u03c5 (ergou, \u201cwork\u201d). In this case, \u201cevery good\u201d became \u201cevery good work.\u201d Once the reading made its way into the tradition, it remained, being found in more than 120 manuscripts. Although found in all these manuscripts, the addition appears to be secondary in light of external evidence.
\nScribal and authorial considerations must be examined as well. In terms of inscriptional probabilities (i.e., what Paul might have written), the addition of work might appear to go against the grain in terms of what Paul usually communicated to his congregations. Although Paul emphasizes salvation being free from \u201cworks of the law,\u201d good works more broadly is not a concept foreign to his teaching. Paul uses a form of the phrase \u1f14\u03c1\u03b3\u03bf\u03bd \u1f00\u03b3\u03b1\u03b8\u03cc\u03bd (ergon agathon, \u201cgood work\u201d) twelve times in his letters, each time in a positive sense (Rom 2:7; 2 Cor 9:8; Eph 2:10; Phil 1:6; Col 1:10; 1 Tim 2:10; 5:10; 2 Tim 2:21; 3:17; Titus 1:16; 3:1). In this instance, therefore, it is possible that Paul could have penned \u201cevery good work.\u201d
\nIn terms of transcriptional probabilities, scribes might have added work to clarify the somewhat ambiguous good. \u201cOf all good\u201d or \u201cof every good thing\u201d is open ended, which could have led some scribes to insert a clarifying noun. Conversely, if work were present initially, its deletion (thus creating a more ambiguous phrase at the end of an already awkward verse) seems less likely. Also, six of the twelve instances of the phrase \u201cgood work\u201d occur in the Pastoral Epistles, located right before Philemon in most manuscripts. Due to close proximity, scribes could have added work to good to correct the perceived error of its absence because they were accustomed to seeing the phrase.
\nThe secondary nature of the phrase \u201cevery good work\u201d makes it a prime candidate for a variant that can be used as commentary for this particular verse. The original text \u201cevery good\u201d by itself cries out for something to complete the thought. In this case, scribes inserted work, giving the otherwise ambiguous phrase some kind of tangible action or result. Paul\u2019s prayer was that the participation of Philemon\u2019s faith would become effective in the knowledge of not just all goodness but specifically good deeds and actions. That this reading found staying power in about 20 percent of the Greek manuscript tradition indicates its acceptance as a legitimate way to understand the text here.
\n\u201cIn us\u201d or \u201cin you.\u201d The second-to-last phrase in Philemon 6 includes the phrase \u201cin us\u201d in the text portion of modern Greek critical editions. The textual tradition, though, is split about two-thirds to one-third in total number of manuscripts that read \u1f21\u03bc\u1fd6\u03bd (h\u0113min, \u201cin us\u201d) and \u1f51\u03bc\u1fd6\u03bd (humin, \u201cin you [plural]\u201d), respectively, which is an unusual split percentage for this letter. Important manuscripts also are split almost evenly in number: 02, 04, 06, 044, 048, 81, and the majority text read \u201cin us,\u201d while P61, P139, 01, 010, 012, 025, 0278, 33, 1739, and 1881 read \u201cin you.\u201d One manuscript, 38, reads \u201cin me,\u201d which is just a strange reading. Only considering external evidence, the edge belongs to \u201cin you,\u201d as the manuscripts that contain the reading are of slightly better quality, especially with the fourth-century witness of P139.
\nWith external evidence slightly favoring one reading over another, scribal and authorial evidence must be considered as perhaps a deciding force. As far as authorial considerations are concerned, Paul uses \u201cin you (plural)\u201d 164 times and used \u201cin us\u201d only thirty-five times in his letters, which is a split of about 80 percent to 20 percent. Interestingly, he uses first-person personal pronouns 826 times in his letters and second-person personal pronouns 841 times, which is nearly an even split. When he uses personal pronouns in the dative case, he prefers to use \u201cin you (plural).\u201d So Paul more than likely would have used \u201cin you (plural)\u201d in this instance. As far as scribal considerations are concerned, scribes would have been more likely to change \u201cin us\u201d to \u201cin you (plural)\u201d because of the presence of \u201cof you\u201d earlier in the verse. If \u201cin you (plural)\u201d were present initially, the shift to \u201cin us\u201d is difficult to explain. Authorial and scribal considerations in this case are inconclusive.
\nThis variant is often discussed in commentaries. The general consensus favors \u201cin us\u201d as original, mostly due to assimilation. Bruce Metzger argues that \u201cin us\u201d is original due to its expressive nature and the fact that scribes would have been more likely to assimilate the text to other second-person personal pronouns in the surrounding verses. Barth and Blanke devote one and a half pages of their commentary to discussing the implications of both readings, which is far more coverage than any other commentary. They argue that \u201cin you\u201d should be the adopted reading based on intrinsic grounds. Referencing Ernst Lohmeyer, they point out that in Paul\u2019s thanksgiving sections, he normally does not lump himself with the congregation to which he is writing. Further, they argue that Paul never thanks God for granting him love and faithfulness to an outstanding degree. To this end, all good things would be experienced in the congregation, not simply in an ideal and abstract manner.
\nBased on manuscript, scribal, and authorial considerations, both readings deserve to be considered by interpreters. In other words, here is a place where no major exegetical point should be made solely on this particular word. Both readings have compelling evidence, and neither can be completely settled. Both readings are adapted by good scholars. Both readings, therefore, should be considered when completing the exegetical task of Philemon 6. This particular place of variation among the Greek manuscript tradition is not simply a place where we decide the original and throw out the variant. In the text of Philemon, we have a place where we cannot say for certain which reading is initial to the text of Paul\u2019s letter.
\nSummary. While we have the luxury of examining nearly two millennia of manuscripts and critical editions of the Greek New Testament, churches throughout the centuries used the texts they had for the public reading of Scripture. In some instances, awareness of different readings existed (especially in the early period). For the most part, though, texts with variations were read and accepted as authoritative in churches. For Philemon, we can see several places where textual variation can act as commentary. Textual variations can help us to understand how the text was understood in some settings. As such, textual variations that have been deemed as completely secondary should not be jettisoned completely by scholars, pastors, and laypeople but should be seen as a part of the interpretive Technicolor tapestry of the church being woven for nearly two thousand years now.<\/p>\n
CONCLUSION<\/p>\n
The text of Philemon, like most of the New Testament, enjoys a complicated history, more so than many realize. Through the painstaking process of reading and examining every single letter and mark of every single extant Greek manuscript, the picture becomes somewhat clearer. Paul\u2019s letter to Philemon survives in 570-plus extant Greek manuscripts to which we currently have access. Of those, only twenty-three survive from the period before the standardization of the Majority Text. Even among those, only three contain the entire text of Philemon from the earliest period of the copying process.
\nUpon comprehensive examination of the text to Philemon, at least 330 new readings have been discovered that are not currently published in critical editions from the past three hundred years, when serious study of textual criticism began in earnest. All variant readings help in understanding not only the transmission of the text but also how churches and communities understood these texts throughout the years. It is safe to say that we have the entire initial text of Paul\u2019s letter, even if we have two places where the initial text may be uncertain between two words. We do not know which reading is original\u2014\u201cto us\u201d or \u201cto you (plural)\u201d in Philemon 6, and the presence or absence of and in Philemon 11. For these places, which reading is original really does not matter for interpretation. It does not matter because it did not matter to the early church. We cannot hold ancient manuscript culture to the same standard as our modern print culture. This is not to say that we cannot have confidence in the text\u2014may it never be! While we do not possess many early copies of Philemon, most textual variation is easily explainable in terms of correcting awkward syntax. In fact, Philemon 14\u201317 contains relatively little textual variation.
\nThe question of the text of the New Testament is not settled. More research and analysis needs to be completed. The study of manuscripts goes beyond the reconstruction of the initial text. Textual variants themselves can be windows into beliefs, or at the very least into how readers have read the text across the centuries. As more and better images become available, more data can be mined than has ever before been possible. The field of New Testament textual criticism is alive and well, with more detailed study to be done. The manuscripts are plentiful, but the laborers are few.<\/p>\n
Key Takeaways
\n\u25ba There are more than 570 manuscripts of Philemon, and only three of them are complete manuscripts written before circa AD 700. If we had only manuscripts copied more than nine hundred years after Paul wrote Philemon, our text of Philemon would change very little.
\n\u25ba An examination of all Greek witnesses to the text of Philemon has revealed that the textual tradition contains more variants than some might expect, even if most of them have no serious merit.
\n\u25ba There are really only two places where the initial text may be uncertain: \u201cto us\u201d or \u201cto you\u201d in Philemon 6, and the presence or absence of and in Philemon 11.
\n\u25ba Textual variations that aren\u2019t original can still help us to understand how the text was understood in some settings.<\/p>\n
CHAPTER TEN<\/p>\n
MYTHS ABOUT VARIANTS<\/p>\n
WHY MOST VARIANTS ARE INSIGNIFICANT AND WHY SOME CAN\u2019T BE IGNORED<\/p>\n
Peter J. Gurry<\/p>\n
The New Testament text has come down to us in thousands of copies with many more thousands of textual differences between them. No one disputes this basic fact. It presents an apparent problem, however, for those of us who believe that the New Testament is foundational for faith in Jesus Christ. After all, what we know of Jesus we know primarily through the New Testament writings. If we cannot know something as basic as what those writings say about Jesus, then there is little point in further debating whether what they say is true. Because of this, the textual differences between our witnesses to the New Testament text have been a recurring concern in debates about Christianity\u2019s viability.
\nReading the recent literature, however, brings one into contact with two very different opinions of how many and how important the differences between our New Testament manuscripts are. Take, for instance, Bart Ehrman, who began to think in seminary that the New Testament cannot be God\u2019s word because \u201ceven if God had inspired the original words, we don\u2019t have the original words.\u201d From this he concluded, \u201cThe doctrine of inspiration was in a sense irrelevant to the Bible as we have it, since the words God reputedly inspired had been changed and, in some cases, lost.\u201d In short, Ehrman concluded that there are too many variants to believe the Bible\u2019s claims about its own inspiration.
\nOn the other hand, it is common to find Christian authors who believe that textual variation poses no threat to Christian belief in the inspiration of Scripture since \u201cfor over 99 percent of the words of the Bible, we know what the original manuscripts said\u201d and \u201cfor most practical purposes, then, the current published scholarly texts of the Hebrew Old Testament and Greek New Testament are the same as the original manuscripts.\u201d Sometimes this near-perfect percentage is applied in the reverse so that it is 99 percent of all the variants that are unimportant rather than 99 percent of the original words that are known accurately, a slightly different metric. Either way, the impression is that, despite Ehrman\u2019s doubts, textual variants present no real challenge at all to Christian confidence in the Bible. In the words of Matthew Barrett, any textual uncertainties raised by our disparate manuscripts are \u201calways in matters insignificant, never having to do with Christian doctrine or the credibility of the biblical text.\u201d
\nWhich of these competing views is true? Is the Bible\u2019s inspiration a moot point because the original text is lost in the mists of time, or do we have nothing to worry about because our modern editions match the original text in 99 percent of cases? In this chapter, we will show why both claims are overstated. In the first case, it is true that a large majority of our vast number of variants really are trivial for modern Bible readers; but we also hope to show why giving the impression that no variants matter for Christian doctrine gives an equally false impression. Some variants, despite being tucked away in Bible footnotes or commentaries, really do touch on important doctrines and so cannot be ignored by Christians who treasure the Bible as the Word of God.
\nThe first matter of concern, then, is to consider what seem like wild claims about the number of variants in our manuscripts and some helpful (and not so helpful) ways to put this estimated number in context. Next, we consider whether these variants affect Christian doctrine and practice before concluding with some reflections on how these variants do matter.<\/p>\n
THE NUMBER OF VARIANTS<\/p>\n
The first point to consider when thinking about whether textual variants affect the Christian faith is how many we have. Over the centuries, New Testament scholars have given many suggestions about how many variants there are in our manuscripts. Some of the earliest printed editions of the Greek New Testament marked variants in the margin, but the number recorded was kept quite small. This changed most dramatically with the publication of John Mill\u2019s momentous edition of 1707, a work of thirty years that recorded something like 30,000 textual variants. A century after Mill and Richard Bentley, Frederick Nolan could claim that further works since Mill\u2019s day had added another 100,000 variants to our knowledge. Another century would add yet another 100,000 estimated variants, until our own day, when the most common number is given as 400,000. At least one noted textual critic has even suggested a number as high as 750,000. To put this in perspective, some like to note that there are only around 138,020 words in the standard Greek New Testament, leaving us with more variants than actual words from which they vary.
\nA better estimate. The problem with all these estimates is that we are not told where they come from. Ehrman, for example, is the first prominent scholar to raise the number to 400,000, but when it comes to telling us how he arrived at this number he, like others, attributes the number to nameless \u201cscholars\u201d and offers no further detail. If we take the most comprehensive collections of variants available, however, it is possible to arrive at a more responsible estimate.
\nWe now have nearly complete and accessible collations of our Greek manuscripts for at least three sections of the New Testament: John 18, Philemon, and Jude. If we count the number of variants in these sections and divide by the number of words in the standard Greek New Testament, we arrive at the rate of variation. The results are 3.86 individual variants per word for John 18, 3.53 for Philemon, and 3.67 for Jude. Multiply these rates by the 138,020 words in the same standard Greek New Testament, and you arrive at between 488,220 and 533,584 variants. Since we are dealing with an estimate here, it is probably best to leave it as a round half-million nonspelling variants in our Greek manuscripts.
\nEach qualification matters. This estimate does not include spelling differences, or variants found only in other sources, such as patristic citations of the New Testament or early translations into Syriac, Latin, and the like. This is due to limitations both in our source data and in the very nature of these other witnesses. Versions, for example, can be notoriously difficult to translate back into the Greek, and the varied ways in which the church fathers cite the New Testament makes cataloging and counting their variants difficult too. We should also note that our estimate counts \u201cvariants\u201d that could be the original reading. This is because it counts variants among our manuscripts rather than variants from the original text, since the latter is, after all, the very thing we are trying to identify in the first place.
\nInterpreting the estimate. Now that we have a reliable estimate, the question remains as to what to make of it. Like any good statistic, the estimated number of variants requires context, and several have been offered. Many of those, however, are more than a little problematic. In our introduction, we saw that some scholars compare the total estimated variants in all our manuscripts to the total number of words in some particular edition of the Greek New Testament. While this makes for a dramatic comparison, it actually tells us very little because it uses a fixed number on one side of the equation (words in one particular New Testament) with a variable one on the other (variants in all our manuscripts). Others have instead tried comparing the number of variants to the number of manuscripts to provide perspective. Here again, we have a weak comparison since manuscripts vary widely in terms of their size and contents. Some of our New Testament manuscripts are, after all, hundreds of pages long and contain the entire New Testament (such as Sinaiticus), while others are mere fragments containing a single verse (such as P12). A comparison that suggests that both manuscripts have the same number of variants is pretty unhelpful.
\nWhat both of these comparisons miss is that scribes create variants from a text only by making additional copies of it. The more copying, the more variants, generally speaking. But wrapped in this curse is a blessing because, counterintuitively, more manuscripts mean the potential for more data with which to resolve their differences. Such was the point made already by Richard Bentley in response to concerns about Mill\u2019s estimated thirty thousand variants published in 1707. Bentley noted that the more manuscripts the better since \u201cby a joint and mutual help all the faults may be mended; some copy preserving the true reading in one place, and some in another.\u201d In other words, the very cause of the problem (copying) can also supply us with its solution (more copies).
\nFortunately, we can go well beyond Bentley\u2019s general principle because we have far more\u2014and far more precise\u2014data available than in his day. For example, if we compare the number of variants in the Greek continuous-text manuscripts of John 18 with the number of estimated words in those same manuscripts, then we get a good picture of just how many unique variants scribes created for each word that they hand copied. Since we have 3,058 variants in 1,659 manuscripts of John 18, and since the average copy of John 18 has roughly eight hundred words, then the result is just one distinct variant per 434 words copied, or 3,058 distinct variants out of 1,300,000 words. If we cut out the 1,360 of these variants that are nonsensical, we are left with 1,698 meaningful textual variants. That, of course, is still a large number for professional textual critics to wade through when deciding which of these are original in the hundreds of places where a decision needs to be made. (Most of them will be quite easy to decide, of course, either because they are attested by so few manuscripts or because they are such obvious corruptions.) But, from the scribes\u2019 perspective, only one new variant per 434 words copied is pretty good when you realize that every one of these came with the opportunity to create a new, otherwise unattested reading.
\nEven with these variants, we still need to ask just how many of these variants in John 18 are important to the meaning. To gain some sense of that, consider that the NA edition, which is designed for academic study, includes 154 variant readings in this chapter of John. The UBS edition, designed specifically for Bible translators, includes just ten. The more recent Tyndale House Greek New Testament lists twelve variant readings. If we consider a few well-regarded commentators on John, we note that D. A. Carson discusses just three variants, and C. K. Barrett eight. How many of these deserve the notice of typical English Bible readers? Based on the major modern translations, the answer is zero. Not a single note about any variant in John 18 is to be found in the ESV, NIV, NRSV, or even the richly footnoted NET. The translators of these versions are right in this since all of the major variants in John 18 are easily resolved, do not significantly affect the meaning, or both.
\nNaturally, other chapters of John and other books will differ, but this example illustrates the basic point that only a tiny sliver of the total number of variants are really worth a translator\u2019s time, let alone the average Bible reader\u2019s. By my calculations, only about 0.3 to 2.8 percent of all our estimated variants are in the UBS edition designed for translators. And while some of these are left out from editorial oversight or out of sheer necessity, the number of variants that might meaningfully affect translation cannot be far from this.
\nConjecture. Before we consider some specific variants, it is worth mentioning how rarely editors of the Greek New Testament have to guess or \u201cconjecture\u201d what the original text is\u2014in other words, how rarely editors have concluded that all our manuscripts are wrong and the original text is simply not to be found in any of them. Such conjecture is fairly common practice among editors of classical works and, it should be said, has long been practiced by New Testament editors too. Yet when we consider the mainstay scholarly edition of the Greek New Testament, the NA, we see a decreasing use of conjecture over time. In the thirteenth edition published in 1927, editor Erwin Nestle (son of Eberhard Nestle) introduced eighteen conjectures in the apparatus that he thought \u201cmust be considered original.\u201d Contrast that to the NA, released in 2012, where we find only two conjectures the editors think are original (one at Acts 16:12 and one at 2 Pet 3:10). Other editors, such as those of the Tyndale House edition, reject all conjectures on principle, although there remains debate about the justification for this. Still, even the eighteen conjectures in Nestle\u2019s thirteenth edition bear witness to the remarkable reliability of the New Testament text. The original reading is usually, if not always, there somewhere in our witnesses; the challenge, where such exists, is in settling convincingly on where exactly it is.<\/p>\n
SOME DIFFICULT AND IMPORTANT VARIANTS<\/p>\n
Despite what has been said, many protest that merely counting variants does not tell the whole story. Kenneth Clark, for example, writes, \u201cCounting words is a meaningless measure of textual variation, and all such estimates fail to convey the theological significance of variable readings.\u201d Likewise, Ehrman says of his self-styled \u201corthodox corruptions\u201d that their \u201csignificance cannot simply be quantified; it is pointless, for example, to calculate the numbers of words of the New Testament affected by such variations or to determine the percentage of known corruptions that are theologically related.\u201d
\nTo be sure, we do not need to go so far as to say that counting is meaningless, especially when such statistics are put in proper context. But it is true that variants, like manuscripts, should be weighed and not merely counted. In light of this, let us sample the most difficult and important variants in the New Testament. This should provide us with a more realistic sense of whether textual variants pose a serious threat to the Christian faith in general or inspiration in particular.
\nDefinitions. When thinking about the significance of textual variants, it is helpful to keep two categories in mind. The first is whether the variant is important for interpretation. By all accounts, most variants do not affect the meaning of the text. This obviously applies to spelling differences (does it change the meaning if we spell John\u2019s name with one or two nus in Greek?), but it also applies to many other smaller variants, which merely make the implicit explicit or the ambiguous clear. These types of variants occur throughout our manuscripts, and any major edition of the Greek New Testament will bear this out on page after page. These types of variants pose no threat to the Christian faith or to the Bible\u2019s inspiration. They merely show that scribes or readers were at times willing to make the text read more clearly.
\nAn example may help. In Acts 13:33, there is a knotty problem in Paul\u2019s speech at Pisidian Antioch. There Paul refers to the fulfillment of God\u2019s promises in Jesus by saying that \u201cwhat God promised to the fathers\u201d has now been fulfilled in the resurrection. But the people to whom that promise is fulfilled is not so clear. It is either \u201cto us, their children,\u201d or \u201cto our children,\u201d or perhaps \u201cto us, the children.\u201d The first and last of these make the most sense in context. The second is altogether awkward since we wouldn\u2019t expect the fulfillment of this promise to be among the children of Paul\u2019s audience. The problem is that the first reading is the latest attested, the second is the earliest, and the third is not attested at all\u2014it\u2019s a conjecture. Regardless of how we resolve this particular issue, this variant is one that does affect Paul\u2019s precise meaning, but also one that in no way affects the importance of the resurrection, still less the fact of the resurrection. No one would be foolish enough to suggest that because this verse has a variant, the truth of the resurrection itself is in peril.
\nWhat we are interested in, then, is variants that are genuinely difficult to resolve and that have some level of bearing on the text in a way that might affect Christian claims. Let us now turn to a few illustrative examples.
\nMark 1:1. In the very first verse of what is thought by most to be the very first Gospel, we find a difficult and significant variant. The Gospel opens with what reads like the title to the whole book: \u201cThe beginning of the gospel of Jesus Christ, Son of God.\u201d The variant involves the words \u201cSon of God,\u201d as some important witnesses omit them. In Mark\u2019s Gospel, Jesus\u2019 sonship is an important theme, highlighted at his baptism (Mk 1:11), then again when he confronts unclean spirits (Mk 3:11), at the transfiguration (Mk 9:7), at his trial before the Sanhedrin (Mk 14:61), and climactically on the lips of a Gentile centurion (Mk 15:39). The issue, then, is not whether Mark presents Jesus as the Son of God but whether Mark wants us to read his account of the good news about Jesus with this in mind from the first line on.
\nIn the standard critical editions (the UBS and NA), the words are placed in brackets indicating that \u201ctextual critics today are not completely convinced of the authenticity of the enclosed words.\u201d This scholarly uncertainty is confirmed when we consider the two other most recent critical editions. The SBL Greek New Testament omits the words completely, whereas the Tyndale House Greek New Testament has them without brackets. Within the UBS, the rating given to this reading is a C, indicating that the editors had difficulty deciding. The reason is easy to see by going no further than the first manuscript listed in our apparatus, Codex Sinaiticus (01). This important manuscript has both readings, the shorter reading found as the first-written text (01*), with the longer reading found there too as a correction by one of the original scribes of the manuscripts. Both readings can also be found in early Christian authors. Origen, for example, attests the shorter reading and Irenaeus the longer on several occasions. Our earliest witness to this portion of Mark is actually an amulet from the late third or fourth century (P.Oxy. 76.5073), and it does not have the longer reading. Most of the manuscript and versional evidence, it must be said, comes down on the side of the longer reading.
\nWhat makes this variant particularly difficult is the contradictory evidence for how each reading might have originated in the first instance. On the one hand, it is not hard to imagine scribes copying the shorter reading and, either out of reverence for Jesus or knowledge of Mark\u2019s Gospel, adding the words \u201cSon of God.\u201d On the other hand, the series of letters in this first verse of Mark make an accidental omission easy to explain because six of the words end in the same letter. When the names for Jesus are written as abbreviations, or nomina sacra, and without spaces (\u0399\u0305\u03a5\u0305\u03a7\u0305\u03a5\u0305\u03a5\u0305\u03a5\u0305\u0398\u0305\u03a5\u0305), it is quite easy to imagine two of them being dropped by accident. Scholars call such an omission homoioteleuton (literally, \u201csimilar endings\u201d) because the similar endings were the cause.
\nSome have objected to this explanation, however, because they note how the nomina sacra are designed to give words special attention. Moreover, since this is the start of the book, the suggestion is that a scribe would be most alert at just this point, probably having taken a break just before starting here.
\nWhile this may have been true in some cases, we actually have clear examples of just this omission happening at the start of Mark\u2019s Gospel in some later manuscripts. One such example is a manuscript housed in Ferrara, Italy. GA 582 (or Manuscript Biblioteca Comunale Cl. II, 187, III) is a fourteenth-century complete New Testament copied by a scribe with a particularly bad penchant for leaving things out by accident. When I visited the manuscript in July 2016, I counted more than 130 omissions that were subsequently added in the margin. These were places where the scribe had left text out only to later discover his error and correct it. Of these 130 omissions, nearly 60 percent are cases where homoioteleuton is an obvious cause. Not surprisingly, we find two such cases on the very first page of Mark, one in the very first line, where the phrase \u201cson of God\u201d (\u03c5\u1f31\u03bf\u1fe6 \u03c4\u03bf\u1fe6 \u03b8\u03b5\u03bf\u1fe6, huiou tou theou) is clearly omitted from the main text and just as clearly added back in the margin. In other words, neither the use of nomina sacra nor the start of a book was enough to keep this scribe from accidental omission. And what happened in this manuscript could certainly explain the omission in 01* and other manuscripts.
\nWhatever the original reading is here (and I think the evidence points toward the longer reading), this variant is, by all accounts, a difficult and important variant. Certainly it is one that English translations should continue to note for their readers. Furthermore, it is one that Christians cannot ignore when closely reading Mark\u2019s Gospel even if, in the full analysis, the sonship of Jesus is not what\u2019s at stake.
\nLuke 23:34. A second and even more difficult and important variant than Mark 1:1 involves one of the Bible\u2019s best-known sayings. In Luke 23:34, as Jesus is crucified between two criminals, he utters words that are as astonishing as they are famous, \u201cFather, forgive them, for they know not what they do.\u201d With these words, Jesus models the response to persecution that he commands for his followers (Lk 6:28) and that his followers will later emulate (Acts 7:60) and further commend (1 Pet 2:21\u201323).
\nThe problem, textually, is that in some very early and important manuscripts, Jesus does not pray this remarkable prayer at all; the words are simply not there. This is true in our earliest manuscript, P75 (second to third century) and in Codex Vaticanus (fourth century), which, when agreeing with P75, likely attests a text that reaches back to the second century or earlier. Here, however, the omission is not explainable by similar endings, as in Mark 1:1. Once again Codex Sinaiticus attests both readings. In this case, the original scribe wrote the longer reading, a second scribe marked the prayer for erasure, and a third tried to erase the erasure! Codex Bezae (fifth century) shows a correction too, with the first scribe omitting the prayer and a later, sixth-century scribe adding it in the bottom margin of the page.
\nThe reading is not found in the Latin text of 05 either originally or by correction. This along with several other majuscules (032, 038, 070, 0124), two minuscules (579, 1241), and then some of the earliest Syriac, Latin, and Coptic translations amounts to all the evidence for the omission of the prayer. On the other side of the ledger, the prayer is found in most of our witnesses, of which the earliest and most important Greek manuscripts are 01*, 02, and 04.
\nThe evidence from the second century shows that the verse was known in its longer form by Irenaeus (Haer. 3.18.5), while it is not clear whether Marcion\u2019s version of Luke contained it. There is also a tantalizing statement of the same prayer attributed to James the brother of Jesus at his martyrdom, by a second-century church father named Hegesippus. The account is recorded for us only later in the fourth century by Eusebius (Hist. eccl. 2.23.16). The problem is that we do not know whether this was the source for Luke 23:34 or whether Luke 23:34 was the source for the prayer that Hegesippus (per Irenaeus) attributes to James.
\nThe manuscript evidence is somewhat split, although the agreement of P75, 03, and 05* weighs heavily in most text critics\u2019 minds. No doubt, this explains the double brackets in NA\/USB5 and the confidence of the UBS committee in giving their decision an A rating. The strongest argument in favor of the prayer\u2019s omission was stated as long ago as 1881, when Westcott and Hort wrote, \u201cWilful excision, on account of the love and forgiveness shown to the Lord\u2019s own murderers, is absolutely incredible: no various reading in the New Testament gives evidence of having arisen from any such cause.\u201d In addition, it is often pointed out that without the prayer, the text flows smoothly from the soldiers crucifying Jesus (Lk 23:33) to casting lots for his garments (Lk 23:35).
\nOn the other hand, Jesus\u2019 prayer sits comfortably with Luke\u2019s second volume, wherein Stephen prays a similar but not identical prayer as he is stoned (Acts 7:60). If Luke 23:34 were added under the influence of Stephen\u2019s prayer, we might expect their wording to agree more closely. In addition, the theme of ignorance that we see in Jesus\u2019 prayer (\u201cfor they know not what they do\u201d) is one that Luke returns to multiple times in Acts (Acts 3:17\u201319; 13:27; 17:30). Such ignorance does not imply innocence, since the prayer is for forgiveness. Rather, the point seems to be that the soldiers (and the Jews?) do not realize the significance of who they are crucifying or how God intends to reverse their evil for a far greater good (see Chrysostom, Homiliae in epistulam i ad Corinthios 7.5).
\nUltimately, the decision must be made based on which reading best explains the origin of the other. (This is the fundamental principle of all New Testament textual criticism.) If the prayer is not original, then where did it come from? If the prayer is original, why would anyone want to remove it? One answer to the first question is that it comes from Stephen\u2019s prayer. But we have already noted that the wording is not as close as we would expect if that were the case. On the other hand, it could be suggested that the prayer is authentic to Jesus but not to Luke. In this scenario, the saying was a \u201cfloating tradition\u201d that eventually found its way into Luke. But, if that were the case, we are left to explain why it was only ever added to Luke and not to any of the other Gospels.
\nTo the second question\u2014why it would be removed\u2014many answers have been given, all having to do with the prayer\u2019s apparent theological problems. The theological problems in question, we must remember, are not ours but those of the early church. They range from the fact that (1) the prayer seems to go unanswered given that God judged the Jews by destroying Jerusalem in AD 70, (2) Jesus offers forgiveness to the unrepentant, (3) the verse implies unfairness to those acting in innocence, or (4) that anti-Semitism made some early Christians dislike such mercy being offered to the Jews. Christian writers certainly did write about these larger issues in general, and in some cases these concerns are even hinted at when discussing this particular prayer. It must be said, however, that these very same discussions show that Christian writers could easily address their concerns about the prayer without excising it from Luke. In other words, what these explanations for excision forget is that actual commentary\u2014rather than textual editing\u2014was the Christians\u2019 method of choice (then as now) for dealing with problem passages.
\nIn the end, a decision on this variant is extremely difficult. The early and weighty support of P75, 03, and 05 is hard to ignore on the one side. On the other side, it really is difficult to explain where this prayer in this context came from if not from Luke himself. Whichever the original reading, this is another case of a variant like Mark 1:1 that Christian readers cannot ignore.
\nJohn 1:18. Having discussed two difficult and important variants, perhaps we might balance our discussion with the less interesting\u2014but, just for that reason, important\u2014places where no difficult variants are found in the New Testament. Take, for instance, the famous beginning of John\u2019s Gospel, where the deity of Christ is set forth with unique clarity and power. In this passage, we get one famous textual variant in John 1:18 concerning whether Jesus is called the \u201conly-begotten God\u201d (\u03bc\u03bf\u03bd\u03bf\u03b3\u03b5\u03bd\u1f74\u03c2 \u03b8\u03b5\u03cc\u03c2, monogen\u0113s theos) or the \u201conly-begotten Son\u201d (\u03bc\u03bf\u03bd\u03bf\u03b3\u03b5\u03bd\u1f74\u03c2 \u03c5\u1f31\u03cc\u03c2, monogen\u0113s huios) who is at the Father\u2019s side. The difference is one letter when written as nomina sacra (\u0398\u0305\u03a3\u0305 vs. \u03a5\u0305\u03a3\u0305).
\nIn isolation, this variant is quite significant, as it seems to be a choice between Jesus\u2019 divinity and his unique sonship. But, of course, this variant does not occur in isolation. It comes only after John\u2019s rich theological introduction, in which he makes clear that the preexistent Logos is divine (Jn 1:1) and became incarnate (Jn 1:14). In none of these first fourteen verses is there a theologically significant variant. In fact, these verses are so textually stable that they agree word-for-word between the very first published Greek New Testament (Erasmus\u2019s, published in 1516) right up to the most recent (the Tyndale House edition published in 2017). There is not even a single letter different between them. These two editions span hundreds of years and are based on very different manuscripts and editorial principles. Whichever reading is original, then, Jesus is clearly divine in the introduction to John\u2019s Gospel. Numerous other nonvariant passages that we might call theologically load-bearing could be cited.
\nThese three variants are certainly not the only ones in the Gospels, but they are illustrative. It is important to stress that most books of the New Testament have only a handful of variants that combine this level of importance and difficulty. Others we could mention in the Gospels occur at Matthew 12:47; 19:9; 21:29\u201331; 24:36; 26:28; Mark 1:2; 16:9\u201320; Luke 2:14; 10:1, 17; 11:1\u20134; 22:43\u201344; and John 5:3\u20134; 7:53\u20138:11. Moving outside the Gospels, we find such variants in Acts 20:28; Romans 5:1; 14:23\/16:25\u201327; Ephesians 1:1; 2 Thessalonians 2:7; 1 Timothy 3:16; Hebrews 2:9; 2 Peter 3:10; and Jude 5. It is worth noting that there is no attempt to hide these variants. They are plainly visible right in the footnotes of most any modern English translation of the Bible. Additionally, their merits are discussed in places such as Bruce Metzger\u2019s textual commentary cited above, the \u201cTC\u201d notes of the NET Bible, and, of course, in the more detailed Bible commentaries. In other words, there is no conspiracy of silence about them; they are well known.<\/p>\n
VARIANTS AND DOCTRINE<\/p>\n
These examples along with the data about the estimated number of variants in our manuscripts illustrate the problem we face in discussing them fairly. On the one hand, it is clear that most variants really are easily resolved or ignored. On the other hand, we have seen examples in Mark 1:1 and Luke 23:34 where variants simply cannot be ignored. How, then, should we best present the evidence?
\nSometimes the impression from the apologetic literature is that variants do not matter at all. Others are more careful to claim only that \u201cno orthodox doctrine or ethical practice of Christianity depends solely on any disputed wording.\u201d Daniel Wallace is even more precise, admitting that some \u201cnoncentral\u201d beliefs or practices seem to be affected by viable variants but that \u201cno viable variant affects any cardinal truth of the New Testament.\u201d Both qualifications (\u201cviable\u201d and \u201ccardinal\u201d) are important and match what we have here called difficult and important variants. In this sense, Wallace is surely right that no core Christian doctrine (e.g., the resurrection, the deity of Christ, salvation, the Trinity) is based solely on a textually difficult passage. Even Bart Ehrman grants that his own view is not a problem for this conclusion. He has said publicly that his view is not at odds with that of his mentor, Bruce Metzger, which is that \u201cessential Christian beliefs are not affected by textual variants in the manuscript tradition of the New Testament.\u201d
\nThat is not to say, however, that no passage that addresses or touches on core doctrines is textually suspect. Some certainly are, such as 1 John 5:7\u20138, which says in the King James Version, \u201cThere are three that bear record in heaven, the Father, the Word, and the Holy Spirit, and these three are one.\u201d That is as explicit a definition of the Trinity as one finds in the Bible. Yet no serious textual critic today accepts this reading as authentic, and neither do evangelical theologians, who are still quite able to make a biblical case for the doctrine of the Trinity. In other words, the fundamental doctrine of the Trinity in no way depends on this variant reading even though the variant in question certainly addresses that doctrine. As for variants touching on matters of Christian practice, we might mention the text-critical debates over Romans 16:7 and 1 Corinthians 14:34\u201335 and the bearing they have for some on the question of women\u2019s ordination.
\nIn light of such cases, we cannot claim without qualification that variants never affect texts that touch on Christian doctrine or practice. Sometimes they clearly do. Yet no one would claim that an issue such as the Trinity or the ordination of women is hanging in the balance because of these disputed texts. It would be better to say, then, that no Christian doctrine or practice\u2014major or minor\u2014is determined by a textually difficult passage.
\nThe reason for this will be obvious to anyone with even a basic knowledge of Christian theology. That reason is that Christians do not base their theology on a single verse here or there, let alone a single word or two within them. Instead, theology at its best is built on a web of biblical evidence\u2014that is to say, on the \u201cwhole counsel of God\u201d (Acts 20:27). As theologian John Frame notes in discussing the impact of textual variants on theology, \u201cScripture is highly redundant, in a good way,\u201d such that \u201cthe doctrines of the Christian faith are never derived from a single text.\u201d In this way, when one passage on the Trinity is suspect, many others rush in to take its place. The thick web of theology is not destroyed for lack of one strand that turns out not to be silk. In the case of 1 John 5:7, for example, it is helpful to realize that the church\u2019s most important doctrinal statements of the Trinity, such as the Nicene Creed, were produced completely without reference to this text for the simple reason that it did not exist in Greek manuscripts when these statements were formulated. Richard Bentley noted exactly this point as far back as 1717, writing that if 1 John 5:7 is not original, then it follows that \u201cArianism in its Height was beat down without the Help of that Verse: And let the Fact prove as it will, the Doctrine [of the Trinity] is unshaken.\u201d
\nHaving said this, few if any would go on to claim that Jesus\u2019 prayer from the cross in Luke 23:34 has no effect whatsoever on our theology or practice. To be sure, it does not change the fact that Jesus teaches us to love our enemies (Mt 5:43\u201348) and to forgive them (see Mt 6:14\u201315; 18:21\u201335). But just as certainly it does have some effect on how we think about this truth, how we apply it, and how we teach it. Likewise, we may be content to say that the story of the woman caught in adultery (Jn 7:53\u20138:11) is not the only place where we see Jesus\u2019 mercy on full display or that the key elements in Mark 16:9\u201320 are recorded elsewhere, but would any go so far as to say that these texts have no effect on our teaching, preaching, or Christian living? The very fact that they continue to draw so much interest tells against such a conclusion. Clearly, then, they have some effect, even if it is small when put in perspective of the whole of Christian faith and practice.<\/p>\n
CONCLUSION<\/p>\n
In the final analysis, it is best to admit that, in relatively rare cases, variants really do have some bearing on some doctrines and ethical practices of the Christian faith, but none of these doctrines or ethical practices is established from these disputed texts. Nor are any of them in jeopardy because of these disputed texts. Mark 1:1 is a good example where the variant matters for how we read Mark\u2019s Gospel, but the Sonship of Jesus himself is not riding on this variation, not even in Mark. The same could be said of Jesus\u2019 ethical teaching on forgiveness with respect to Luke 23:34 and on his divinity in the case of John 1:18. Jesus clearly wants us to forgive our enemies in the New Testament, and he is just as clearly presented as divine with or without these important readings. Furthermore, as we saw in John 18, the vast majority of variants, upward of 99 percent, are awfully boring for most Bible readers, are easily resolved, or both. The exceptions, which do exist and should not be ignored, are nevertheless few and far between.
\nWe are safe, then, to claim that neither the Christian faith nor the Bible\u2019s inspiration is threatened by textual variants. The words of Stephen Neill from a half-century ago remain true: \u201cIndeed, I think it is no exaggeration to say that the very worst Greek manuscript now in existence \u2026 contains enough of the Gospel in unadulterated form to lead the reader into the way of salvation.\u201d If the very worst manuscript can do this, then we are that much more secure in having New Testaments based on the very best of our Greek manuscripts.
\nIt is true that our Bibles could change slightly in the future as more research sheds greater light or as methods of textual criticism change. It is also true that knotty and important textual decisions remain and we should not ignore these. Yet because of the overall fidelity of scribes over fifteen hundred years and because of the Herculean efforts of textual scholars for hundreds beyond that, we can sing the words of the eighteenth-century hymn:<\/p>\n
How firm a foundation, ye saints of the Lord,
\nis laid for your faith in His excellent Word!
\nWhat more can He say than to you He hath said,
\nto you who for refuge to Jesus have fled?<\/p>\n
Key Takeaways
\n\u25ba The estimated number of variants in just our Greek manuscripts is around half a million, not including spelling differences. Nearly half of these are meaningless mistakes.
\n\u25ba Only a tiny fraction of the known variants are ever discussed by commentators. Fewer still deserve a footnote in a modern translation. In John 18, not one of the more than three thousand variants makes it into a footnote of the ESV, NIV, NRSV, or even the NET.
\n\u25ba It is true, then, that most variants do not affect the meaning of the text or the Christian faith in general. A few dozen do, however, and some of these are theologically important, as in Mark 1:1; Luke 23:34; and John 1:18.
\n\u25ba We should not give the impression that New Testament variants do not matter at all for Christian theology or practice; we can and should, however, recognize that no doctrine is in jeopardy because of a serious variant.<\/p>\n
CHAPTER ELEVEN<\/p>\n
MYTHS ABOUT ORTHODOX CORRUPTION<\/p>\n
WERE SCRIBES INFLUENCED BY THEOLOGY, AND HOW CAN WE TELL?<\/p>\n
Robert D. Marcello<\/p>\n
When dealing with any debated topic, it is easy to discuss it in terms that favor a certain position. The same is true when it comes to the reliability of the text of the New Testament. Almost everyone who is critically engaged in the topic now has access through online and text resources to the same data. However, the packaging of that data often changes how one perceives the details. And the fact is, packaging matters. With the proliferation of interest in this topic, particularly among those not engaged in critical study, packaging matters even more. It matters because many misappropriate this information and draw conclusions that are unjustifiable given the evidence. One particular example of this is the area of scribal changes or corruptions of the text. A common saying is, \u201cWe can\u2019t trust the text of the New Testament, because scribes edited it.\u201d As Kurt Eichenwald puts it, \u201cScribes added whole sections of the New Testament, and removed words and sentences that contradicted emerging orthodox beliefs.\u201d
\nYes, scribes did introduce corruptions, or intentional changes of the text for theological reasons, into the manuscript tradition of the text. There is ample evidence to support that conclusion. However, what about the first part of that slogan? Is the corruption that we find in the New Testament manuscript tradition so pervasive that it renders the text untrustworthy? Are these corruptions that significant? These later implications are often a part of the packaging of the data. If the data are presented in certain ways, it is easy to draw such conclusions. At the same time, it is also easy for the data to be dismissed by apologists who also have theological motivations to tamper with the reality of the text\u2019s history. This chapter will investigate just that issue by showing some real examples of theologically motivated textual changes (or \u201corthodox corruptions\u201d) along with some not-so-real examples of such corruptions. In working through these examples, hopefully one can gain an accurate appreciation for the history of the text along with the ability to articulate what is at stake in this discussion so that the packaging matches the facts.<\/p>\n
BACKGROUND TO THE QUESTION<\/p>\n
When it comes to corruptions, one of the most significant commentators on the topic was scholar F. J. A. Hort, who deals with the subject briefly in the appendix of his classic work, coauthored with B. F. Westcott, The New Testament in the Original Greek. Here he states,<\/p>\n
It will not be out of place to add here a distinct expression of our belief that even among the numerous unquestionably spurious readings of the New Testament there are no signs of deliberate falsification of the text for dogmatic purposes. The licence of paraphrase corruption, where scribes allowed themselves to change language which they thought capable of dangerous misconstruction; or attempted to correct apparent errors which they doubtless assumed to be due to previous transcription; or embodied in explicit words a meaning which they supposed to be implied \u2026 in a word, they bear witness to rashness, not to bad faith.<\/p>\n
The opinion that scribes did not change the text for dogmatic purposes, here summarized by Hort, became a common perspective in determining textual decisions within the New Testament. Westcott and Hort\u2019s position became even more important because of the publication of their Greek New Testament, which was a standard text and used widely by subsequent scholars. However, Hort also clarifies, \u201cIt is true that dogmatic preferences to a great extent determined theologians, and probably scribes, in their choice between rival readings already in existence.\u201d The distinction seems to be that scribes did not invent new readings for dogmatic purposes but may have been inclined to favor an already existing reading that corresponded with an orthodox interpretation of the text. While Hort did not see a role for intentional dogmatic changes, many others have seen that as a possible feature of transmission and argued against Hort\u2019s view in particular.
\nIt appears Hort\u2019s view on this issue was short-lived among specialists even though his text continued to be widely used. In fact, many scholars came to the conclusion that scribes did in fact alter the text for doctrinal purposes. Even with this change in course, most scholars still come short of ascribing motive to the scribes\u2014a key distinction. In fact, some go in the opposite direction and indicate a positive motive on the part of scribes. For example, Metzger comments on a specific change in his Textual Commentary on the Greek New Testament, saying that at times some changes were \u201cintroduced from a sense of reverence for the person of Jesus.\u201d As such, the debate began a slow transition from the belief that scribes did not invent readings for dogmatic purposes to a growing body of evidence that indicates that scribes may have in fact done just that.
\nKenneth W. Clark\u2019s comments on the concept of intentional corruption are significant here because he distinguishes between the scribes\u2019 actions and their intentions. Here he states,<\/p>\n
We can agree with Hort that \u201cperceptible fraud\u201d is not evident in textual alteration, that \u201caccusations of wilful tampering \u2026 prove to be groundless,\u201d and that dogma has not motivated \u201cdeliberate falsification.\u201d But these are heinous faults such as we should never allege, and these are not the terms that we should employ. Willful and deliberate, yes. But not tampering, falsification, and fraud. Alteration, yes; but not corruption. Emendation, yes; but not in bad faith. These denials of evil or unethical intention can well be sustained, but such intention is not a proper allegation by the textual critic. He must analyze the text constructively to understand the theological value of any variation, and its place in historical theology.<\/p>\n
Clark indicates that the type of alteration found within the New Testament is not of a malicious nature, intended to invent a specific reading. While it is true that dogmatic preferences inclined scribes to prefer the textual option that they deemed most orthodox from among the available readings, the nature of that influence was not intended to create a knowingly false text but rather a good-faith clarification. Thus, Clark distinguished the falsification of readings from good-faith choices, thereby vindicating the scribes\u2019 motives for most scribal alterations.
\nOthers have argued that there is indeed clear-cut dogmatic influence within the text, and it was intentional. Eldon Epp provided a thorough study of Codex Bezae in Acts. There he showed that in 40 percent of the variants there was \u201cthe unmistakable result\u2014a clearly anti-Judaic tendency in the D-text of Acts.\u201d More and more, the motivation of the scribe came into question, since now there were clear examples of a theological influence on textual alteration. This changed the subsequent landscape of research into the text of the New Testament in general and scribal alterations in particular. Bart Ehrman, in his landmark work The Orthodox Corruption of Scripture, took this issue head-on and then popularized much of the information in Misquoting Jesus. In his study he presents four key christological issues of orthodoxy\u2014antiadoptionistic corruptions, antiseparationist corruptions, antidocetic corruptions, and antipatripassianist corruptions. Through these categories, he demonstrates how theological controversies found their way into the transmission history of the text. He argues, \u201cScribes occasionally altered the words of their sacred texts to make them more patently orthodox and to prevent their misuse by Christians who espoused aberrant views.\u201d Basing his argument on the thesis of Walter Bauer about early Christianity, which states that there were multiple forms of Christianity competing for what would later be considered orthodoxy, Ehrman situates the scribes within the same framework. In so doing, he provided the first lengthy study of theologically motivated alterations and demonstrated that such alterations were a legitimate category for understanding the transmission of the text.
\nOne voice providing further entry into this topic is Wayne Kannaday. He was a student of Ehrman and provides a study outlining pagan criticisms along with the rebuttals of early apologists. He then compares those with scribal alterations in early Christianity. Kannaday concludes, \u201cIn the course of defending their text, their faith, and their Lord against pagan assault the scribes engaged in what I have termed \u2018scribal apologetics.\u2019 In so doing, they repaired and renovated the Gospel handed down to them, making it to pagan readers and critics more palatable, on the one hand, and more resistant to challenge, on the other.\u201d Kannaday\u2019s point is to demonstrate that scribes were agents in defending orthodoxy\u2014not only those contemporary apologists.
\nDavid C. Parker is another voice who has gone further in describing the scribal contributions to the text of the New Testament. Parker provided a provocative thesis when he stated, \u201cThere is no definitive text to be discovered.\u201d By declaring that it would be a fruitless\u2014even unattainable\u2014enterprise to determine an early definitive text, he shifts the focus of the text-critical endeavor to an emphasis on scribal habits rather than the original text. In doing so, he focuses on explaining why scribes would come up with certain readings, since the text was not a fixed document but a \u201cliving\u201d entity. As such, Parker removes the concept of an original text from the equation and indicates that scribes, within a theological community, produced variants, and all variants were products of their community. Thus, as he states elsewhere, the pursuit to find a \u201cpure tradition\u201d is unsuccessful, as both heretical and orthodox corruptions were introduced into the text. The text critic\u2019s task ought to be understanding the scribes within their context over against the pursuit of an original text form.
\nAll of these scholars have contributed to the conversation about theologically motivated corruptions in one form or another. They provide the foundation for understanding how the discussion has evolved and why certain claims about the reliability of the text are currently circulating. Equally important is a discussion of method. In order to have a robust understanding of the text\u2019s reliability, one must have a fair method that accounts for all the evidence in the manuscript tradition. The following section will address this issue specifically.<\/p>\n
METHOD MATTERS<\/p>\n
Ascertaining a motive for anyone\u2019s actions from their writing is difficult even when considering contemporary authors. Doing so for someone living in a time and place in history with unknown influences, unknown sources, and an often-unknown identity makes the task even bleaker. It is true that some have been able to offer helpful observations on a scribe\u2019s tendency throughout a manuscript. However, at times there seems to be a general assumption among scholars that scribes were well intentioned even in the midst of adopting a theological reading. When apologetic intentions are declared, such claims are met with ample caution since those claims often are simply lacking in sufficient evidence for such a historical claim. This is especially true when variant readings are read in isolation from the scribe\u2019s general tendency throughout an entire manuscript. Certainly, reconstructing the social context of a scribe becomes helpful since understanding contemporary debates provides a further window into the reason for a scribe\u2019s actions.
\nHowever, many questions arise out of such an endeavor. Did a reading arise because a scribe was influenced by apologetic purposes for orthodoxy or for heresy? Without studying the scribe\u2019s actions throughout a manuscript, it is difficult at best to make a claim one way or the other. Does the scribe consistently alter the manuscript or demonstrate any pattern of alteration? Is there a consistent theological tendency, or is the author picking data points to undergird a thesis? The latter seems to be the case when it comes to Ehrman\u2019s method. One of the key weaknesses of his study is that his examples of scribal corruptions are often isolated from discussions about the manuscript in which they are contained. As Wallace notes regarding Ehrman\u2019s Misquoting Jesus, \u201cFirst, there is next to no discussion about the various manuscripts. It is almost as if external evidence is a non-starter for Ehrman. Further, as much as he enlightens his lay readers about the discipline, the fact that he does not give them the details about which manuscripts are more trustworthy, older, and so on, allows him to control the information flow.\u201d This pattern creates a false understanding of the scribal actions and raises doubts about Ehrman\u2019s conclusions in general. One clear issue of method is that one must have a good understanding of the quality and nature of the manuscripts containing the alteration.
\nThis leads to another methodological issue, which can often be seen as a basic element of Ehrman\u2019s Orthodox Corruption. Consistently, the least orthodox reading is thought to be original, and the orthodox reading is deemed the corruption. While it may be the case that the orthodox reading was the corruption, using that assumption for one\u2019s method presupposes what the original text was, what the reading was in the manuscript\u2019s ancestor, that the scribe had intention to alter the text and implemented it, and many other factors. All of these dynamics are largely unknown in the majority of available manuscripts\u2014resulting in a real lack of evidence to support such a premise. If one does not note the scribe\u2019s tendency throughout a witness, and if one consistently views the orthodox reading as the corruption, then the conclusions one is likely to draw are questionable. Again, this does not deny that there are real cases of dogmatic influences on the text; however, the issue here is one of method and how one is able to determine such influences. Accordingly, the implicit working notion\u2014that the least orthodox reading is preferred\u2014is flawed due to lack of historical evidence to support such a claim.
\nFurthermore, a truism of textual criticism by Westcott and Hort, which has stood the test of time, is that \u201cknowledge of documents should precede final judgment on readings.\u201d Knowing the ancestor from which a manuscript was copied would, of course, tell us whether a variant was simply adopted from its exemplar or whether it was generated by the scribe. Obviously, one is rarely privy to such information in the manuscript tradition; however, claims about dogmatic motivation at a point in history should be held to the same strict standard of historiography as other historical claims. This strict standard does not always seem to be followed consistently in discussions of scribal motivations or claims regarding corruption. Knowing a scribe\u2019s source provides invaluable information for reconstructing part of the reason for a variant. Without such knowledge, in particular when we are missing other pieces of evidence, strong claims of probability are thereby weakened considerably. This is the reason that a robust method that takes into account a multitude of factors such as those discussed above is essential when claiming that a scribe corrupted the text.
\nWhile one must admit that there are specific examples of corruption within the manuscript tradition of the New Testament, the way scholars have determined whether a text is corrupt and the level of probability of that corruption is often overplayed. As one can see, certain data and methodological considerations must be taken into account to provide a more accurate picture of scribal activity. Without such data or when an investigation lacks such considerations, the results are also open to criticism. Ehrman, while not alone in his claims of corruption, provides an example of this type of investigation, since there are some weaknesses in his method. That he has popularized the concept and delivered it in a well-packaged book opens his work up to specific criticism, as it is the work with which most readers are familiar. Nevertheless, an honest discussion also acknowledges that, while his claims may be too broad and his methodology at times weak, the central thesis that there are indeed orthodox corruptions within the manuscript tradition remains true. Therefore, the following will examine both aspects of this\u2014pointing out real and not-so-real examples of corruption.<\/p>\n
EXAMPLES OF VARIOUS KINDS OF CORRUPTIONS<\/p>\n
As has been said, there are good examples within the available manuscripts, and that reality has allowed certain scholars to find highly likely examples of individual dogmatic changes. This section is not meant to provide an exhaustive list of examples but to demonstrate what textual corruptions look like and how they affect our understanding of the text. Also, while it is true that scribes changed the text of the New Testament for dogmatic reasons, the practice was not nearly as widespread or prevalent as some indicate. Therefore, it is important to look at some good examples that have been called corruptions but that have compelling counterexplanations.
\nCases of real corruption and their significance. It is easy to dismiss a concept that makes theology messy or more difficult. If scribes corrupted and not merely preserved the text, then can anything really be trusted? Apologists, pastors, teachers, and academics may struggle with how to deal with this question, and in doing so, some have dismissed the concept of scribal corruptions prematurely and without all the evidence. What follows are two examples of well-accepted corruptions within the manuscript tradition. They provide a window into the world of ancient scribes and show how they were motivated by theological influences and adjusted their texts accordingly.
\nScribal corruption in Codex Bezae. One of the clear examples of dogmatic influence on the text of a manuscript comes from Eldon Epp\u2019s groundbreaking work on Codex Bezae. Within this work, Epp explains how what he calls the Western D-text of Acts is a prime sample to mine for theological tendency. Furthermore, the study\u2019s method is good because it restrains itself to a single early manuscript\u2014one that has long been known for its unique character and textual makeup of the so-called Western text. One of the reasons Epp chose this text as a key one to mine for examples was the work of P. H. Menoud, who stressed that the Western text\u2019s emphasis is \u201cthe newness of the Christian faith as regards Judaism.\u201d
\nEpp picks up on this and takes the variant tradition of the Western \u201cD\u201d text of Acts further than the \u201cmore notable variants\u201d and also considers smaller variants \u201cin an effort to determine whether or not they support the interests discerned in the larger and better-known D-readings.\u201d As noted, his conclusion is that 40 percent of the variants taken into consideration have an \u201canti-Judaic\u201d bias. He clarifies what is meant by the term anti-Judaic by saying that it is used broadly to describe<\/p>\n
that religious complex out of which Christianity arose and contemporary with the earliest period of the new faith.\u2026 [It] involves the concept of Israel as the distinctive and exclusive people of God and also, at times, refers to the official religious system, including the regulations, customs, and institutions of both \u201cPalestinian\u201d and \u201cDiaspora\u201d Judaism, though cult does not largely figure into this study.<\/p>\n
His study finds that these variants \u201creveal this tendency directly, are contributory to it, or at least comport with such a viewpoint.\u201d Metzger notes, \u201cSome of the evidence Epp adduces is certainly impressive.\u2026 [Acts] 2:23; 3:13\u201314; 13:28\u201329, involve more or less plain examples of variant readings in which the Jews\u2019 relationship to the condemnation of Jesus to death is more sharply described in the Bezan text, which places greater emphasis on the Jews\u2019 responsibility and on their hostile attitude toward him.\u201d
\nThese examples are just some of the numerous instances evaluated by Epp demonstrating this recurring theme. Here, not only do the variants seem to contain a specific bias but also considering (1) the amount of them, (2) the fact that both significant and insignificant variants function the same way, and (3) the universal nature of the bias throughout the manuscript, one is hard-pressed to argue that the scribe of Bezae does not have such a bias. This example is a clear illustration of a study demonstrating scribal corruption at work within a specific manuscript and by a specific scribe.
\nP72 and theological tendencies. P72, or what has been referred to as the Bodmer Miscellaneous Codex because of its multiple seemingly unrelated works bound together in the same codex, is a late third- to-mid-fourth-century codex. The codex has been discussed with regard to its theological tendencies and the significance of those tendencies in recent years. These examinations have shown that there is good reason to view its variants as having a high Christology and stressing the divinity of Jesus. It seems that these changes also reflect some of the contemporaneous dialogue happening within proto-orthodox circles in regards to heretical claims. Therefore, not only do the variants consistently demonstrate a high Christology, and do so across books (Jude and 1-2 Pet), but they also do so right in line with what would be expected within their communities given the concurrent theological discussions.
\nOne such example is a reading from 2 Peter 1:2, where the scribe omits the conjunction and (\u03ba\u03b1\u03af, kai) between \u201cof God and Jesus our Lord\u201d (\u03c4\u03bf\u1fe6 \u03b8\u03b5\u03bf\u1fe6 \u03ba\u03b1\u1f76 \u1f38\u03b7\u03c3\u03bf\u1fe6 \u03c4\u03bf\u1fe6 \u03ba\u03c5\u03c1\u03af\u03bf\u03c5 \u1f21\u03bc\u1ff6\u03bd, tou theou kai I\u0113sou tou kyriou h\u0113m\u014dn). In doing so, the scribe creates a reading where the syntax now equates Jesus and God. Earlier in the codex, at 1 Peter 5:1, the scribe also substitutes the word God where every other manuscript except for one uses the word Jesus. Such cases make it clear that the scribe of this manuscript is undoubtedly providing a high Christology. The combination of a scribe\u2019s tendency throughout an entire manuscript along with solid specific examples and ample social implications offer a clear illustration that models an effective method.
\nCases of not-so-real corruption. While there are some clear examples of corruption, the practice is easily exaggerated when it comes to actually demonstrating instances of corruption. In what follows one can see two not-soreal examples illustrating various reasons to question the claim that they model theologically motivated corruptions.
\nMatthew 24:36 and the ignorance of the Son. One commonly cited example of orthodox corruption is the exclusion of the phrase \u201cnor the Son\u201d in Matthew 24:36 in a discussion on who knows the time of Jesus\u2019 return. The inclusion of the phrase has been widely accepted as being original, due in large part to its being well attested in some of the earliest manuscripts. Also, with the phrase removed there are grammatical issues, and some have argued that a scribe would be more likely to delete the phrase than to add it based on theological reasons. A phrase that would seem to question the omniscience of Christ would obviously be suspect to a scribe. Some argue that this theological bias would have driven scribes to remove the phrase from the text, thereby eliminating the dilemma.
\nHowever, this raises significant questions regarding scribal motivations and actions. First, why would scribes remove the phrase \u201cnor the Son\u201d but leave the word \u201calone\u201d in Matthew\u2019s text? The two cases functionally accomplish the same thing in regard to meaning. If the Father alone knows the time of Christ\u2019s return, then tacitly the Son does not. Second, in almost all instances where the text is found in a manuscript containing both Matthew and Mark, Mark\u2019s text is left unaltered, whereas Matthew is missing the phrase. Why wouldn\u2019t the same scribe omit a christologically significant phrase from both Gospels if the purpose of the alteration were theological? Further, the longer reading suffers from grammatical and redactional issues as well. These problems raise significant questions about the originality of the longer reading. Even if it were original, the assumption that the shorter reading was theologically motivated becomes suspect considering the presence of the phrase in Mark and the use of \u201calone\u201d in the same verse. Wallace has demonstrated that it is most likely Matthew, not a later scribe, who removed the phrase and added \u03bc\u03cc\u03bd\u03bf\u03c2 (monos, \u201conly\u201d), thereby softening the concept of the Son\u2019s ignorance. This follows the general pattern of how Matthew uses Mark\u2019s Gospel. The point is that this text, while purported to be a significant example of theologically motivated corruption, does not marshal the support often attributed to it.
\n\u201cGod\u201d or the \u201cSon\u201d in John\u2019s prologue? Another important text for this discussion is John 1:18. This text has been used to support the concept of a theologically motivated corruption created to combat the belief that Jesus was adopted as God\u2019s son at some point during his life, or what is referred to as adoptionism. Here the question revolves around which variant was original\u2014either \u03bc\u03bf\u03bd\u03bf\u03b3\u03b5\u03bd\u1f74\u03c2 \u03b8\u03b5\u03cc\u03c2 (monogen\u0113s theos, \u201cthe only God\u201d) or \u03bc\u03bf\u03bd\u03bf\u03b3\u03b5\u03bd\u1f74\u03c2 \u03c5\u1f31\u03cc\u03c2 (monogen\u0113s huios, \u201cthe only son\u201d). In some witnesses, the form of the text with \u03b8\u03b5\u03cc\u03c2 (theos, \u201cGod\u201d) also can be found with the article. Ehrman has argued that \u201cthe only Son\u201d was original and that the alternate variants were \u201ccreated to support a high Christology in the face of widespread claims, found among adoptionists.\u201d The debate on the original reading largely boils down to the question of whether \u201cGod\u201d or \u201cSon\u201d is the correct reading. The inclusion of the article seems to be a scribal clarification since the lack of the Greek article is a harder reading. The inclusion of the article could have been due to a harmonization with other passages (Jn 3:16, 18), and the shorter reading has strong support. Central to claims of corruption is the idea that the text was changed to support a specific theological point of view.
\nHowever, reconstructing that social context is difficult at best with little literature to support such claims. If one is going to claim that a scribe was making an intentional change to combat a specific heresy in the early church, then one must demonstrate the contemporary relevance of that heresy within the cultural or theological dialogue of that specific scribe. Just because a heresy existed at the time of writing, it does not mean a scribe both had it in mind and was arguing against it. It is this point Benjamin Burkholder\u2019s work addresses by demonstrating that the early church did not use this verse in the manner that Ehrman describes. Rather,<\/p>\n
It was employed to substantiate their notion that God was transcendent and the one in his bosom made the knowledge of God possible. When \u03bc\u03bf\u03bd\u03bf\u03b3\u03b5\u03bd\u1f74\u03c2 \u03b8\u03b5\u03cc\u03c2 [monogen\u0113s theos, \u201conly God\u201d] does become a litmus test for orthodoxy, it occurs at such a late date that it cannot aid in determining how the reading came into existence.<\/p>\n
He further shows that the way in which the verse was used in early church writings would not entice a scribe to make an alteration to the text, thereby undercutting the claim for a theologically motivated corruption.
\nNot only does this specific claim of corruption struggle on social-historical grounds, it also struggles to provide the best explanation of the internal evidence. One key issue here is a grammatical one. Ehrman argues that an insurmountable issue for the \u201cthe only God\u201d reading is what it could mean in the first century. Particularly, he addresses the \u201centirely implausible\u201d understanding that the adjective \u03bc\u03bf\u03bd\u03bf\u03b3\u03b5\u03bd\u03ae\u03c2 (monogen\u0113s, \u201conly\u201d) is being used as a noun, which would produce the reading \u201cthe only one, himself God.\u201d Nevertheless, this seemingly impossible situation is entirely possible for the New Testament writers. Wallace provides numerous examples of the very construction Ehrman claims does not happen. In doing so, the internal argument for \u03bc\u03bf\u03bd\u03bf\u03b3\u03b5\u03bd\u1f74\u03c2 \u03b8\u03b5\u03cc\u03c2 (monogen\u0113s theos, \u201conly God\u201d) becomes stronger. This is particularly true when viewed in light of the fact that (1) \u03bc\u03bf\u03bd\u03bf\u03b3\u03b5\u03bd\u03ae\u03c2 is used as a noun a few verses earlier (Jn 1:14) and (2) the reading \u201cunique God\u201d makes little sense in both the immediate content of the prologue\u2014equating Jesus and the Father\u2019s divine status in John 1:1\u2014and John\u2019s broader content in the Gospel. Finally, John\u2019s return to referring to the Father from John 1:1 would form an inclusio, bracketing the prologue and providing focus for his readers. All of these internal clues provide further support to the manuscript evidence that \u201cthe only God\u201d is the reading that best explains the evidence. As such, the claim that this text is a good example of orthodox corruption is severely undercut.<\/p>\n
CONCLUSION<\/p>\n
The text of the New Testament does not come down to our generation without years of history. That history, like all of history, has aspects of corruption in it, which many would like to forget. When packaging this history and presenting it to interested audiences, we do them no service by denying or whitewashing these facts. Equally, by overblowing the significance of corruptions or the prevalence of the practice, we do no service either. The reality is that there are some clear instances of corruption within the text\u2019s history that can be determined with a responsible method. Also, there are instances where individuals have claimed that the text is corrupt when there is ample evidence to the contrary. These claims do not accurately reflect a text that, while it does have certain instances of corruption, has a history that is remarkably stable. The actual facts do not need to be whitewashed for claims of the text\u2019s reliability to stand. In fact, it is with a solid methodology that we are able to determine that corruptions do exist. If we can determine an intentional change has been made by some theologically motivated scribe, then we can conversely also determine what the text looked like prior to that change by evaluating the alternate readings.
\nWe are not, then, left wondering whether such corruptions have happened everywhere; we can pinpoint where they did and why they arose. We are not left with a hopelessly corrupt text but a textual tradition that, while including corruptions, includes the genuine text. We need to be fair to the data and to our audiences by not creating a perception that the history of the text is unscathed. When such statements are easily debunked, the result is often to throw out the good with the bad. Likewise, by packaging the data as a hopeless enigma, we overstate the severity and pervasiveness of the theological influence within the manuscript tradition. Both extremes must be avoided, and those who are charged with packaging this data for audiences must be faithful to the facts.
\nWhile it is true that the text was sometimes changed for theological reasons, instances of this practice are few and far between. What is more important is that, by the same process that we are able to identify such changes, we are also able to identify the original text. While some packaging may make this task appear bleak, the reality is the ability to uncover the original text enables us to continue to trust the reliability of the text of the New Testament.<\/p>\n
Key Takeaways
\n\u25ba Scribes did sometimes change the text for theologically motivated reasons; however, not a few textual variants that might appear to be theologically motivated are better explained by other factors.
\n\u25ba Determining the intention behind textual variants is much more difficult than some surmise. This means we should be appropriately skeptical about bold claims of theologically motivated variation or \u201corthodox corruption.\u201d
\n\u25ba The best approaches to determining when this happened are based on a knowledge of an entire manuscript and the context in which it was made, not simply on isolated occurrences from otherwise unrelated manuscripts.<\/p>\n
CHAPTER TWELVE<\/p>\n
MYTHS ABOUT PATRISTICS<\/p>\n
WHAT THE CHURCH FATHERS THOUGHT ABOUT TEXTUAL VARIATION<\/p>\n
Andrew Blaski<\/p>\n
In their now-classic work The Text of the New Testament, Bruce Metzger and Bart Ehrman put forth the following proposition regarding the use of patristic scriptural quotations in the field of textual criticism:<\/p>\n
Besides the textual evidence derived from New Testament Greek manuscripts and from early versions, the textual critic has available the numerous scriptural quotations included in the commentaries, sermons, and other treatises written by early church fathers. Indeed, 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.<\/p>\n
Of these two sentences, the first is certain, while the second is wholly ambiguous. What, after all, constitutes an \u201cearly\u201d church father? What constitutes a \u201ccitation\u201d? What does \u201cpractically\u201d the entire New Testament entail? In recent decades, variations of this argument have made their way into countless Christian apologetic books aimed at defending the integrity of the New Testament text, but to what degree is the claim true? To what degree is it helpful? We will explore these questions and more in this chapter, ultimately exposing what has become one of the longest-standing myths in the field of text criticism.<\/p>\n
A MYTH OF EPIC PROPORTIONS: LORD DALRYMPLE AND THE MISSING ELEVEN VERSES<\/p>\n
Despite their statement\u2019s problems, Metzger and Ehrman have actually presented the tamer version of a much bolder and far more prevalent claim. To this we turn first. In the Baker Encyclopedia of Christian Apologetics, under the heading \u201cNew Testament Manuscripts,\u201d Norman Geisler states, \u201cIf we compile the 36,289 quotations by the early church fathers of the second to fourth centuries, we can reconstruct the entire New Testament minus 11 verses.\u201d Contrary to the ambiguity we saw above, it is precisely the specificity of this claim that lends it its apparent validity. But where did the claim originate? Geisler was not the first to employ it, nor was he the last. The greatest challenge in assessing its veracity is therefore locating its source. It can be found in a multitude of popular-level apologetic volumes, but if one is fortunate enough to find an accompanying citation at all, the citation will lead typically to yet another popular-level apologetic putting forth the same unsubstantiated declaration. There appears to be no hard data whatsoever. Yet the claim has become so widely accepted and so frequently employed by apologists in recent decades that very few have ever stopped to ask the obvious question: Is it true?
\nPerpetuating the problem is the peripheral nature of the claim. Rarely, if ever, has it been used as a central argument for establishing the trustworthiness of the New Testament text. Rather, apologists consistently employ it as a kind of add-on argument, as if to say, \u201cEven if these other impressive arguments don\u2019t convince you, we could also reconstruct the entire New Testament text (minus eleven verses) from patristic quotations alone.\u201d Perhaps for this reason above all the claim has gone unchecked over the years. It tends to serve as more of a buffer for other, better arguments than as a key argument in its own right. Where did the claim originate, then? The story is a fascinating one and is itself a prime example of how easy it is to misconstrue or misremember a person\u2019s words or deeds over time.
\nIn 1841 Scottish author Robert Philip published a book titled The Life, Times, and Missionary Enterprises of the Rev. John Campbell, which, as suggested, narrates the life of John Campbell (1766\u20131840), a prominent Scottish philanthropist and missionary to South Africa. In his book Philip includes a number of lengthy and detailed quotations from Campbell\u2019s personal journals and correspondences, giving it an overall autobiographical quality and a wealth of firsthand accounts. The importance of this lies in the fact that one of the many journal entries reveals the origins of our \u201cmyth.\u201d In a section titled \u201cAnecdote of Lord Hailes,\u201d Campbell narrates an event that, in Philip\u2019s words, \u201chad much influence in satisfying [Campbell\u2019s] own mind upon the perfection of the New Testament.\u201d It is worth reading in full:<\/p>\n
I remember distinctly an interesting anecdote referring to the late Sir David Dalrymple, (better known to literary men abroad by his title of Lord Hailes,) a Scotch judge. I had it from the late Rev. Walter Buchanan, one of the ministers of Edinburgh. I took such interest in it, that though it must be about fifty years since he told it, I think I can almost relate it in Mr. Buchanan\u2019s words:
\n\u201cI was dining some time ago with a literary party at old Mr. Abercrombie\u2019s, (father of General Abercrombie who was slain in Egypt, at the head of the British army), and spending the evening together. A gentleman present put a question, which puzzled the whole company. It was this: \u2018Supposing all the New Testaments in the world had been destroyed at the end of the third century, could their contents have been recovered from the writings of the first three centuries?\u2019 The question was novel to all, and no one even hazarded a guess in answer to the inquiry.
\n\u201cAbout two months after this meeting, I received a note from Lord Hailes, inviting me to breakfast with him next morning. He had been of the party. During breakfast he asked me, if I recollected the curious question about the possibility of recovering the contents of the New Testament from the writings of the three first centuries? \u2018I remember it well, and have thought of it often without being able to form any opinion or conjecture on the subject.\u2019
\n\u201c \u2018Well,\u2019 said Lord Hailes, \u2018that question quite accorded with the turn or taste of my antiquarian mind. On returning home, as I knew I had all the writers of those centuries, I began immediately to collect them, that I might set to work on the arduous task as soon as possible.\u2019 Pointing to a table covered with papers, he said, \u2018There have I been busy for these two months, searching for chapters, half chapters, and sentences of the New Testament, and have marked down what I have found, and where I have found it; so that any person may examine and see for themselves. I have actually discovered the whole New Testament from those writings, except seven or eleven verses (I forget which), which satisfies me that I could discover them also. Now,\u2019 said he, \u2018here was a way in which God concealed, or hid the treasures of this world, that Julian, the apostate emperor, and other enemies of Christ who wished to extirpate the gospel from the world, never would have thought of; and though they had, they never could have effected their destruction.\u2019
\n\u201cThe labor in effecting this feat must have been immense; for the gospels and epistles would not be divided into chapters and verses as they are now. Much must have been effected by the help of a concordance. And having been a judge for many years, a habit of minute investigation must have been formed in his mind.\u201d<\/p>\n
Here we have the origin of our myth, and quite a myth it is. By Campbell\u2019s own account, Lord Dalrymple, a local judge engaging in a bit of amateur biblical research, communicated the results of that research to Walter Buchanan over breakfast, who then (sometime later) passed on a version of those results to John Campbell, who proceeded to record them in his journal fifty years after the fact. This is, of course, not to mention that the journal entry itself comes to us via Robert Philip\u2019s book (yet another degree of separation). Remarkably, Campbell is not even certain about the number of \u201cmissing\u201d verses (seven or eleven)\u2014one of two critical pieces of information that have successfully made their way into modern apologetic texts.
\nAs for the other, is it true that there are exactly 36,289 quotations of the New Testament in the works of the second- to fourth-century fathers? This number, it seems, did not originate with Dalrymple but rather from the work of another nineteenth-century figure: John William Burgon, dean of Chichester Cathedral. In a quest not altogether different than Dalrymple\u2019s own, Burgon created a sixteen-volume index of patristic biblical quotations taken from Christianity\u2019s first three to four centuries, which is now held in the British Museum. Though the work has not been published, Frederic Kenyon included the numerical results of Burgon\u2019s study in his own Handbook to the Textual Criticism of the New Testament. Taken together, the number of quotations he separately lists for Justin Martyr, Irenaeus, Clement of Alexandria, Origen, Tertullian, Hippolytus, and Eusebius add up to exactly 36,289 (hence the very specific number). However, Kenyon himself is highly critical of Burgon, noting that the editions Burgon relies on are \u201ccomparatively uncritical texts of the Fathers.\u201d Indeed, the majority of editions at that time were reliant on a few comparatively late manuscripts.
\nThe claim, then, appears to be a conflation of two unsubstantiated, unpublished nineteenth-century studies, and yet it has appeared in literally dozens of current apologetic texts, with even the most basic questions left unanswered: Which editions of the fathers were these men using? Which edition of the New Testament? How did they define a quotation? Why did they focus on those particular writers and not others? Where can I find the original studies? Still, what is perhaps most striking about this myth, or rather the prevalence of this myth, is that it stands in direct opposition to what biblical scholars have long known and taught regarding the attestation (or lack thereof) of entire books of the New Testament in the first two to three centuries of Christianity\u2019s growth. The epistle of James is a prime example. James was rarely cited in the earliest centuries, was not included in the famous Muratorian Canon (second to third century), and was even listed as a \u201cdisputed\u201d book by Eusebius of Caesarea as late as the mid-fourth century (Hist. eccl. 3.25.1\u20136). This was due in part to its apparent lack of influence in the Christian literature of the time. Though there is evidence that it inspired portions of the Shepherd of Hermas, it is not until the third century that we find any direct quotations. Yet the claim that the entirety of the New Testament can be reconstructed from the writings of this period alone (minus eleven verses) still circulates with regularity.
\nAs far back as 1905, a committee of the Oxford Society of Historical Theology set out to \u201cprepare a volume exhibiting those passages of early Christian writers which indicate, or have been thought to indicate, acquaintance with any of the books of the New Testament.\u201d The committee limited its search to the authors sometimes called the \u201cApostolic Fathers\u201d (a loosely given title generally signifying the second generation of Christian writers after the apostles), and employed a four-tiered grading system of A, B, C, and D to indicate the degree of probability that an author was citing or even referring to a New Testament text. Between the twenty-seven books of the New Testament and the eight works chosen by the committee (the Epistle of Barnabas, the \u201ctwo ways\u201d portion of the Didache, the remainder of the Didache, 1 Clement, 2 Clement, the letters of Ignatius, the writings of Polycarp, and the Shepherd of Hermas), there were 216 points of possible overlap. The results were telling. Out of those 216 possibilities, less than 3 percent were given an A rating, indicating certain knowledge of a New Testament book. Conversely, more than 60 percent of the possible points of overlap received no rating, indicating no (visible) knowledge whatsoever. The point is not that these writers must have been unaware of any works they did not explicitly mention but rather that it is often highly difficult to positively prove that they were aware (making Campbell\u2019s anecdote all the more impossible).
\nChristian apologists must remember that perpetuating baseless claims such as the one above will ultimately only provide their critics with additional ammunition. In 2006, for example, an Islamic apologetic organization known as Islamic Awareness took it on themselves to get to the bottom of Campbell\u2019s anecdote. Not surprisingly, they found hard evidence that the claim had become greatly exaggerated throughout the course of its transmission. Beginning with the assumption that Dalrymple had indeed recorded the results of his research (as suggested by Campbell), they discovered that a multitude of Dalrymple\u2019s unpublished papers were being held at the National Library of Scotland. Among those papers, they found catalogued thirteen notebooks on the Greek New Testament, as well as a number of loose-leaf folios, which collectively serve to reveal the exact timing and extent of Dalrymple\u2019s work. What they ultimately uncovered was evidence of a study that spanned a time period of about four years (1780\u20131784), consisting of at least four different attempts (in three collations) to discover which portions of the New Testament could be reconstructed using only the patristic writings of the first three centuries. Each attempt appears to draw from different sources (with some overlap), to contain different emphases, and in some cases to examine different books of the New Testament. Presumably, Dalrymple intended eventually to synthesize the totality of his results into a single work worthy of publication, but this synthesis never came to fruition.
\nA number of these findings stand in contrast to Campbell\u2019s report, including the timespan of Dalrymple\u2019s work (four years rather than two months), but chief among them is the number of \u201cmissing verses.\u201d After methodically charting the number and percentage of missing verses in each of Dalrymple\u2019s efforts (available for viewing online), members of Islamic Awareness came to the conclusion, \u201cEven if we admit the best-case scenario of the least number of missing verses in each of the books of the New Testament as seen in the three collations, we obtain 4336 verses (~54%) absent in the Patristic citations of the New Testament.\u201d
\nRegardless of the accuracy of their numbers, these apologists knew that the \u201celeven missing verses\u201d claim had no basis in reality (and they showed it). The irony in what they found, of course, is palpable. In transmitting Dalrymple\u2019s claim that the original wording of the New Testament is partially attainable through patristic quotations, Campbell ended up altering the original wording of the claim itself. It seems we can put this form of our myth to rest.<\/p>\n
QUALIFYING THE ARGUMENT<\/p>\n
What, then, if the argument is modified, or better defined? Daniel Wallace, for example, has made claims similar to those above but adds that the patristic quotations at our disposal \u201cdate as early as the first century and continue through the thirteenth century.\u201d At first glance, this seems far more promising. Surely, adding a full millennium of Christian writings onto the question posed to Lord Dalrymple helps its overall cause. However, Wallace\u2019s more realistic argument demands an additional inquiry: If the title of \u201cancient church father\u201d can be applied to writers living as late as Thomas Aquinas (ca. 1225\u20131274), what meaning or importance does the argument even retain? Surely, the only reason the claim was ever compelling was that the writings of the earliest fathers were composed within a few generations of the apostles themselves and that they were contemporaneous with some of the earliest extant New Testament manuscripts. Indeed, more time elapsed between Jesus and Thomas Aquinas than between Thomas Aquinas and Daniel Wallace. While Wallace\u2019s argument is therefore more realistic, it is simultaneously less valuable. Herein lies the fundamental problem with the argument as a whole, whether as expressed by Metzger, Wallace, or anyone else. As soon as we begin to give actual content to the terms of the argument, such as \u201cearly,\u201d \u201cchurch fathers,\u201d \u201cpractically,\u201d or even \u201cquotations,\u201d the argument ceases to function.
\nThe word quotations is a prime example. Textual critics must be able to recognize a quotation as a quotation in order to make use of it. How, then, is a quotation defined? One of the more traditional methods has been to divide patristic scriptural references into one of three categories: citations (direct quotations, perhaps with a very small degree of variation), adaptations (significantly modified scriptural references), and allusions (references containing only reminiscences of the original wording). As an example, we can look to Origen of Alexandria (ca. 185\u2013254) and his famous Commentary on the Gospel of John. In the course of that work, Origen refers to John 14:9 (\u201cHe who has seen me has seen the Father\u201d) multiple times, and in more than one context. The following are three examples:<\/p>\n
It is through this mirror that Paul and Peter and their contemporaries see God, because [the Savior] says, \u201cHe who has seen me has seen the Father who sent me.\u201d (13.153)<\/p>\n
And perhaps because they saw the image of the invisible God, since he who has seen the Son has seen the Father, they have been recorded to have seen God and to have heard him, in that they have perceived God and heard God\u2019s words in a manner worthy of God. (6.19)<\/p>\n
On the one hand, insofar as he who has seen the Son has seen the Father who sent him, one sees the Father in the Son. (20.47)<\/p>\n
In their analysis of these passages, Bart Ehrman, Gordon Fee, and Michael Holmes have determined that the first example is a citation, the second is an adaptation, and the third is an allusion. But how? The only way for them to be certain (aside from instances in which the writer explicitly states that he is quoting from a manuscript) is to compare these references to what is found in the manuscript tradition. Of course, scholars may disagree with Ehrman, Fee, and Holmes over the definition of a quotation, or over the number of possible categories (some include categories such as locution, or even lemma), but by necessity, any method of identification must rely at least in part on comparison. As the editors of the Editio Critica Maior recognize, \u201cA true quotation is one where the wording of the Father\u2019s text is identical with a reading found in the manuscript tradition.\u201d If, then, we were to attempt a reconstruction of the entire New Testament using only the scriptural \u201cquotations\u201d of the church fathers, we would first have to determine which of those quotations are in fact actual quotations, or citations, rather than adaptations, allusions, or other types of references. Doing so requires a trustworthy point of comparison, and so the original argument falls flat once more. A line is judged to be crooked because it fails to conform to a straight line. Similarly, the chief way to determine that a scriptural quotation is a quotation (or not) is to compare it to the New Testament manuscripts we already possess.
\nWe can see that, on further reflection, each of the terms of the original argument breaks down. However, we may wish to ask one further question: Why are there so many inexact quotations and references in the writings of the Fathers? Why is it so difficult to identify an actual quotation? Certainly, faulty memory and a scarcity of accessible manuscripts played an enormous role, but is there more to it than that? As we shall see, there are certain fundamental differences between modern Christian perceptions of the New Testament (and its interpretation) and the perceptions held by some of the early Christian writers mentioned above. Those differences affect the very manner in which the church fathers quoted Scripture and therefore equally affect our ability to make use of their quotations.<\/p>\n
A PATRISTIC THEOLOGY OF QUOTATIONS<\/p>\n
Among the major conceptual differences between early and modern Christians is the very notion of a concrete and precisely defined \u201csomething\u201d called \u201cthe New Testament.\u201d This is not to say that patristic writers did not recognize the existence of authoritative apostolic texts or that they did not perceive these authoritative texts to be \u201cScripture\u201d or \u201ccanonical\u201d from a very early date. Rather, in Christianity\u2019s earliest centuries, the New Testament did not exist as a set of twenty-seven authoritative books in a definitively closed list. There was a certain fluidity, even within the boundaries Christians increasingly came to recognize. On the face of it, this may not seem relevant to the way in which an early Christian writer would cite Scripture, but contrasted with today\u2019s norms (especially among evangelical Protestants), the relevance soon becomes clear.
\nWhen a canon is closed, particularly in a religious context, the totality of texts within that canon can eventually come to be regarded as itself the object of contemplation. The impact of this process differs from tradition to tradition, but a closed canon by definition engenders a greater mental divide between oneself and the canonical texts. The canonical texts become equivalent to the whole of the truth rather than the authoritative set of lenses through which one enters into a larger truth. As a result, one often ceases the attempt to carry on in the same tradition as the authoritative books, or to use those books as authoritative tools, and begins instead to seek out ways to penetrate the barrier between oneself and the canonized books, whether historical, geographical, cultural, or linguistic.
\nThe point, simply, is this: for Christianity\u2019s earliest thinkers, the mental divide between text and interpreter was not as great as it tends to be today, which (as we shall see) allowed those thinkers to be more free and even playful with the text itself. Such things cannot be reduced purely to the \u201cclosing\u201d of the canon, but because many early Christians lacked a totality of authoritative words, the object of Christian contemplation in the earliest centuries tended to be Christ the authoritative Word. He is clothed in language but not bound by it. Though stated in overly simplistic terms, this distinction between text and person is arguably visible even within the pages of the New Testament. John Behr, a prominent patristics scholar, puts it this way:<\/p>\n
Thus, in the material which comes to be collected together as the canonical New Testament, reflection on Christ is an exegetical enterprise. But, it is very important to note that it is Christ who is being explained through the medium of Scripture, not Scripture itself that is being exegeted: the object is not to understand the \u201coriginal meaning\u201d of an ancient text, as in modern historical-critical scholarship, but to understand Christ himself, who, by being explained \u201caccording to the Scriptures\u201d becomes the sole subject of Scripture throughout\u2014he is the Word of God.<\/p>\n
Regardless of the accuracy of Behr\u2019s statement (no doubt many biblical scholars would object), this is an accurate depiction of what the Fathers perceived in the apostolic writings. The object of interpretation is Christ the Word, and all other words conform themselves to him, thereby serving to reveal him. When the Fathers encountered Paul, for example, they tended not to see a man concerned with demonstrating the \u201coriginal intent of the author\u201d or even the \u201cexact words\u201d of the author. Rather, they saw a man perfectly comfortable conforming biblical passages, and indeed entire biblical narratives, to the christological tradition he had received. It is Paul, after all, who teaches that the story of Sarah and Hagar is an \u201callegory\u201d (Gal 4:24) of the two covenants, corresponding to the \u201cchildren of slavery\u201d (Gal 4:25) and the \u201cchildren of promise\u201d (Gal 4:28). It is Paul who teaches that Christ is the true \u201cpaschal lamb\u201d (1 Cor 5:7) and that Adam was a \u201ctype of the one who was to come\u201d (Rom 5:14). It is Paul who states that the rock from which the Israelites drank in the desert \u201cwas Christ\u201d (1 Cor 10:4) and that the marital union of husband and wife \u201crefers to Christ and the church\u201d (Eph 5:32).
\nThe examples go on, and though the degree to which one finds such figural readings varies from book to book (and though modern interpreters would no doubt explain each of those passages differently), the Fathers saw in these passages an underlying interpretive principle that is no less present in the Gospels and the other apostolic writings. In John 5:39, Jesus is recorded as saying, \u201cYou search the Scriptures because you think that in them you have eternal life; and it is they that bear witness about me\u201d (ESV). Then, in Luke\u2019s account, he appears to apply this principle personally with two of his disciples on the road to Emmaus, where, \u201cBeginning with Moses and all the Prophets, he interpreted to them in all the Scriptures the things concerning himself\u201d (Lk 24:27 ESV).
\nIf the apostles interacted with the biblical texts in such a creative manner, their successors saw no barrier to prevent them from doing the same. If, as Paul states, \u201cThe letter kills, but the Spirit gives life\u201d (2 Cor 3:6), then the \u201cletter\u201d always kills, and the \u201cspirit\u201d always gives life. For the fathers, this notion manifested itself practically in a multitude of ways, including the very nature and shape of their scriptural citations. In short, there is a very close parallel between the manner in which the New Testament writers interacted with the Old Testament and the manner in which the church fathers interacted with the New Testament. Quite often, those interactions do not conform to modern standards of quotation. Could we, then, reconstruct the entire New Testament using only the quotations of the earliest church fathers? We ought perhaps to respond in the form of another question: Could we reconstruct even a significant portion of the Old Testament text using only the quotations of the New Testament? The two questions are deeply connected, and taken in conjunction, they provide a helpful framework for what follows.
\nComposite citations. To show that the Scriptures theologically conform to Christ and his teachings, Christian writers occasionally ensured that they were physically conformed to him as well. That is, at times, these authors carefully selected, shifted, and paired together the phrases and passages of Scripture in order to make more explicit the manner in which they reveal Christ or a central element of Christian belief. We find, for instance, numerous examples of composite citations throughout the pages of the New Testament, wherein two or more Old Testament passages are fused together and conveyed as though they are one. In Romans 3:9\u201318, Paul unites a number of different passages, drawn from the Psalms, Proverbs, and Isaiah in order to underscore the point that all people are \u201cunder sin\u201d:<\/p>\n
What then? Are we Jews any better off? No, not at all; for I have already charged that all men, both Jews and Greeks, are under the power of sin, as it is written:<\/p>\n
\u201cNone is righteous, no, not one;
\nno one understands, no one seeks for God.
\nAll have turned aside, together they have gone wrong;
\nno one does good, not even one [Ps 14:1\u20133; 53:1\u20133].\u201d
\n\u201cTheir throat is an open grave,
\nthey use their tongues to deceive\u201d [Ps 5:9].
\n\u201cThe venom of asps is under their lips\u201d [Ps 140:3].
\nTheir mouth is full of curses and bitterness [Ps 10:7].
\nTheir feet are swift to shed blood,
\nin their paths are ruin and misery [Prov 1:16],
\nand the way of peace they do not know [Is 59:8].
\nThere is no fear of God before their eyes [Ps 36:1]. (RSV)<\/p>\n
Nowhere does this text, in this form, appear in the Old Testament, and yet Paul seems perfectly comfortable declaring, \u201cIt is written.\u201d Of course, these words were \u201cwritten\u201d in the straightforward sense of the word, but the real point is that we would find ourselves sorely disappointed if we attempted to use Paul\u2019s \u201cquotation\u201d to reconstruct an original Old Testament text. Such a text does not exist. In continuity with the apostles, then, some early Christian writers applied the same methods in their interpretation of the apostles. One of the more prominent examples is Justin Martyr (ca. 100\u2013165), a second-century philosopher and apologist who spent the majority of his career teaching in Rome. In one of his best-known works, the First Apology, Justin consistently uses composite citations in order to deliver his message concisely and effectively. In a section detailing Christ\u2019s central moral teachings, for example, Justin writes,<\/p>\n
[Jesus] said, \u201cGive to everyone who asks and turn not away from him who wishes to borrow [Lk 6:30]. For if you lend to those from whom you hope to receive, what new thing do you do? Even the publicans do this [Lk 6:34]. Lay not up for yourselves treasure upon earth, where moth and rust corrupt and thieves break in; but lay up for yourselves treasure in heaven, where neither moth nor rust corrupts [Mt 6:19]. For what will it profit a man, if he should gain the whole world, but lose his own soul? Or what will he give in exchange for it? [Mt 16:26]. Lay up treasure therefore in heaven, where neither moth nor rust corrupts [Mt 6:20].\u201d (First Apology 15)<\/p>\n
Again, nowhere in the Gospels does Jesus say these exact words in this exact order, and using them for the reconstruction of a New Testament passage would prove highly problematic without a solid point of reference. To the modern critic, such methods may seem reckless. Indeed, ancient Christian writers understood the potential dangers involved, warning against the ways in which such a principle might be abused. Irenaeus of Lyons (second century), for example, lambasts the Valentinian \u201cgnostics\u201d for disregarding the \u201corder and connection\u201d of the Scriptures by \u201ctransferring passages, and dressing them up anew, and making one thing out of another.\u201d He compares the oracles of Scripture to a collection of precious jewels, which a skilled artist has carefully constructed into the mosaic of a king. The Valentinians, he complains, \u201crearrange the gems, and so fit them together as to make them into the form of a dog or of a fox\u201d (Haer 1.8.1). They then go about tricking the ignorant into believing that the image of the dog or the fox was the image originally crafted by the artist.
\nDespite his complaint, it cannot be that Irenaeus intrinsically objects to the rearranging of passages, for aside from Paul and the other New Testament writers, he himself does so in the context of this very same work. Indeed, the assumption is that the jewels (or \u201coracles\u201d) within the mosaic do actually conform to some image, and that this image is not one that can always be discovered simply by reading the text in a straightforward, linear manner. As he goes on, he reveals that the fundamental problem is not the rearranging of passages itself but that the Valentinians \u201cadapt the oracles of God to their baseless fictions\u201d (Haer. 1.8.1). That is, their starting point is faulty and erroneous. The \u201chypothesis\u201d with which they begin is without warrant, and as a result they only compound the problem and deceive others by forcing the scriptural evidence to fit their defective assumptions.
\nFor Irenaeus, the proper starting point is what some refer to as the \u201crule of faith,\u201d or the \u201ccanon of truth.\u201d This \u201ccanon,\u201d remarkably similar in order and content to the later creeds, is nothing less than the apostolic \u201chypothesis\u201d required to see rightly the image of the king. It is a verbal expression of Christ himself, crucified, risen, and preached \u201caccording to the Scriptures\u201d (meaning the Old Testament). It is the canon (rule or standard) that precedes, underlies, and gives shape to what would become the more concrete canon (list) of New Testament books. Provided, then, that the words of Scripture were arranged in such a way that they served to demonstrate and reveal this hypothesis, rearranging them was not inherently problematic. Indeed, it was often beneficial for the purposes of communicating one\u2019s exegetical, theological, or apologetic point.
\nUltimately, this could affect even the way in which a writer approached textual variation. Occasionally, provided a textual variant reading did not explicitly run contrary to the apostolic deposit of faith, patristic writers were happy to allow multiple readings to stand in their interpretations. In his Commentary on Romans, for example, Origen notes that there are two possible readings of Romans 3:5. It is either \u201cIs God unjust who inflicts wrath against men?\u201d or \u201cIs God unjust who inflicts wrath? I am speaking according to man.\u201d Yet he appears to make no effort to determine which was the \u201coriginal.\u201d Rather, he simply goes on to accept and explain the meaning of both textual variants in a manner consistent with his theology (Commentary on Romans 3.1).
\nAugustine of Hippo (ca. 354\u2013430), when faced with a similar problem in the various Latin translations, remarks, \u201cWhich of these two followed the original words one cannot tell, unless one reads copies of the original language. But all the same, for those who are shrewd readers, something important is being suggested by each version\u201d (Doctr. chr. 2.12.17). Of course, both Origen and Augustine were known for urging their students to familiarize themselves with the original texts (where they could be known), and yet they did not always perceive variants as inherently problematic. Provided that the variant words speak truly about the invariant Word, they felt free to accept multiple readings.
\n\u201cIn other words.\u201d As we have already begun to see, many early Christian writers believed that while one finds meaning through the words of Scripture, meaning is not necessarily equivalent to the words of Scripture. Granted, no one today would suggest that the mere possession of the words guarantees correct interpretation, but many would suggest that if one understands the basic historical contexts and definitions of the words, a correct interpretation will be well within reach. In other words, the literal sense dominates. This was not necessarily so in the ancient world. No early figure better personifies the difference than Origen, arguably Christianity\u2019s first great scriptural commentator and by far its most prolific. In his most famous and controversial work, On First Principles, Origen teaches the following: \u201cThe weakness of our understanding cannot discover the deep and hidden thoughts in every sentence; for the treasure of divine wisdom is concealed in vessels of poor and humble words, as the apostle points out when he says: \u2018We have this treasure in earthen vessels, that the greatness of the divine power may shine forth the more\u2019 (2 Cor 4:7)\u201d (On First Principles 4.1.7).
\nFor Origen, the primary error of Jewish interpretation is its insistence on taking the text at face value, or too literally, such that the Messiah ought \u201cactually and visibly to have \u2018proclaimed release to captives.\u2019 \u201d The same holds true for the \u201cheretics,\u201d such as Marcion, who took the notion of God\u2019s \u201canger\u201d or \u201cjealousy\u201d in the plain sense, and ended up rejecting the \u201cCreator God\u201d in favor of a \u201cmore perfect God\u201d revealed in Jesus Christ (On First Principles 4.1.8). By contrast, Origen teaches that every passage of Scripture must be interpreted in a pneumatic, or spiritual sense. Even with regard to the Gospels, he asks, \u201cIs there not also hidden in them an inner meaning which is the Lord\u2019s meaning, and which is only revealed through the grace that was given to him who said, \u2018We have the mind of Christ, that we may know the things that were freely given to us by God\u2019?\u201d (4.1.10). For Origen, Scripture is a historical, literal text only in the most superficial sense. It is important to understand the history, but it is far more important to understand the meaning (and unlike certain streams of modern thought, the two are not equivalent).
\nThe natural outcome of such a mindset, in contrast to certain modern norms, is a greater propensity for putting things into \u201cother words.\u201d This includes a greater propensity for adaptations and allusions that skillfully communicate Christian truth. In this, too, the Fathers were not without biblical warrant. One of the more prominent examples of a New Testament author intentionally adapting an Old Testament text to his christological convictions is found in Hebrews 10:5\u20137, which draws on Psalm 40:6\u20138. The text begins, \u201cConsequently, when Christ came into the world, he said, \u2018Sacrifices and offerings you have not desired, but a body have you prepared for me\u2019 \u201d (Heb 10:5 ESV). Remarkably, the original psalm does not include any form of the phrase \u201ca body you have prepared for me\u201d but rather reads \u201cears you have dug for me.\u201d Many seeming \u201cmisquotations\u201d in the New Testament are merely the result of the text having been translated into Greek, but in this case, the author appears to have intentionally put the text into other words to better communicate one of his overarching points (the importance of the sacrifice of \u201cthe body of Jesus Christ\u201d in Heb 10:10). Furthermore, the author attributes the words of the psalmist directly to Jesus, adding an additional layer of intrigue. Do these changes dramatically affect the meaning or intention of the psalmist? Perhaps not, but this in fact illustrates the overarching point. Even in the ESV Study Bible, for example, one finds the following commentary: \u201cNT quotations of OT texts are not always precise; NT authors often reword them or adapt them to suit their own purposes, yet always in a way that is compatible with their original meaning.\u201d Unlike certain patristic writers, however, we would be highly uncomfortable replicating something like this today.
\nFor the Fathers, this overall mindset also resulted in a greater openness to the use of translations that effectively communicate Christian theology in their choice of language. Translations were commonplace and were often considered spiritually authoritative in their own right. Even among the authors of the New Testament, the Septuagint was the most common source of Old Testament quotations (yet another reason why it would be difficult to reconstruct the Old Testament from the New). This alone suggests that for some, lacking the \u201coriginal words\u201d of the text was not an inherent obstacle to right interpretation. With few exceptions, this conviction grew stronger in the earliest centuries of Christianity, when the Septuagint remained dominant and few Christians possessed the capacity to read Hebrew.
\nAs Christianity began to spread throughout the ancient world, the Scriptures (including what would become the New Testament) were gradually translated into languages such as Latin, Syriac, and Coptic, resulting in a multitude of differences between manuscripts and versions. This made it increasingly more difficult to ascertain what the \u201coriginal\u201d text said. In response, Augustine of Hippo (among others) urged translators to familiarize themselves with Greek and Hebrew, being troubled by the \u201cinfinite variety of Latin translations.\u201d In his classic work On Christian Doctrine, Augustine states that the \u201cproper meaning of a passage, which several translators attempt to express \u2026 can only be definitely ascertained from an examination of it in the language they are translating from\u201d (2.13.19). Clearly, Augustine understood the importance of the original languages, yet, despite his strict instructions, he retained the belief that the Septuagint version of the Old Testament has \u201cgreater authority\u201d than the Hebrew. Why? Because the translation was inspired by the Holy Spirit: \u201cAnd for correcting any Latin versions at all, Greek ones should be employed, among which, as regards the Old Testament, that of the Seventy Translators has the greatest authority. These are said, throughout all the more learned Churches, to have been so directed by the Holy Spirit in their translations, that while being so many they had but a single mouth\u201d (2.15.22).
\nTranslations are nothing less than extended interpretations, or controlled articulations of meaning. They are not equivalent to saying \u201cin other words,\u201d but they are quite literally in other words. Augustine, like many early Christians, believed that the Septuagint was \u201cinspired\u201d or \u201cdirected\u201d by the Holy Spirit, not because the original Hebrew wording was somehow inadequate or \u201cless inspired\u201d but because inspiration must be located also in the mind of the interpreter(s). Origen states this point succinctly, writing, \u201cOn this point the entire Church is unanimous, that while the whole law is spiritual, the inspired meaning is not recognized by all, but only by those who are gifted with the grace of the Holy Spirit in the word of wisdom and knowledge\u201d (On First Principles, preface 8). In short, uninspired interpretations of inspired texts will not get you very far, even if you possess the original words.
\nFor many early Christian writers, then, translations and adaptations\/allusions (as well as adapted translations) were not always perceived as being inadequate or even limiting. They were often considered an inspired means of communicating inspired meaning. The point here is not that the Fathers all played fast and loose with the Scriptures (though some might be worthy of the accusation). Many were in fact quite precise with their quotations. The point, rather, is that one\u2019s very theology of Scripture and scriptural interpretation affects the way in which one quotes and uses it.
\nWe therefore ask yet again, Could we reconstruct virtually the entire New Testament using only the quotations of the earliest church fathers? In response, we also ask yet again, Could we reconstruct even a significant portion of the Old Testament text using only the quotations of the New Testament? Between all the adaptations, allusions, composite citations, and Greek translations common among the New Testament authors (partially resulting from their very theology of Scripture), the answer must be a definitive no. That is, we could not do so without the Old Testament manuscript tradition as a point of reference. Without that manuscript tradition, one would simply be left guessing. Because the Fathers tended to imitate the New Testament authors in their interpretation and style of quotation, the same generally holds true for their writings.
\nDoes all of this suggest that the original wording did not matter to early Christian writers? Certainly not. Many (perhaps most famously Jerome) insisted on the primacy of the Hebrew with regard to the Old Testament, and most Christians understood the importance of reading the apostolic writings in Greek. Even Origen, who explicitly taught that one must move beyond the bare words (or plain meaning) of Scripture, was a forerunner to modern textual criticism. He tended to be more precise in his quotations and is famous for having gathered together not only the Hebrew text(s) of the Old Testament but also the recognized Septuagint of his day along with a variety of other Greek translations such as those of Aquila, Symmachus, and Theodotion. Origen placed each of these versions, along with a Greek transliteration of the Hebrew, into side-by-side parallel columns. This monumental work came to be known as the Hexapla (\u201csix-columned thing\u201d) and was preserved in the library of Caesarea for centuries before its likely destruction. This is hardly the work of a man indifferent to the original wording of Scripture. But again, it is crucial to understand that his concern to possess the original wording does not, for him, suggest that attaining the original wording was his ultimate goal. Rather, the original wording best enabled him to discover the meaning hidden within the wording (the treasure hidden in \u201cearthen vessels\u201d).<\/p>\n
THE SILVER LINING<\/p>\n
It seems that we can finally put this myth to rest. But despite everything we have seen in this chapter, it is important to recognize that the quotations of the church fathers can and do play an important role in New Testament textual criticism. Specifically, they can be helpful in determining when and where a certain textual variant became prevalent or was being read.
\nTake, for example, the letters of Cyprian (ca. 210\u2013258), bishop of Carthage in the mid-third century. The scriptural quotations in those letters appear almost always to agree with the text preserved in the Old Latin manuscript known as VL 1 (Codex Bobiensis; k). The manuscript itself does not date until the fourth or fifth century, at least a century after Cyprian\u2019s time. As Cyprian could not possibly have consulted a copy produced after his own death, scholars have theorized that VL 1 must have descended from an earlier copy located somewhere in the proximity of North Africa in the mid-third century. In cases such as this, patristic quotations are enormously helpful.
\nAnother example relates to the establishment of the oldest form of a text, as in the case of Luke 3:22. The majority of witnesses, including P4 (late second\/early third century), record that at the baptism of Jesus, the voice from heaven declared, \u201cYou are my beloved Son, in you I am well pleased,\u201d while other witnesses, beginning with Codex Bezae (fifth century) have the voice proclaiming, \u201cYou are my Son, today I have begotten you.\u201d Judging by the Greek manuscript evidence alone, one might conclude that the latter reading was a later innovation, or at least a minority reading. However, the patristic evidence suggests otherwise. Justin Martyr, Clement of Alexandria, Origen, Methodius, the Gospel according to the Hebrews, the Gospel according to the Ebionites, and the Didascalia all make use of the latter reading, serving to flip the conclusion on its head. As Metzger and Ehrman put it, \u201cThis means that, with the exception of P4, all of the surviving witnesses of the second and third centuries appear to have known this alternative form of the text.\u201d Again, without access to these patristic witnesses, our knowledge of the transmission history of this text would be greatly diminished.
\nFinally, there are a multitude of instances in which a church father explicitly comments on a textual variant in the context of one of his commentaries, homilies, letters, or other theological or apologetic works. These have been thoroughly documented in recent years, and so one brief example should suffice. In Irenaeus\u2019s great work Adversus Haereses (Against Heresies), while discussing the antichrist and the great apostasy to come, he offers up an explanation of the number of the beast, which is famously 666 in Revelation 13:18. However, he notes that he is aware of a variant reading of this passage that instead has the number 616. His rejection of this variant is absolute, and his reasoning is as follows: First and foremost, he states that 666 is found in \u201call the most approved and ancient copies.\u201d Second, he notes that this reading has been upheld by \u201cthose men who saw John face to face.\u201d Finally, he comments on the internal meaning and logic of 666. Six, the number of apostasy, is repeated three times, specifically indicating an apostasy \u201cwhich occurred at the beginning, during the intermediate periods, and which shall take place at the end.\u201d He appears to attribute the variant to the error of a copyist but explicitly warns against intentionally perpetuating it, as it might lead the faithful into error regarding the identity of the beast (Haer. 5.30.1). Though Irenaeus is not the sole witness to this variant (the third-century P115 is the earliest manuscript witness), his commentary provides additional insight into its date, location, and prevalence. Furthermore, examples such as this one offer a special glimpse into the way early Christians thought about textual variation, including why some variants were accepted while others were rejected, or why certain manuscripts were considered more reliable than others.
\nWhile we may not, therefore, be able to reconstruct the entire New Testament text using only the quotations of the church fathers, textual critics can certainly do a great deal of important work with them, given the right set of circumstances and resources.<\/p>\n
CONCLUSION<\/p>\n
The myth at the heart of this chapter has proven to be exactly that, but there is a sense in which the Fathers themselves would have affirmed the underlying sentiment it represents. What would we do if every single copy of the New Testament were suddenly destroyed? To what (or whom) would we turn? We would no doubt turn to the Fathers of the church. Not because they possess the original wording of the New Testament text within their writings, but because they possess Truth. At least, that is how Irenaeus would have answered the question. Indeed, he poses a variation of that same question in Adversus haereses, asking, \u201cWhat if the apostles had not left us the Scriptures?\u201d That is, what if there were no writings available to us at all? (Note how the question itself presupposes the supreme importance of Scripture.) Irenaeus replies, \u201cOught we not, then, to follow the disposition of tradition, which they handed down to those to whom they entrusted the churches?\u201d (Haer. 3.4.1). The churches, and those to whom they have been entrusted, carry within them that which (or he whom) the Scriptures were always meant to articulate. At the heart of our myth, then, is the desire to know that Christianity would survive, in possession of the truth, even if it turned out that all the New Testament manuscripts were inaccurate. This hypothetical situation naturally stretches and pushes us, and Christians of different theological stripes may respond to it in different ways (with some taking offense at the very proposition). Yet thanks to the many great scribes and text critics who have worked tirelessly over the centuries, such a hypothetical has never become reality. We do not need to reconstruct the New Testament text without manuscripts. In fact, thanks to them, the myth deconstructed in this chapter was never really necessary in the first place.<\/p>\n
Key Takeaways
\n\u25ba The argument that we can reconstruct all but eleven verses of the New Testament from 36,289 quotations by the church fathers is not only false but it is a conflation of two different arguments that are both riddled with problems. It should not be used.
\n\u25ba Even if true, the argument fails because it is viciously circular: we have to know what the text of the New Testament is before we can identify patristic quotations of it.
\n\u25ba In general, patristic theology of Scripture affected the value of patristic quotations of Scripture. For instance, when a variant reading did not run contrary to the rule of faith, patristic writers were sometimes happy to allow multiple readings to stand in their interpretations.
\n\u25ba Many early Christian writers believed that while one finds meaning through the words of Scripture, meaning is not necessarily equivalent to the words of Scripture. Consequently, many patristic writers felt free to be more fluid in their wording when they \u201cquoted\u201d Scripture than many Christians today would be.<\/p>\n
CHAPTER THIRTEEN<\/p>\n
MYTHS ABOUT CANON<\/p>\n
WHAT THE CODEX CAN AND CAN\u2019T TELL US<\/p>\n
John D. Meade<\/p>\n
The ancient codex (the precursor to the modern book form) was invented around the turn of the era and perhaps was in use by the middle of the first century. Before the codex, the primary writing medium was the roll or scroll. Why did early Christians adopt the codex form? Some scholars believe that the development from scroll to codex caused the concept of canon to become more concrete, for with this new book technology scribes could now organize many works in order, and for the first time choices over what books to include and exclude were made. The idea is that codex implies canon, so that the multibook codex creates or at least corresponds with the development of the canon. In considering just what the codex can and cannot tell us about the New Testament canon, we must (1) consider how early Christians described their religious literature and (2) summarize the contents of our codices, paying close attention to their contents. The following lines of evidence show that the canon lists reveal the New Testament canon, and that while the codices sometimes correspond to the New Testament canon, more often they simply reflect what books Christians were reading. Thus, if one wants to know the contours of the early canon, one should consult the lists. If one wants to become acquainted with the early Christian\u2019s repository or library, one should consult the contents of manuscripts.<\/p>\n
CANON LISTS AND THEIR IMPORTANCE<\/p>\n
Before turning to the codices, we need to supply the early context for canon, the canon lists. These lists are exclusive lists of authoritative books, and their importance lies in that they provide rather specific information on the contents and boundaries of the canon of Scripture. The lists can be partial (e.g., only Gospels, as in Irenaeus\u2019s Haer. 3.1.1) or complete (e.g., the entire New Testament, Old Testament, or both). Furthermore, the church fathers\u2019 or scribe\u2019s commentary in and around the list of books often provides crucial information that a crude reproduction of the list cannot convey (see the tables at the end of this chapter).
\nCanon lists represent the culmination of a process\u2014not the beginning of one. Therefore, the presence of lists does not establish a canon but rather evinces a time when the complete canon crystallized. For example, even though Athanasius in 367 was the first to include all twenty-seven New Testament books into one canon list (though see Origen below), he did not invent this list. He compiled a complete list of books that were already recognized as canonical (four Gospels, Acts, fourteen epistles of Paul, seven Catholic Epistles) and included Revelation\u2014a book at the edges, that is, a book that most but not all churches would have recognized as canonical.
\nThe number of lists and what they are. From the early period (up to ca. AD 400), early Christians drafted eighteen canon lists and presented them as complete and exclusive lists of New Testament books\u2014nine in Greek, eight in Latin, and one in Syriac. The lists and their dates are as follows:<\/p>\n
\u2022 Origen (ca. 250; Hom. Jos. 7.1)
\n\u2022 Eusebius (ca. 325; Hist. eccl. 3.25)
\n\u2022 Cyril of Jerusalem (ca. 350; Catechetical Lectures 4.36)
\n\u2022 Athanasius (367; Festal Letters 39.18)
\n\u2022 Synod of Laodicea 59 (343\u2013380)
\n\u2022 Apostolic Canons 85 (375\u2013380)
\n\u2022 Gregory of Nazianzus (381\u2013390; Carmen 12)
\n\u2022 Amphilochius (ca. 380; Epistula Iambica ad Seleucum lines 280\u2013319)
\n\u2022 Epiphanius (ca. 376; Panarion 76.22.5)
\n\u2022 Muratorian Fragment (second, third, or fourth century)
\n\u2022 Codex Claromontanus (ca. 300\u2013350)
\n\u2022 Mommsen Catalogue (before 365)
\n\u2022 Jerome (395; Ep. 53)
\n\u2022 Rufinus (ca. 404; Commentarius in symbolum apostolorum 35)
\n\u2022 Breviarium Hipponense (ca. 393)
\n\u2022 Augustine (397; Doctr. chr. 2.8.13.29)
\n\u2022 Pope Innocent I (405; Ep. 6.7)
\n\u2022 Syriac List of St. Catherine\u2019s Monastery (350\u2013400)<\/p>\n
Canonical books. These lists agree on the four Gospels, on Acts, and mostly on the fourteen epistles of Paul (Hebrews is omitted in the Muratorian Fragment, Mommsen, and perhaps Claromontanus, although textual corruption may be the cause of the omission of some books in this list). This core canon was anticipated already in the second century. If the Muratorian Fragment belongs to the second century, then the Gospels, Acts, and thirteen epistles of Paul are clearly listed, though our evidence of this core canon does not depend on the early dating of the Fragment. Around 180, Irenaeus speaks of the established four Gospels by his time (Haer. 3.1.1; 3.11.8). From the late second century, Christians such as Irenaeus (Haer. 3.13.3), Tertullian (Against Marcion 5.1\u20134), and Clement of Alexandria (Stromata 5.12.82) recognized Acts as part of their Scriptures. The church had Paul\u2019s fourteen letters by 200, since Tertullian describes Paul\u2019s thirteen letters (Hebrews omitted) and treats them as a collection of his epistles (Against Marcion 5), and Clement of Alexandria (d. ca. 215) attributes them to Paul and cites all fourteen letters except Philemon (perhaps due to its brevity).
\nOn the other hand, our canon lists reveal dispute over the Catholic Epistles. In Greek and Latin traditions, twelve lists contain the traditional seven Catholic Epistles, which shows most of the churches recognized their canonicity, probably from an earlier time than the fourth century; only Eusebius, Amphilochius, Apostolic Canons, the Muratorian Fragment, and Mommsen deviate from this pattern. Eusebius says only 1 Peter and 1 John are recognized and places the other five among the disputed books. Amphilochius accepts James, 1 Peter, and 1 John but notes disputes over 2 Peter, 2-3 John, and Jude. The Muratorian Fragment lists Jude and the letters of John (1-2 John or 1-3 John?) and lacks the others. Mommsen lacks James and Jude. Although the tendency was to reduce the number of these letters, the Apostolic Canons adds 1-2 Clement among them. The Syriac list of St. Catherine\u2019s Monastery does not include the Catholic Epistles, showing the early disputes over these books within the Syriac church. Cyril of Jerusalem (ca. 350) is the first to list all seven Catholic Epistles without dispute.
\nOf the eighteen canon lists, Revelation is omitted or disputed in Origen, Eusebius, Cyril of Jerusalem, Synod of Laodicea, Apostolic Canons, Gregory of Nazianzus, Amphilochius, and the Syriac list\u2014that is, most of the Eastern lists. However, even among these lists, the situation is complex. Gregory may include an allusion to the book of Revelation in his list, while Amphilochius lists it but notes, \u201cThe majority say it is spurious.\u201d The textual transmission of Origen\u2019s Homiliae on Josuam has some textual witnesses that include Revelation, while others do not. Eusebius lists it twice, once among his \u201crecognized (= canonical) books\u201d and again among his \u201cdisputed-spurious books,\u201d while Cyril, the Synod of Laodicea, and Apostolic Canons omit it (though the Ethiopic version of Apostolic Canons includes it). The complex reception of Revelation has been documented elsewhere.
\nBooks at the edges of the canon. The Greek lists of Eusebius and Athanasius place other books in secondary lists that were not finally included in the twenty-seven-book New Testament. The Apostolic Canons includes extra books within its primary list. Eusebius includes in his list the following books as \u201cdisputed-spurious\u201d: the Acts of Paul, the Shepherd of Hermas, the Apocalypse of Peter, the Epistle of Barnabas, the Didache, and a Gospel According to the Hebrews. In his secondary list, Athanasius lists the Shepherd and the Didache as part of his \u201cbooks to be read to catechumens.\u201d The Apostolic Canons includes 1-2 Clement among the canonical books.
\nThe Latin lists of the Muratorian Fragment, Claromontanus, and Rufinus also contain more books than those associated with the traditional New Testament. The Muratorian Fragment includes the Epistle to the Laodiceans (\u201cforged\u201d), the Epistle to the Alexandrians (\u201cforged\u201d), the Apocalypse of Peter (\u201cprivate reading according to some\u201d), and the Shepherd of Hermas (\u201cprivate reading only\u201d). Claromontanus includes Barnabas, the Shepherd of Hermas, the Acts of Paul, and the Apocalypse of Peter with an obelus or lance (\u2192), usually used in ancient manuscripts to indicate misplaced text. In his list of ecclesiastical books, Rufinus includes the Shepherd, the Two Ways (the Didache?), and the Judgment of Peter.
\nEarly Christian description of religious literature. The preceding section raises the question about how early Christians described their religious literature. Generally, they did not describe books in a binary way: canonical and noncanonical. Rather, they employed a tri-fold description of religious literature: canonical scripture, useful\/intermediate scripture, and apocryphal. In Festal Letters 39.20\u201321, for example, Athanasius describes this literature as follows:<\/p>\n
But for the sake of greater accuracy, I add this, writing from necessity. There are other books, outside of the preceding, which have not been canonized (\u03ba\u03b1\u03bd\u03bf\u03bd\u03b9\u03b6\u03cc\u03bc\u03b5\u03bd\u03b1, kanonizomena), but have been prescribed by the ancestors to be read (\u1f00\u03bd\u03b1\u03b3\u03b9\u03bd\u03ce\u03c3\u03ba\u03b5\u03c3\u03b8\u03b1\u03b9, anagin\u014dskesthai) to catechumens wanting to be instructed in the word of piety.\u2026 Nevertheless, beloved, the former books are canonized (\u03ba\u03b1\u03bd\u03bf\u03bd\u03b9\u03b6\u03bf\u03bc\u03ad\u03bd\u03c9\u03bd, kanonizomen\u014dn); the latter are (only) read (\u1f00\u03bd\u03b1\u03b3\u03b9\u03bd\u03c9\u03c3\u03ba\u03bf\u03bc\u03ad\u03bd\u03c9\u03bd, anagin\u014dskomen\u014dn); and there is no mention of the apocryphal books (\u1f00\u03c0\u03bf\u03ba\u03c1\u03cd\u03d5\u03c9\u03bd, apokryph\u014dn).<\/p>\n
In accordance with the ancestors, Athanasius lists the canonical books of Old and New Testament, several books that are to be read (e.g., Tobit, the Shepherd, the Didache), and the category of apocryphal books but does not list them. Other early Christian fathers who presented a similar tri-fold description include Amphilochius (Epistula Iambica ad Seleucum 251\u2013260), Epiphanius (De mensuris 4), Jerome (Preface to the Books of Solomon), and Rufinus (Commentarius in symbolum apostolorum 36). For example, in the context of a canon list, Jerome says the Shepherd is \u201cnot in the canon\u201d (Prologus Galeatus), but in another context, he calls it \u201cuseful\u201d and says many ancient writers quoted it authoritatively (Vir. ill. 10). Therefore, early Christians had an intermediate category of books\u2014books considered somewhere between canonical and apocryphal; specifically, these books were considered useful for instruction in piety but not for establishing ecclesiastical doctrine.
\nWould early Christians refer to these books that were useful but not canonical as \u201cScripture\u201d or as \u201cinspired\u201d? Would they cite them as Scripture alongside canonical texts? They did. For example, Irenaeus calls the Shepherd \u1f21 \u03b3\u03c1\u03ac\u03c6\u03b7 (h\u0113 graph\u0113), usually translated \u201cScripture\u201d (Haer. 4.20.2). Clement of Alexandria believed the Shepherd to be a genuine revelation, thus making Hermas \u201ctechnically a prophet.\u201d Origen refers to the Shepherd as very useful and divinely inspired (Commentarii in Romanos 10.31) but does not include it in his canon list (Hom. Jos. 7.1). Athanasius cites the Shepherd 26.1 alongside Genesis 1:1 (De incarnatione 3.1), indicating to some scholars that he considered it canonical, at least earlier in his career. However, in De decretis 18.3.2 and his canon list he says clearly the book is not in the canon, showing that he referred to the book earlier as \u201cuseful\u201d but not as \u201ccanonical.\u201d
\nTherefore, patristic biblical theory included a tri-fold description of religious literature. An ancient author\u2019s canon list did not include all the books that he considered to be Scripture, and the scope of scriptural books was often wider than his canon list. Although patristic biblical theory included distinctions between canonical and useful scriptures, a church father\u2019s exegetical practice sometimes appears to blur his theoretical distinctions, as in the case of Origen\u2019s use of the Shepherd (though his early third-century context must be remembered). Athanasius also says that the Shepherd is useful and quotes it authoritatively. Yet neither Father included the book in his canon list.
\nConclusions. From these data, we see that in the second century there was already a core New Testament canon consisting of the four Gospels, Acts, and thirteen or fourteen epistles of Paul. By the mid-fourth century (though see Origen\u2019s canon list from the mid-third century), probably some churches recognized seven Catholic Epistles, with others recognizing only two or three, and others, as seen in Apostolic Canons, recognizing more. The Revelation of John was recognized early, but later doubts about its authorship caused it to be omitted from many Eastern lists in the fourth century. By the end of the fourth century, these disputes appear to have subsided as the twenty-seven-book New Testament canon crystallized. This major formative stage did not cause the secondary lists of other important books to disappear, since they continued to be listed well after the fourth century.
\nThese other important books circulated from the late first and second century AD. Church fathers referred to them and cited them in ways similar to how they cited canonical books. But their biblical theory accounts for this phenomenon. In the third- and fourth-century fathers, canon lists did not exhaust all scriptural works, and other scriptural works spilled over the canonical boundaries, as evidenced by the secondary lists of scriptural books. The debate over whether a tiered view of New Testament Scripture existed in the second century continues, but recent scholarship shows its plausibility.
\nAs we turn to the material evidence of our manuscripts, we need to remember that the New Testament canon is clearly observed from the early Christian lists. We also need to heed the early Christian statements on specific books and their overall biblical theory to avoid committing historical anachronism by assigning canonical significance to works found in codices that the ancient Christians would not have recognized as canonical.<\/p>\n
THE CONTENTS OF OUR NEW TESTAMENT MANUSCRIPTS<\/p>\n
Bearing in mind that the canon lists and other such explicit statements provide the evidence for the New Testament canon, in this section we look at the evidence of codices\u2014single and multiple books\u2014from AD 100 to 700 and include several relevant Latin and Coptic multibook codices (each is clearly labeled in the tables below). Most studies limit the evidence to Greek and manuscripts dated to the end of the third century, but this study expands the date range and the languages, attempting to show whether the codex format helped determine the canon or at least reflected it. Some scholars have argued that it did. For example, J. K. Elliott writes,<\/p>\n
Canon and codex go hand in hand in the sense that the adoption of a fixed canon could be more easily controlled and promulgated when the codex was the means of gathering together originally separate compositions.\u2026 We must assume that the authorities behind Codex Sinaiticus and Codex Alexandrinus considered these works canonical [the Shepherd, Barnabas, 1-2 Clement] and wished to promote them as such. Certainly, the user of these codices would have accepted all the texts in their Bible codex as having equal status.<\/p>\n
Elliott thus holds that what is in the codex is what was believed to be in the canon. In explaining why Christians preferred the codex, Michael Kruger agrees with Elliott that there is a relationship of some kind between the two, writing, \u201cThe most plausible suggestions [for why Christians preferred the codex over the roll] link the codex with the early development of the New Testament canon.\u201d Next, he tells us what two distinctive functions the codex performed: \u201c(1) Positively, it allowed certain books to be physically grouped together by placing them in the same volume; and (2) negatively, it provided a natural way to limit the number of books to those contained within the codex; that is, it functioned as a safeguard.\u201d
\nLarry Hurtado is somewhat broader but still thinks there is a connection between \u201cscripture\u201d and codex, writing, \u201cIn any case, it is clear that Christians favored the codex particularly for the writings that they treated as scripture.\u201d A little later, he offers this important caveat:<\/p>\n
We should, however, be wary of simplistic conclusions. For example, in light of the clear preference for the codex generally, it would be unsound to assume that if a text was copied in a codex this signals that the text was used as scripture. On the other hand, given this general Christian preference for the codex, particularly for scriptures, plus a noteworthy readiness to use the roll for a variety of other Christian texts, it is reasonable to judge that the use of a roll to copy a text signals that the copyist and\/or user for whom the copy was made did not regard that text (or at least that copy of that text) as having scriptural status.<\/p>\n
Thus, not every literary work copied in codex form implies its scriptural status. But in the case of multibook codices, Hurtado does suggest that the codex form reflects a connection between the works included. As he writes, \u201cI suggest specifically that the physical linkage of texts in one manuscript probably reflects a view of them as sharing some common or related subject matter or significance for readers.\u201d Therefore, on this view, the physical grouping of books helped to determine or develop scriptural collections or a canon in a binary manner: a book is either included between the covers or it is not. But around the same time that many of these codices were produced and read, Christians had already developed the tri-fold description of literature with its middle category of useful books that could be placed alongside those books they acknowledged as canonical. Therefore, it will be difficult to interpret the data of the codices presented below according to the statements of the early Christians given above, for early Christians maintained conceptual distinctions between their books (useful and canonical books) that physical linkages between them in codices could not maintain.
\nIn what follows, I argue that the codex did not serve to demarcate the Christian canon, nor did it help it develop in any notable way, even though at times the contents of our codices overlap with the canon lists. Thus, the codex does not furnish a rule that would only generate exceptions. The early New Testament canon can be observed from the early statements and lists, and furthermore, that evidence should help us interpret the material evidence of the codices.
\nSingle-book codices. Work has been undertaken on the interpretation of the fragmentary remains of codices. From one extant page, can we know whether that codex contained more than one book? An even further question is, What would the fragmentary evidence reveal about the putative early collections of books in one codex?
\nMichael Dormandy recently wrote an article in which he has placed the Greek codices from the second to the sixth century in six categories: (1) certainly collection-evident; (2) plausibly collection-evident; (3) certainly or plausibly one work; (4) plausibly multiwork, indeterminably collection-evident; (5) plausibly not collection-evident; and (6) definitely not collection-evident. His third category is relevant for our study. In this category Dormandy includes the fragments of codices deemed to contain one book, and it is significant that in each century (second to sixth) the majority of manuscripts were categorized as containing one work. Part of Dormandy\u2019s conclusion captures the difficulties involved in this interpretation: \u201cAlthough the number of single-work manuscripts identified by this study is high, this is partly due to my decision, discussed above, to assume that a manuscript is single-work, unless there is evidence to the contrary.\u201d But he finally concludes that single-book codices were \u201ccommon enough,\u201d and this datum calls into question that \u201cthe works of the NT very commonly circulated together.\u201d Indeed, Dormandy refers to the analysis of Francis Watson that, if the four-Gospel codex were the norm, we might expect an equal number of fragments of each Gospel, but fragments from Matthew and John far outnumber fragments from Mark and Luke.
\nAs we will see below, multibook codices did exist, but these were probably not the norm or very common in the patristic period. However, if the single-book codex was the norm, it would be a challenge to see the correspondence between canon and codex. But what do multibook codices tell us? We turn to that evidence next.
\nMultiple-book codices. There are some sixty New Testament multiple-book codices, that is, codices that have one book joined to another. Most of these types preserve only two books but some more than two. Once again, we will examine the evidence according to the collections established from the canon lists above, treating evidence of codices with two or more books. We will examine the pandects (complete New Testament) in their own section below.
\nGospels. Most of the evidence of multiple-book codices is from the Gospels. Matthew is joined with other Gospels thirteen times and once with Acts, as follows:<\/p>\n
LDAB
\nGA
\nDate
\nContents
\n2899
\n064 + 074 + 090
\n500\u2013599
\nMatthew + Mark
\n2901
\n043
\n500\u2013599
\nMatthew + Mark
\n2987
\n0104
\n600\u2013699
\nMatthew + Mark
\n2988
\n0107
\n650\u2013699
\nMatthew + Mark
\n2990
\n042
\n500\u2013599
\nMatthew + Mark
\n128512
\n067
\n400\u2013499
\nMatthew + Mark
\n2936
\nP4 + P64 + P67
\n150\u2013250
\nMatthew + Luke
\n2982
\n0171
\n175\u2013225
\nMatthew + Luke
\n2978
\n1276
\n500\u2013599
\nMatthew + John
\n2979
\nP44
\n500\u2013699
\nMatthew + John
\n2989
\n087 + 092b
\n500\u2013599
\nMatthew + Mark + John
\n62323
\n1601
\n400\u2013599
\nMatthew + Mark + John
\n2984
\n078
\n500\u2013599
\nMatthew + Luke + John
\n2981
\nP53
\n350\u2013399
\nMatthew + Acts<\/p>\n
Table 13.1. Manuscripts with Matthew and other Gospels or Acts<\/p>\n
Three times Matthew is grouped with Old Testament and other books:<\/p>\n
LDAB
\nGA
\nDate
\nContents
\n8942 (Latin)
\n425\u2013499
\nActs of Pilate + Matthew + Infancy Gospel of Jesus
\n2993 (Greek + Coptic)
\nP62
\n300\u2013399
\nMatthew + Daniel 3:51\u201353 (= Ode 8)
\n3315
\n035
\n500\u2013599
\nIsaiah + Matthew<\/p>\n
Table 13.2. Manuscripts with Matthew and books outside the New Testament<\/p>\n
Mark + John occurs once:<\/p>\n
LDAB
\nGA
\nDate
\nContents
\n2927
\nP84
\n550\u2013599
\nMark + John<\/p>\n
Table 13.3. Manuscript with Mark and John<\/p>\n
Luke appears with John and other material:<\/p>\n
LDAB
\nGA
\nDate
\nContents
\n2895
\nP75
\n275\u2013325
\nLuke + John
\n2896 (Coptic + Greek)
\nP2
\n500\u2013699
\nLuke + John
\n2897 (Coptic + Greek)
\n070 + 0110 + 0124 + 0178 + 0179 + 0180 + 0190 + 0191 + 0193 + 0194 + 0202
\n550\u2013650
\nLuke + John
\n2898 (Coptic + Greek)
\n029 + 0113 + 0125 + 0139
\n500\u2013599
\nLuke + John
\n2932
\n026
\n400\u2013499
\nLuke + John
\n107965 (Coptic)
\n300\u2013399
\nExodus + Luke + Shepherd of Hermas<\/p>\n
Table 13.4. Manuscripts with Luke and John or Luke and books outside the New Testament<\/p>\n
In addition to the above, John appears in multiple-book codices three times without any of the other Gospels:<\/p>\n
LDAB
\nGA
\nDate
\nContents
\n2806 (Greek + Coptic)
\nP6
\n400\u2013450
\n1 Clement (Coptic) + John (Greek-Coptic) + James (Coptic)
\n777
\n500\u2013699
\nJohn + Commentaries on Psalms
\n2763 (Coptic + Greek)
\n275\u2013350
\nJohn + mathematical school exercise<\/p>\n
Table 13.5. Manuscripts with John and books outside the New Testament<\/p>\n
The four Gospels appear together four times and with Acts two times (though not always in the same order):<\/p>\n
LDAB
\nGA
\nDate
\nContents
\n2904
\n024
\n500\u2013599
\nMatthew + Mark + Luke + John
\n2905
\n022
\n550\u2013599
\nMatthew + Mark + Luke + John
\n2986
\n1043
\n400\u2013499
\nMatthew + Mark + Luke + John
\n10657
\n500\u2013599
\nMatthew + Mark + Luke + John
\n2980
\nP45
\n200\u2013250
\nMatthew + Mark + Luke + John + Acts
\n2929
\n05
\n400\u2013450
\nMatthew + Mark + Luke + John + Acts + 2 John (Latin)<\/p>\n
Table 13.6. Manuscripts with the four Gospels (and Acts)<\/p>\n
Most of these multiple-book codices contain only two or three books. A relative few codices evince the four-Gospel collection or a Gospel collection with Acts. Matthew was grouped with the Infancy Gospel of Jesus and the Acts of Pilate. Luke was grouped with material including the Shepherd of Hermas. John was grouped with 1 Clement and some commentarial literature and school exercises. Thus, the evidence both comports and contrasts with the early evidence for the New Testament canon presented above. Christians had already acknowledged the four-Gospel collection by 180, as Irenaeus says. The canon lists indicate no disputes over this collection. Thus, the relative few codices that incorporate the four Gospels comport with the evidence above. But some codices did contain a Gospel with some unexpected material. In Greek, there was no evidence of the canonical Gospels being grouped with apocryphal Gospels such as the Gospel of Thomas. But in Latin, Matthew was joined to the Infancy Gospel of Jesus and the Acts of Pilate (!), and in Greek and Coptic, Luke and John were grouped with the Shepherd and 1 Clement, respectively.
\nPauline Letters. What do the multiple-book codices show us about the Pauline letters? I present the evidence from shortest to longest collections.
\nPaul\u2019s letters were probably formed into a collection by AD 200, according to the statements of early Christians. The material evidence coheres with the early testimonia since these codices appear almost exclusively to include traditional Pauline material, even though no one codex includes all of them except the pandects (see below).<\/p>\n
LDAB
\nGA
\nDate
\nContents
\n3060
\nP34
\n500\u2013699
\n1 Corinthians + 2 Corinthians
\n3008
\nP92
\n250\u2013350
\nEphesians + 2 Thessalonians
\n3002
\n0208
\n400\u2013499
\nColossians + 1 Thessalonians
\n3017
\nP30
\n175\u2013225
\n1 Thessalonians + 2 Thessalonians
\n3054 (Coptic + Greek)
\nP205
\n500\u2013599
\nTitus + Philemon
\n3016
\nP15 + P16
\n300\u2013399
\n1 Corinthians + Philippians
\n3001
\n088
\n400\u2013599
\n1 Corinthians + Titus
\n3007
\n0209
\n600\u2013699
\nRomans + 2 Corinthians + 2 Peter
\n3063
\nP61
\n600\u2013699
\nRomans + 1 Corinthians + Colossians + Philippians + 1 Thessalonians + Titus + Philemon
\n7312
\n0285 + 081
\n500\u2013599
\nRomans + 1-2 Corinthians + Ephesians +
\n1 Timothy + Hebrews + Philippians
\n3011
\nP46
\n200\u2013250
\nRomans + Hebrews + 1-2 Corinthians + Ephesians + Galatians + Philippians + Colossians + 1 Thessalonians
\n7152
\n015
\n500\u2013550
\n1-2 Corinthians + Galatians + Colossians + 1-2 Thessalonians + Hebrews + 1-2 Timothy + Titus
\n3044
\n016
\n500\u2013599
\n1-2 Corinthians + Galatians + Ephesians + Philippians + Colossians + 1-2 Thessalonians + Hebrews + 1-2 Timothy + Titus + Philemon<\/p>\n
Table 13.7. Manuscripts with the Pauline Epistles<\/p>\n
There were two other groupings:<\/p>\n
LDAB
\nGA
\nDate
\nContents
\n2844
\n0296
\n500\u2013599
\n1 John + 2 Corinthians
\n3030
\nP99
\n400\u2013499
\nGlossary on 2 Corinthians + Galatians + Ephesians; Greek verb conjugations + alphabet<\/p>\n
Table 13.8. Manuscripts with Pauline Epistles and non-Pauline books or material<\/p>\n
Acts. In addition to Acts joined to the Gospels above, it is also joined with the following books:<\/p>\n
LDAB
\nGA
\nDate
\nContents
\n2893
\n0166
\n450\u2013550
\nActs + James
\n2894
\nP74
\n500\u2013699
\nActs + James + 1-2 Peter + 1-3 John + Jude
\n2907
\n093
\n550\u2013599
\nActs + 1 Peter
\n2906
\n048
\n400\u2013499
\nActs + James + 1-2 Peter + 1-3 John + Romans +
\n1-2 Corinthians + Ephesians + Philippians + Colossians + 1 Thessalonians + 1-2 Timothy + Titus + Philemon + Hebrews<\/p>\n
Table 13.9. Manuscripts with Acts and the Catholic Epistles<\/p>\n
The book of Acts appears in single-book codices and is joined with the Gospels and Catholic Epistles. Interestingly, the material evidence does not reflect a joining of Acts to the Pauline Epistles.
\nCatholic Epistles. In addition to being joined with Acts, these books are also joined to the following:<\/p>\n
LDAB
\nGA
\nDate
\nContents
\n3070
\n0247
\n400\u2013599
\n1-2 Peter
\n2840
\n0251
\n400\u2013599
\n3 John + Jude
\n2565 +
\n[220465]
\nP72 (= Jude + 1-2 Peter)
\n310\u2013350
\n[325\u2013399]
\nNativity of Mary \/ Protoevangelium of James + 3 Corinthians + Ode of Solomon 11 + Jude + Melito\u2019s Peri Pascha + 1-2 Peter + [Apology of Phileas + Psalms 33\u201334]
\n107771
\n(Coptic)
\n300\u2013399
\nPeri Pascha (Melito) + 2 Maccabees + 1 Peter + Jonah + Pachomian homily on Easter
\n108050
\n(Coptic)
\n600\u2013699
\nPsalms + Hebrews + James + end of a martyrium<\/p>\n
Table 13.10. Manuscripts with the Catholic Epistles or Catholic Epistles and non-New Testament books<\/p>\n
LDAB 2565 (includes P72, P.Bodm. VII [Jude] + P.Bodm. VIII [1-2 Peter]) is a fourth-century composite codex with an array of contents and an unusual compilation and codicology. Although most New Testament scholars treat P72 (Jude and 1-2 Peter) as a \u201csingle-papyrus continuous\u201d manuscript, recent codicological analysis shows that \u201cP.Bodmer VIII was only joined to this unit and the other quires of the composite codex at a later time,\u201d even if P.Bodm. VII and P.Bodm. VIII are considered to be from the same copyist. Brent Nongbri concludes that it appears that the interval between the production of these two works was not long, but his study does call into question the notion of P72 as one continuous manuscript. Thus, 1-2 Peter (P.Bodm. VIII) were together added secondarily to the codex that already contained, with continuous pagination, P.Bodm. V (Protoevangelium of James) + P.Bodm. X (3 Corinthians) + P.Bodm. XI (Ode of Solomon 11) + Jude (P.Bodm. VII), along with the other books in the codex. Therefore, Jude and 1-2 Peter were not part of the same manuscript originally, even though the current form of the codex reveals that they were eventually joined together. But, of course, the final compilation of the codex reveals that these Catholic Epistles were joined together along with many other noncanonical works as well.
\nPerhaps this situation does justify the claim that \u201cthis manuscript seems to be an oddity in that it also includes a number of noncanonical works.\u2026 Thus P72 appears to be the exception that proves the rule\u201d or \u201cthat it was made for private use and not to be taken as typical of early Christian manuscripts.\u201d But since the primary compilation had already included Jude with noncanonical books, some scribes clearly had no issue with combining canonical and noncanonical material together. Even when scribes made a secondary compilation, they had no issue with adding 1-2 Peter to this codex that contained works of mixed status. Therefore, we cannot claim, as Kruger does, that in P72 \u201cwe do have 1 and 2 Peter and Jude preserved in the same third-century manuscript,\u201d unless we mean in a secondary, later compilation. Its \u201coddity\u201d lies in its complicated compilation history, not necessarily in its inclusion of noncanonical books, since both the primary and later compilations attest to a codex with mixed contents.
\nRevelation. Revelation is not joined with any other New Testament book in Greek except in the pandects (see below). The scribe of LDAB 2786 (includes P18; dated to 200\u2013299), which includes Revelation 1:4\u20137, used the blank side of LDAB 3477 (Ex 40:26\u201338; dated to 200\u2013299).
\nSignificance. The codices from 100 to 700 reveal a tendency for the surviving evidence most commonly to be single pages from manuscripts of unknown length. Presumably, these copies would be more portable and user-friendly than the scroll or larger codices in which many books were joined together. But if this conclusion is true, then what might it mean for the theory that early Christians adopted the codex for a canonical purpose? Would it not suggest that early Christians did not use the codex for this purpose?
\nThe relatively fewer examples of multiple-book codices from the same period reveal some interesting features. First, most of the multiple-Gospel codices have only two or three Gospels; some have all four. In Greek manuscripts, there is no evidence of a noncanonical Gospel being joined to a canonical Gospel. Second, no codex contains all the traditional letters of Paul, but codices containing Paul\u2019s letters usually contain only Pauline letters\u2014not other works (see discussion of pandects below). Third, some early evidence of the Catholic Epistles remains, but there is almost no evidence of an early collection of these letters (LDAB 2894 is dated between 500 and 699). Fourth, Revelation is never joined to other New Testament books from this period in Greek except for the pandects (discussed below). Fifth, some codices show that early Christians did join books from the traditional twenty-seven-book New Testament with other important works such as 1 Clement and the Shepherd, and even works such as Protoevangelium of James, the Infancy Gospel of Jesus, and the Acts of Pilate. Therefore, without the canon lists, the material evidence could be used to show a fluid New Testament canon: (1) single-book codices could be used to show an open canon and (2) the codices containing books of mixed status could be used to show that the New Testament was in constant flux. But we have the canon lists, and they do help us interpret the contents of our manuscripts.
\nWe now need to examine the early pandects. Given this background, these codices stand out in several ways.
\nThe early pandects. Finally, we turn to the early pandects, that is, the megacodices that contained the Old and New Testament. Because Codex Vaticanus\u2019s fourth-century text breaks off at Hebrews 9:14, we will leave it aside in this discussion; its original contents are simply too uncertain. The contents of Codex Sinaiticus (fourth century) and Codex Alexandrinus (fifth century) will be considered.Codex Sinaiticus (01) has Gospels, Paul, Acts and the Catholic Epistles, Revelation, Barnabas, and the Shepherd, while Codex Alexandrinus (02) has Gospels, Acts and the Catholic Epistles, Paul, Revelation, and 1-2 Clement. Both manuscripts have books added to the end of the traditional twenty-seven-book New Testament. How did the person who ordered these codices esteem these books? We have already seen that 1 Clement and the Shepherd were joined to other New Testament books above in 2806 (400\u2013450) and 107965 (300\u2013399), respectively. Second Clement is found elsewhere only in Codex Hierosolymitus and a Syriac version, while Barnabas is preserved in Hierosolymitus, nine late Greek manuscripts, a papyrus fragment, and a Latin version. Therefore, what does their inclusion in a biblical codex mean? Did early Christians believe that, by including books in these codices, each book was canonical, as Elliott says? Since we have the early Christian canon lists and, in some cases, specific statements on these books, we are able to evaluate the less clear evidence of their appearance in the pandects with the more explicit statements from these lists.
\nCodex and canon: differences between them. In this final section, we describe the differences between canon and codex and suggest a theory for interpreting the contents of early Christian codices. The codex is a material document that contains one book or many. The Gospels and Pauline letters, respectively, were sometimes grouped together into one codex, but more often were not. In unusual cases, collections of the Gospels, Paul, Acts and Catholics Epistles, and Revelation and other books could be combined into one pandect along with the Old Testament books. Thus, we might compare an ancient pandect to a library or a collection of books. Early Christians read many kinds of books, and this is reflected in the remains of their codices.
\nA canon list, on the other hand, represents a church father\u2019s or scribe\u2019s list of authoritative books. The list may be presented as final, closed, and exclusive, as in the case of Gregory of Nazianzus, or as a simple, exclusive list of books, as in the case of Origen. The list is exclusive and contains books recognized by most or all churches as authoritative for church doctrine. Therefore, one would expect to discover differences between the contents of a codex and the canon. But do the differences amount to different canons, one represented by the codex and another represented by the canon list? I propose that the early Christian tri-fold description of religious literature accounts not only for the exclusive canon lists but also for the inclusion of useful books from the middle category of books in the codex. Thus, when we compare early Christian canon lists with the Christian pandects, we see useful books alongside canonical books, and this picture agrees with their own description.
\nThe canon lists as key. When we examine the evidence of the pandects, we cannot forget what we learned from the canon lists previously. Only one late fourth-century canon list, Apostolic Canons 85, includes some of these four extra books in its canon: 1-2 Clement. But this one list shows that only a sector of churches in Syria thought 1-2 Clement were in the New Testament canon; that is, 1-2 Clement enjoyed a very limited canonical reception. It should be noted that 02 places 1-2 Clement at the end and does not join them to the Catholic Epistles, as in the Apostolic Canons; thus, 02 is probably not following the Apostolic Canons. Therefore, given the other canon lists and the tri-fold description of religious literature, whoever ordered Codex Alexandrinus probably considered 1-2 Clement as useful but not canonical scripture. That is, early Christians placed the two kinds of scripture between the same covers but kept them conceptually distinct.
\nThe Shepherd was the most popular book of these four works, given its manuscript attestation and citations. But church fathers never recognized it as canonical. Athanasius describes it as \u201crather useful\u201d but does not include it in the canon (De decretis 18.3.2) and later relegated it to books that are to be read but are not canonical (Festal Letters 39). Jerome states it is not in the canon (Prologus Galeatus) but still useful (Vir. ill. 10). Rufinus includes the Shepherd among his \u201cecclesiastical\u201d books that may be \u201cread in the churches but appeal should not be made to them on points of faith\u201d (Commentarius in symbolum apostolorum 36). Eusebius includes the Shepherd among his disputed-spurious books\u2014not his \u201cacknowledged\u201d books (Hist. eccl. 3.25). Codex Claromontanus includes the Shepherd in its list with an obelus, probably indicating that it is not canonical. Although Origen describes the book as \u201cvery useful\u201d and \u201cdivinely inspired\u201d (Commentarii in Romanos 10.31), he does not include the work in his list of New Testament books (Hom. Jos. 7.1), probably because he knew that not all Christians recognized its canonicity. Finally, the Muratorian Fragment designated the Shepherd for private, not public, reading. Thus, the case of the Shepherd shows that a book may enjoy great popularity and even be cited as divinely inspired scripture but not be included in the canon. Given this context, the person who ordered Sinaiticus probably considered the Shepherd an important and useful book but not a canonical one, especially in the fourth century.
\nThe Epistle of Barnabas was also a popular book. Clement of Alexandria wrote a commentary on it (according to Eusebius, Hist. eccl. 6.14.1) and mentions \u201cthe apostle Barnabas\u201d several times in his Stromata. Origen calls it a \u201ccatholic epistle\u201d (Contra Celsum 1.63). However, once again, church fathers do not include the book in their canon lists. Eusebius lists the book among his disputed-spurious books (Hist. eccl. 3.25), and Codex Claromontanus includes it with an obelus. Jerome calls the work an apocryphon (Vir. ill. 6). Therefore, Barnabas may have been considered useful scripture, like the Shepherd, but there is no clear evidence that early Christians reckoned it among the canonical books. Its presence alongside the canonical books in a fourth-century codex only enhances the point that the codex was a sacred library or collection of scriptural books, not the concretization of the closed, biblical canon.<\/p>\n
CONCLUSION: WHAT HATH CANON TO DO WITH CODEX?<\/p>\n
If, as has been argued, the canon lists are the key to interpreting the contents of our codices, why have these lists not been used this way before? The reason may be that the twentieth century saw a wonderful advance in knowledge of the history of the Bible. The field of archaeology matured, and many biblical manuscripts were discovered for the first time. Now, realia could be esteemed more than the \u201ccultural memory\u201d enshrined in texts. In this same spirit, the manuscripts themselves could now be viewed as artifacts and their material remains could be evaluated accordingly. What could they tell us? Helpfully, they do provide wonderful evidence of the biblical text, and they have long been used as such by textual critics. Beyond that, they may tell us which copies were for private or public use. The codex form perhaps reveals what books were scriptural both according to the quantity of manuscripts and the physical linkages between books. Conversely, a book written on media other than the codex (e.g., the scroll) could perhaps reveal a work\u2019s unscriptural status, keeping in mind that not all \u201cscriptural\u201d texts were canonical for the early Christians. Although these advances in the field are welcomed, we need to interpret the material data primarily according to early Christian views of their own literature, a warning that sometimes appears to go unheeded.
\nBut the matter need not end there. Manuscripts provide for us the textual form of the biblical books, such as longer and shorter versions of Mark, and sometimes, the multibook codex agrees with Christian statements (e.g., a four-Gospel codex could be interpreted as corresponding to Christian testimony on the matter). However, as we saw, there are codices, which reflect a more open and fluid canon\u2014but only if we think there is a strong or indicative connection or correspondence between codex and canon. This is because of the manuscripts that show that early Christians copied and promulgated other books not considered to be in their canon. But once we realize that the canon of Christian scripture is to be discovered among the many canon lists and not in manuscripts, the picture is clarified, and we see that the codex was primarily a repository of the varied and many books that Christians read. The canon lists can be used to interpret the various contents of the manuscripts according to the actual early Christian categories used to describe their own literature. In short, Christians did prefer the codex form, perhaps for convenience or portability or other factors, but they did not do so primarily as a means of determining or developing the canon. That important function was reserved for a much more explicit means\u2014namely, the canon lists we find in early Christianity.<\/p>\n
Key Takeaways
\n\u25ba Early Christians preferred the codex form over the bookroll or scroll form.
\n\u25ba Early Christians had a category for books that they considered useful and good but not canonical. Sometimes early Christians placed the two kinds of scripture (canonical and noncanonical-but-still-useful) between the same covers but kept them conceptually different.
\n\u25ba Early Christian canon lists provide the best way to interpret the varied contents of the early Christian codices.
\n\u25ba Just because a book was in the codex does not mean it was therefore deemed canonical.<\/p>\n
Origen
\nEusebius
\nCyril
\nAthanasius
\nLaodicea
\nApostolic Canons
\nEpiphanius
\nSyriac List
\nGregory of Nazianzus
\nAmphilochius
\nMatthew\u2013John
\n4 Gospels
\n4 Gospels
\nMatthew\u2013John
\nMatthew\u2013John
\nMatthew\u2013John
\n4 Gospels
\nMatthew\u2013John
\nMatthew\u2013John
\nMatthew\u2013John
\n1-2 Peter
\nActs
\nActs
\nActs
\nActs
\n14 of Paul
\n14 of Paul
\nActs
\n? Revelation?
\nActs
\nJames
\n14 of Paul\u2019s letters
\nJames
\nJames
\nJames
\nJames
\nActs
\nGalatians
\nActs
\nRomans
\nJude
\n1 John
\n1-2 Peter
\n1-2 Peter
\n1-2 Peter
\n1-3 John
\nJames
\n1-2 Corinthians
\n14 of Paul\u2019s letters
\n1-2 Corinthians
\nLetters of John
\n1 Peter
\n1-3 John
\n1-3 John
\n1-3 John
\nJude
\n1-2 Peter
\nRomans
\nJames
\nGalatians
\nActs
\n? Revelation?
\nJude
\nJude
\nJude
\n1-2 Peter
\n1-3 John
\nHebrews
\n1-2 Peter
\nEphesians
\n14 of Paul\u2019s letters
\n14 of Paul\u2019s letters
\nRom
\nRom
\n1-2 Clement
\nJude
\nColossians
\n1-3 John
\nPhilippians
\n(Revelation)
\n1-2 Corinthians
\n1-2 Corinthians
\nActs
\nRevelation
\nEphesians
\nJude
\nColossians
\nDisputed:
\nGalatians
\nGalatians
\nPhilippians
\n1-2 Thessalonians
\nJames
\nEphesians
\nEphesians
\n1-2 Thessalonians
\n1-2 Timothy
\nJude
\nPhilippians
\nPhilippians
\n[1]-2 Timothy
\nTitus
\n2 Peter
\nColossians
\nColossians
\nTitus
\nPhilemon
\n2-3 John
\n1-2 Thessalonians
\n1-2 Thessalonians
\nPhilemon
\nHebrews
\nSpurious:
\nHebrews
\nHebrews
\nJames
\nActs of Paul
\n1-2 Timothy
\n1-2 Timothy
\n1-?2? Peter
\nShepherd of Hermas
\nTitus
\nTitus
\n1-?2-3? John
\nApocalypse of Peter
\nPhilemon
\nPhilemon
\n? Jude?
\nBarnabas
\nRevelation
\n? Revelation?
\nDidache
\nTo be read:
\n? Revelation?
\nDidache
\n? Gospel according to Hebrews?
\nShepherd of Hermas<\/p>\n
Table 13.11. Eastern canon lists up to AD 400<\/p>\n
Muratorian Fragment
\nClaromontanus
\nMommsen Catalog
\nJerome
\nRufinus
\nBreviarum Hipponensis
\nAugustine
\nInnocent I
\nMatthew\u2013John
\nMatthew
\nMatthew
\nMatthew\u2013John
\nMatthew\u2013John
\n4 Gospels
\nMatthew\u2013John
\n4 Gospels
\nEpistles of John
\nJohn
\nMark
\n14 of Paul
\nActs
\nActs
\nRomans
\n14 of Paul\u2019s letters
\nActs
\nMark
\nJohn
\nActs
\n14 of Paul\u2019s letters
\n14 of Paul\u2019s letters
\n1-2 Corinthians
\n1-3 John
\n1-2 Corinthians
\nLuke
\nLuke
\nJames
\n1-2 Peter
\n1-2 Peter
\nGalatians
\n1-2 Peter
\nEphesians
\nRomans
\n13 of Paul\u2019s letters
\n1-2 Peter
\nJames
\n1-3 John
\nEphesians
\nJude
\nPhilippians
\n1-2 Corinthians
\nActs
\n1-3 John
\nJude
\nJude
\nPhilippians
\nJames
\nColossians
\nGalatians
\nRevelation
\nJude
\n1-3 John
\nJames
\n1-2 Thessalonians
\nActs
\nGalatians
\nEphesians
\n1-3 John
\nRevelation
\nRevelation
\nRevelation
\nColossians
\nRevelation
\n1-2 Thessalonians
\n1-2 Timothy
\n1-2 Peter
\n1-2 Timothy
\nRomans
\nTitus
\nEcclesiastical:
\nTitus
\nPhilemon
\nColossians
\nHerm.
\nPhilemon
\nTitus
\nPhilemon
\nTwo Ways
\nHebrews
\n1-2 Timothy
\n1-2 Peter
\nJudgement of Peter
\n1-2 Peter
\nEpistle to the Laodiceans (forged)
\nJames
\n1-3 John
\nEpistle to the Alexandrians (forged)
\n1-3 John
\nJude
\nJude
\nJude
\nJames
\n2 epistles of John
\n\u2192 Barnabas
\nActs
\nWisdom
\nRevelation
\nRevelation
\nRevelation
\nActs
\nApocalypse of Peter (private reading)
\n\u2192 Shepherd of Hermas
\nShepherd of Hermas (private reading)
\n\u2192 Acts of Paul
\n\u2192 Apocalypse of Peter<\/p>\n
Table 13.12. Latin canon lists up to AD 400<\/p>\n
CHAPTER FOURTEEN<\/p>\n
MYTHS ABOUT EARLY TRANSLATIONS<\/p>\n
THEIR NUMBER, IMPORTANCE, AND LIMITATIONS<\/p>\n
Jeremiah Coogan<\/p>\n
Translations of the New Testament have consistently been part of Christian worship and mission. Over the centuries, Christians have worshiped in languages as diverse as Latin, Syriac, Ethiopic, Hindi, and Mandarin, to name only a few. In this chapter I discuss the earliest translations (\u201cversions\u201d) of the New Testament: those in Latin, Syriac, and Coptic.
\nMy objective in this chapter is not to advance novel claims but to summarize the current state of scholarship. I begin by discussing how translation works and what this means for responsibly using evidence from a translation (\u201creceptor text\u201d) to reconstruct the text from which it was translated (\u201csource text\u201d or Vorlage). I then briefly survey the translations of the New Testament in Latin, Syriac, and Coptic. I focus on major translations and revisions, on manuscript evidence, and on how language and translation technique affect use of these translations for textual criticism of the Greek New Testament.
\nA particular concern will be to make clear what we know about the total numbers of New Testament manuscripts in each language. This is crucial because, while there is a rich body of manuscript evidence for the text of the New Testament in a number of early translations, apologetic and popular literature is filled with unguarded overstatement on this point. For example, James MacDonald says that there are \u201cnearly 25,000 early manuscripts of the Bible,\u201d and he is not alone in claiming this large number. Craig Blomberg also says there are twenty-five thousand, although he does not claim they are all early. In fact, as we will see, while exact numbers for the versions cannot be attained, these estimates are far too large and thus give a misleading picture.
\nFinally, as the conclusion of the chapter, I briefly discuss early translations of the New Testament as witnesses to Christian worship, reading, teaching, and devotion. Translations of the New Testament are not, in the first instance, tools for textual criticism of the Greek New Testament. Their primary function has always been to enable Christians to encounter Scripture as the living Word of God.<\/p>\n
USING TRANSLATIONS FOR TEXTUAL CRITICISM OF THE GREEK NEW TESTAMENT<\/p>\n
Every language is different. Those who have undertaken the difficult task of learning a modern language know this well. Idioms confuse, grammatical structures diverge, articles and prepositions are notoriously tricky. For example, if I wanted to tell someone that I am hungry in German, I might say, \u201cIch habe Hunger.\u201d Woodenly, someone might translate this as \u201cI have hunger.\u201d But no one who really knows both English and German would translate the expression that way. Instead, someone seeking to communicate my statement in good English might simply say, \u201cI\u2019m hungry.\u201d The two expressions perform the same communicative task, but structurally they do not quite match up.
\nSuch challenges increase with longer and more complicated texts. Moreover, English and German are rather closely related to each other. Even readers who do not know German probably recognize cognates (German Hunger and English hunger, for example). By comparison, Greek is much further separated from Syriac (a Semitic language) or Coptic (an Afro-Asiatic language). Even a relatively close cousin such as Latin poses difficulties. For example, while Greek has a definite article (\u1f41\/\u1f21\/\u03c4\u03cc, ho\/h\u0113\/to, only roughly corresponding to English \u201cthe\u201d), Latin does not. Translation requires adaptation in order to convey the source text within the vocabulary and grammar of the receptor language.
\nOften early translations do not offer clear evidence for deciding between two or more Greek readings. This does not mean that Latin, Syriac, Coptic, or other versions of the New Testament are bad translations. Limitations for reconstructing the text of the Greek New Testament result from inevitable differences between the source language (Greek) and the receptor language. It would be similarly difficult to reconstruct the exact Greek source text used for even the most wooden modern translations of the New Testament.
\nWe might summarize the challenges in using a translation to perform textual criticism of the Greek New Testament under four headings.
\nLinguistic differences. First, there are the linguistic differences I have just discussed. Because languages do not line up perfectly, some languages are not well suited to reflect certain features in a source text. To return to the previous example, Latin translations are not a reliable guide for the presence or absence of the Greek article, since there is no corresponding feature in Latin.
\nTranslation technique. This leads to a second challenge: translation technique (often discussed under the German name \u00dcbersetzungsweise). We must attend not only to what the receptor language can do but also to what the translator actually does. Sometimes this works in our favor. Translators occasionally go out of their way to reflect vocabulary, grammatical constructions, or syntactical features from the Greek source text consistently in Latin, Syriac, or Coptic translation. For example, while Latin does not have a definite article (roughly equivalent to English \u201cthe\u201d or Greek \u1f41\/\u1f21\/\u03c4\u03cc, ho\/h\u0113\/to), translators sometimes sought to indicate the presence of the Greek article using the Latin demonstrative hic (roughly equivalent to English \u201cthis\u201d). The Greek \u1f41 \u03ba\u03cc\u03c3\u03bc\u03bf\u03c2 (ho kosmos, \u201cthe world\u201d) is often translated hic mundus (\u201cthis world\u201d). We would be mistaken to conclude that the Latin hic mundus translates a Greek source text including the demonstrative pronoun \u03bf\u1f57\u03c4\u03bf\u03c2 (houtos, \u201cthis\u201d); the Latin translation hic mundus does, however, offer strong evidence for the Greek article in the source text.
\nOften, however, even when the receptor language could have reflected a given feature in a way that would enable us to reconstruct the Greek source text, the translator chose not to do this. Often it is clear that the Latin text resembled the textual tradition of the Greek New Testament, but specific points of variation are impossible to discern. The translator might not think it necessary to represent a word in the source text independently by its own word in the translation. Conversely, the translator may add one or several words in order to render the Greek effectively in the receptor language. Likewise, a given word in Greek can have multiple translation equivalents in the receptor language, or vice versa. In such cases we are aided by studies of translation technique, but these often leave multiple Greek readings as possibilities. As a result of such translation decisions, it is frequently difficult to reconstruct the source text. Many scholars thus prefer to use translations for confirmation rather than for reconstruction, but even here we must be careful. As a rule, we are on firmest ground when using early translations as evidence for additions, omissions, and transpositions of a phrase or more. For example, the absence of Mark 16:9\u201320 is quite clear in the fifth-century Codex Bobiensis (VL 1; k).<\/p>\n
Figure 14.1. Codex Bobiensis (Turin, Biblioteca Nazionale Universitaria, 1163 [G.VII.15]; Gregory-Aland k; Vetus Latina (1) preserves the second half of Mark, followed by the first half of Matthew. The rest of the manuscript is lost. The \u201clonger ending\u201d of Mark (16:9\u201320) was not included in this manuscript, which was copied in Africa in the fourth or fifth century. It offers our earliest manuscript evidence for the New Testament in Latin.<\/p>\n
Reconstructing smaller-scale readings in the Greek source text requires caution and restraint. As a general tendency, we observe that within a given language, translation tendencies shift over time from being oriented to the idiom and grammar of the receptor language to being oriented to the idiom and grammar of the source language. This tendency toward revision demonstrates ongoing concern for both the authoritative status of a Greek New Testament text and the accuracy of the translation.
\nThese first two considerations relate directly to the problem of translation. Languages differ from one another, and translators choose various approaches in order to navigate these differences. These two challenges, however, intersect with two other considerations, related to the nature of textual transmission.
\nDetermining the version\u2019s initial text. A third challenge is the question of the initial text in the receptor language. A sixth-century Latin manuscript or a tenth-century Syriac one will not exactly preserve the initial translation into that language. All the mishaps that affect the Greek New Testament also affect a translation. Scribes make mistakes by skipping or repeating text. Parallel Gospel passages are harmonized, and Old Testament citations are modified. Translations sometimes also undergo stylistic revisions (especially in Latin) or dialectal shifts (Coptic). Before they can be used for textual criticism of the Greek New Testament, then, translations must themselves be studied and edited. If, for example, we lack a critical edition for a biblical book in the Sahidic dialect of Coptic, then it is problematic to use \u201cthe Sahidic version\u201d for reconstructing a Greek text of that book. Textual criticism thus involves not only the New Testament in Greek but also the New Testament in translation. Despite major contributions over the past century, much remains to be done in this area.
\nFurther contact with the Greek. Finally, contact with the Greek New Testament did not cease after an initial moment of translation. There were intentional revisions toward the Greek in a number of languages (for example, the Latin Vulgate, the Syriac Peshitta, the Philoxenian Syriac, the Harklean Syriac), bilingual manuscripts, and Christian contexts where the Greek New Testament was used alongside the New Testament in other languages. Each complicates matters. Which Greek text is reflected in a given translational reading? For example, does a given Latin reading reflect a second-century Greek source text of the Old Latin Gospels, or does it reflect Jerome\u2019s revision in the late fourth century, using a different Greek source text?
\nUsing early translations of the New Testament as witnesses to the Greek New Testament requires caution and hard work. Sometimes a translation does not clearly support a single Greek reading. In other cases, the state of scholarship on a translation or the translation\u2019s complicated transmission history do not permit confidence about what Greek source text a reading in the translated language renders.
\nWhy study the versions? Given these challenges, why study the versions at all? Prior to twentieth-century manuscript discoveries in Egypt, early translations (completed before most then-known manuscript witnesses) narrowed the chronological gap between textual composition and extant evidence. Translations from the second or third century offered a way to push back the date of the earliest evidence. Yet in light of recent manuscript finds, some of which themselves date back to the second and third centuries, this reason is less compelling.
\nThe early translations consistently represent a New Testament text quite similar to what we find in Greek. Yet, as translations, they are limited. Contemporary New Testament textual criticism focuses on relatively minor differences in the Greek New Testament text. Many of these are inaccessible through translation, as a result of either linguistic difference or translation technique. It is true that some translations tend to agree more with certain streams of Greek textual transmission than with others (for example, Coptic translations tend to be similar to the Greek papyri also largely originating in Egypt, while Gothic translations more closely resemble the Byzantine text). Nonetheless, at relatively few points do translations shift the balance of evidence in favor of a particular Greek reading or plausibly attest a lost Greek reading. It is in the case of the exceptions, longer variation units such as Mark 16:9\u201320 or John 7:53\u20138:11, that translations play their most important role in the study of the Greek New Testament text.
\nIn a broader sense, however, the versions attest the remarkable integrity with which the New Testament text was transmitted across time and geography. If the versions reflected more extensive variation, they would be more central to textual criticism of the Greek New Testament. Their relative lack of significance for determining specific readings reflects the remarkable similarity between the Greek New Testament and early translations.<\/p>\n
A SURVEY OF EARLY TRANSLATIONS<\/p>\n
The earliest and most important translations of the Greek New Testament are those in Latin, Syriac, and Coptic (listed in likely order of antiquity). These are the best attested in manuscript evidence, have been studied most thoroughly, and are cited regularly in editions of the Greek New Testament. Other translations, however, quickly followed these first three. By the late fourth century, most or all of the New Testament had been translated into Gothic, Ethiopic, and Armenian. Still other translations followed in late antiquity and the early Middle Ages, including Georgian, Slavonic, Christian Palestinian Aramaic, and Arabic. By discussing the value and challenges of using the three earliest translations, I hope that the reader will be equipped to explore others as well.
\nLatin. Latin is the best-attested and best-studied early translation of the New Testament, so it is a natural place to begin. At least some New Testament books may have been translated by the end of the second century, including the Gospels and Paul. The remaining books were translated by the end of the third century. These translations experienced later revisions, the most important of which is known as the Vulgate.
\nOld Latin. The early Latin translations are known simply as the Old Latin (or Vetus Latina) and attest a form of the New Testament text going back to the second century. Within the Old Latin New Testament, however, we find significant textual diversity. At one time it was thought that multiple independent translations might lie behind this diversity, but this appears unlikely in light of recent research. If multiple independent translations of individual books were made, they left few identifiable traces. Theories of regional text-types have experienced a similar fall from scholarly favor. For much of the twentieth century, scholars distinguished between \u201cAfrican\u201d and \u201cEuropean\u201d texts, each divided into further regional text-types. More recent scholarship has been less enthusiastic about this model; as Philip Burton suggests, a theory of regional text-types oversimplifies complex textual evidence. While early citations of the New Testament locate certain text-forms with particular geographic regions, a number of texts traveled widely, and some regions had more than one circulating text. Furthermore, while clusters of witnesses often exhibit patterns of textual agreement with one another, clearly differentiated text-types seem hard to maintain. While parallel independent translations no longer provide a satisfactory model to explain textual diversity in the Old Latin New Testament, this diversity indicates ongoing contact with and revision toward Greek manuscripts.
\nOngoing revision of the Old Latin New Testament makes sense in light of both social dynamics, which frequently included bilingual Christian communities, and the early Latin approach to translation. The first of these points is illustrated by bilingual manuscripts that place Latin alongside Greek (most famously Codex Bezae, 05).<\/p>\n
Here, however, we focus on the second: the Old Latin translation is oriented toward the idiom of Latin rather than the vocabulary and grammar of the Greek source text. Later revisions, starting as early as the third century, introduce a greater formal equivalence between the Latin and Greek.
\nWhat does this mean in practice? On the one hand, idiomatic tendencies in the Old Latin mean that one must be careful about assuming that small-scale features of the Latin translation reflect particular features of the Greek source text. Studies of translation technique continue to explore translation tendencies that might assist in reconstructing the Greek source text. Furthermore, different books of the Bible and different points of revision demonstrate different tendencies, so we must not impose more unity on the texts we call \u201cthe Old Latin\u201d than they actually exhibit. A side effect of such later revisions was to introduce more diversity from the Greek tradition into the Latin New Testament by using slightly different Greek texts from those that formed the basis for earlier Latin translation. As a result, multiple Latin readings for a given variation unit occasionally reflect different Greek readings. In practice, however, it is difficult to discern when a given Latin reading reflects contact with a different Greek text and when it simply reflects an inner-Latin variation.
\nThese complexities are not unique to the Old Latin; Syriac, Coptic, and other languages also exhibit a general tendency toward idiomatic early translations and later revisions, with greater focus on formal equivalence of vocabulary and grammar. As in the case of the Old Latin, these revisions in Syriac and Coptic introduce further textual diversity from the Greek tradition. In Latin, the overall movement toward formal equivalence continues in the later revision known as the Vulgate.
\nVulgate. While the Vulgate is traditionally associated with Jerome of Stridon (347\u2013420), he is responsible only for the Gospels (completed 383\/384 at the behest of Pope Damasus). Rather than producing a new translation, Jerome revised the Old Latin Gospels. He began with a Latin text that circulated in Italy and revised it toward a Greek text with strong similarities to Codex Sinaiticus (01). Others in the fourth and fifth centuries followed Jerome in revising the Latin New Testament. These revisions are anonymous, but they may be the work of one or more Pelagian scholars. Jerome continues the trend of revising earlier translations toward greater conformity with Greek vocabulary and grammar. With varying degrees of consistency, other Vulgate revisers also mirror the grammar and vocabulary of their Greek base texts more closely than the Old Latin did. This means that, on the whole, the Vulgate is more suited to studying its Greek source text than is the Old Latin.<\/p>\n
Figures 14.2 and 14.3. Codex Bezae (Gregory-Aland 05; Vetus Latina 5) is a bilingual manuscript containing the Gospels (Matthew\u2013John\u2013Luke\u2013Mark), 3 John, and Acts. The manuscript dates from ca. AD 400 and may have been produced in Berytus (modern-day Beirut). The pericope adulterae (John 7:53\u20138:11) is included in the manuscript, as seen in part here with John 7:44\u20138:4<\/p>\n
The Vulgate did not immediately replace the Old Latin. On the contrary, the Vulgate took several centuries to establish itself as a widespread text, and the Old Latin flourished well into the Middle Ages. The texts we possess always reflect mixture between earlier and later translations; there are no \u201cpure\u201d manuscripts of either the Vulgate or the Old Latin. Not only might the text of a given book mingle Old Latin and Vulgate readings, but different parts of a manuscript might reflect different textual traditions. For example, a manuscript in which the Gospels primarily reflect the Old Latin might contain the Pauline letters in a form that primarily reflects the Vulgate (or vice versa). Furthermore, the Vulgate developed its own internal diversity, much of which did not result from direct contact with Greek.
\nManuscripts and citations. The earliest extant Latin New Testament manuscripts date from the fourth and fifth centuries. These are primarily parchment codices, many more or less intact. Unlike in Greek and Coptic, there are few papyrus fragments of the Latin New Testament. Unfortunately, there is no comprehensive list of Latin New Testament manuscripts corresponding to the list of Greek New Testament manuscripts kept by the INTF, so we have no reliable estimate of the total number. For the Old Latin, Houghton lists 111 manuscripts containing the New Testament. For the Vulgate, precision is harder to attain; Metzger\u2019s reference to over ten thousand manuscripts of the Vulgate is often cited, but it is imprecise and possibly inaccurate.
\nEven early and extensive manuscript evidence, however, still leaves a gap of several centuries between the initial Latin translations and the first physical evidence. As a result, citations of the Latin New Testament play a key role in the study of its textual development, especially the geographic and chronological use of particular text-forms. This poses challenges, however, because some Latin authors have eclectic citation habits and because later scribes sometimes modified biblical citations in the works they copied. Using citations to study the Greek New Testament thus requires both critical editions of early Christian writings and careful attention to citation habits. (While citations are less central to reconstructing the development of the New Testament in Coptic and Syriac, we find the same methodological concerns there as well.)
\nStanding somewhere between the categories of manuscript and citation are other phenomena that enrich our understanding of the Latin New Testament. Paratexts such as prefaces, chapter lists, and the Eusebian Gospel apparatus are important for the history and transmission of the New Testament. Lectionaries and other noncontinuous manuscripts are similarly significant, though often neglected. These sources illuminate how audiences encountered the New Testament in Latin. (These features are relevant in other translations of the New Testament as well.)
\nLatin translations and textual criticism of the Greek New Testament. In perhaps every way, the Latin is the easiest version to use in studying the Greek New Testament. There is a rich manuscript tradition in Latin. Due to the indefatigable efforts of scholars over the course of the last century, both the Old Latin and the Vulgate exist in careful critical editions, although the Old Latin project is not quite complete. Early Christian texts that cite the Latin New Testament are also generally well served with critical editions. Studies of Latin translation technique and of citation habits of Latin authors are available in many cases, although there are still many open avenues of research.
\nLatin New Testament tradition exhibits diverse approaches to translation technique, multiple revisions against different Greek New Testament texts, and mixture between major textual clusters in Latin as well as all the possibilities for textual corruption that result from manuscript copying. Latin itself, moreover, has limitations in reflecting features of the Greek New Testament. These complexities require caution when using the Latin New Testament for the textual criticism of the Greek New Testament. While it is frequently impossible to decide which of two or more Greek readings a given Latin phrase renders, the Latin textual tradition reflects a New Testament text quite similar to those transmitted in Greek. As we will see, many of the challenges and complexities that we find in the Latin New Testament are similar to those in Syriac and Coptic.
\nSyriac. Syriac is a Semitic language that flourished across the Middle East and central Asia during late antiquity and the Middle Ages. As a result of the rich and ongoing Syriac Christian tradition and the comparatively arid climates where some Syriac Christians live, the Syriac New Testament is well preserved. Indeed, through the seventh century alone there are some 120 manuscripts of the Syriac New Testament.
\nThe richness of Syriac New Testament translations is remarkable. The earliest translations date to the second century. The scholar Tatian (fl. ca. 160\u2013180) wove together multiple Gospel narratives into a work known as the Diatessaron, which circulated widely in Syriac for centuries. Since we lack a text of Tatian\u2019s project, it seems better to begin with the Old Syriac Gospels. The subsequent history of the Syriac includes at least three major revisions, known as the Peshitta, the Philoxenian, and the Harklean. In some books (especially those with a more tenuous place in the Syriac tradition, such as 2 Peter, 2-3 John, Jude, and Revelation) it is difficult to associate extant translations with any of these reliably. In our discussion of the Latin New Testament, we noted two major points of translation or revision (the Old Latin and the Vulgate); by contrast, Syriac has four major textual forms in some books.
\nOld Syriac. The Old Syriac Gospels may have been translated as early as the second century, although the third century is equally plausible. Parts of three manuscripts remain. Two have been known for more than a century: the fourth-century Sinai palimpsest (Sys = Sinai syr. 30) and the fifth-century fragmentary Curetonian manuscript (Syc = BL Add MS 14,451 + Berlin syr. 8 + Deir al-Surian syr. frag. 9). Recently, a third manuscript, a sixth-century palimpsest with portions of all four canonical Gospels (including passages not previously extant in the Old Syriac), was discovered as part of the New Finds at Saint Catherine\u2019s Monastery (Sinai syr. NF 37 + 39).
\nThe Sinaitic and the Curetonian descend from the same initial translation, but both show signs of later revision toward the Greek New Testament. The two are sufficiently different that scholars generally cite them separately. We await further study of the New Finds manuscript and its textual relationship to the other manuscripts of the Old Syriac Gospels.
\nWhile we have no pre-Peshitta manuscripts of Acts and Paul, citations by fourth-century Syriac authors such as Ephrem, Aphrahat, and the anonymous author of the Liber Graduum (\u201cBook of Steps\u201d) attest that these also circulated. Unfortunately, it is difficult to reconstruct these books since we do not have any extensive texts. Even still, it is likely that\u2014as in the case of the Gospels\u2014the Peshitta revised these translations rather than starting anew. These citations pose challenges, but they offer valuable evidence for the early Syriac transmission of the New Testament. It is possible that the Catholic Epistles and Revelation circulated in Syriac before the Peshitta, but no identifiable traces of this remain.
\nPeshitta. The majority of New Testament manuscripts in Syriac are from the version known as the Peshitta (syp). The Peshitta has been the primary Bible of Syriac Christianity for the past fifteen hundred years, although it exhibits internal variation over time, across regions, and among Syriac Christian communities. The Peshitta also served as the source for translations into other languages in the Middle East and central Asia, including Arabic, Middle Persian, and Sogdian.
\nThe Peshitta New Testament was probably translated around the beginning of the fifth century, although parts may have been translated as early as the fourth. Since the earliest extant Peshitta manuscripts date from the fifth century, the gap between the initial production of the Peshitta text and the earliest manuscript evidence is relatively small.
\nDifferences in translation technique suggest that multiple revisers are responsible for the creation of the Peshitta. Except perhaps for the \u201cmajor\u201d Catholic Epistles (James, 1 Peter, 1 John), the Peshitta is a revision rather than a new translation. (This parallels the relationship of the Vulgate to the Old Latin.) The original translation did not include the \u201cminor\u201d Catholic Epistles (2 Peter, 2-3 John, Jude) or Revelation\u2014or if it did, no identifiable traces remain. Peshitta manuscripts are sometimes supplemented with these books using later translations. The pericope adulterae (Jn 7:53\u20138:11) is likewise not part of the original Peshitta, but it is frequently supplemented from a later Syriac translation. The Peshitta shows a stronger orientation to the Greek source language and text than the earlier Old Syriac translations. It often resembles a Byzantine form of the Greek New Testament. Further editorial work on the Peshitta New Testament is desperately needed.<\/p>\n
Figure 14.4. Sinai syr. NF 37, f. 2v. This sixth-century manuscript of the Old Syriac Gospels contains portions of all four canonical Gospels
\nPhiloxenian and associated translations. In the sixth and seventh centuries, two major projects revised the Peshitta in order to produce greater fidelity to the Greek text. The first of these is known as the Philoxenian (syph) because it was commissioned by Philoxenus of Mabbug (d. 523); colophons indicate that the version was completed in 507\/508. This translation was designed for christological controversy; apparently the Peshitta was seen as insufficiently precise at key textual cruxes. While the Philoxenian was intended to reflect the twenty-seven-book edition of the Greek New Testament (presumably by a careful revision of the Peshitta), only the minor Catholic Epistles and Revelation are preserved for us today. This was the first Syriac translation of the minor Catholic Epistles and Revelation, which were absent in the earlier Peshitta. A number of manuscripts circulated with the Peshitta in most books but the Philoxenian translation for the four minor epistles. The Syriac translation of the pericope adulterae (Jn 7:53\u20138:11) is also sometimes attributed to the Philoxenian. The limited manuscript attestation of the Philoxenian tradition may result in part from the fact that it circulated primarily in the Miaphysite Syriac church. In any case, it was mostly eclipsed by the Harklean version.<\/p>\n
Figure 14.5. BnF ms. syr. 296, no. 1\u00ba is among the oldest extant Peshitta manuscripts. While it lacks a dated colophon, the same scribe also produced BL Add. MS 14,425, a manuscript containing Genesis and Exodus dated by a colophon to 463\/464. Shown here is the end of Acts and the beginning of James<\/p>\n
Harklean. A second revision of the Peshitta was completed in 615\/616 and includes the whole New Testament. Named after the bishop and scholar Thomas of Harqel (ca. 570\u2013640), the Harklean version (syh) aspires to maximal equivalence between Greek source text and Syriac translation. Indeed, it regularly defies the grammar and idiom of Syriac in order to imitate its Greek source text. As a result, the Harklean is especially valuable for reconstructing the Greek base text once one has identified the translation techniques used to render Greek into Syriac. The source text for this revision seems close to the Byzantine text except in the Catholic Epistles. The Harklean left a significant manuscript footprint, especially in the Gospels.
\nSyriac translations and textual criticism of the Greek New Testament. Idiomatic Syriac is quite different from Greek. As a result of linguistic difference, the Old Syriac and Peshitta frequently do not provide reliable evidence for many features in Greek, including word order or case, the presence or absence of articles, conjunctions and other small words, and verbal constructions.
\nAs an example, we might think about Mark 15:45. Here some Greek manuscripts (notably 01, 03, 019, 038, 565, followed by NA) read \u03c4\u1f78 \u03c0\u03c4\u1ff6\u03bc\u03b1 (to pt\u014dma, \u201cthe corpse\u201d), while most others read \u03c4\u1f78 \u03c3\u1ff6\u03bc\u03b1 (to s\u014dma, \u201cthe body\u201d). 05 reads \u03c4\u1f78 \u03c0\u03c4\u1ff6\u03bc\u03b1 \u03b1\u1f50\u03c4\u03bf\u1fe6 (to pt\u014dma autou, \u201chis corpse\u201d). Sys reads \u072b\u0720\u0715\u0717 (\u0161aldeh, \u201chis corpse\u201d); sy reads \u0726\u0713\u072a\u0717 (phagreh, \u201chis body\u201d). NA cites sys in support of \u03c4\u1f78 \u03c0\u03c4\u1ff6\u03bc\u03b1 \u03b1\u1f50\u03c4\u03bf\u1fe6 (= 05) and sy in support of \u03c4\u1f78 \u03c3\u1ff6\u03bc\u03b1 (with the majority of manuscripts). But a closer look at Syriac idiom and vocabulary makes this problematic. Both Syriac nouns appear with an affixed possessive (\u201chis\u201d). Since the suffix is idiomatically obligatory in Syriac, however, this does not offer evidence for the presence of the Greek possessive \u03b1\u1f50\u03c4\u03bf\u1fe6 (autou) in the source text for either Syriac rendering. Thus, there is no reason why \u072b\u0720\u0715\u0717 (\u0161aldeh, noun with possessive) in sys should support \u03c4\u1f78 \u03c0\u03c4\u1ff6\u03bc\u03b1 \u03b1\u1f50\u03c4\u03bf\u1fe6 (to pt\u014dma autou) rather than simply \u03c4\u1f78 \u03c0\u03c4\u1ff6\u03bc\u03b1 (to pt\u014dma). Furthermore, there is no tidy way to link \u03c0\u03c4\u1ff6\u03bc\u03b1 (pt\u014dma) to \u072b\u0720\u0715\u0710 (\u0161ald\u0101) and \u03c3\u1ff6\u03bc\u03b1 (s\u014dma) to \u0726\u0713\u072a\u0710 (phagr\u0101). The first two nouns are used only to refer to dead bodies, while the last two can refer to bodies living or dead. Yet in the narrative context, Jesus\u2019 body is dead. Since the body on the cross is a corpse, the Syriac translator of Mark 15:45 might reasonably use \u072b\u0720\u0715\u0710 (\u0161ald\u0101, \u201ccorpse\u201d) to translate \u03c3\u1ff6\u03bc\u03b1 (s\u014dma). In short, the Sys reading \u072b\u0720\u0715\u0717 (\u0161ald\u0101) might support any of the three Greek readings. The Syriac noun \u0726\u0713\u072a\u0710 (phagr\u0101, \u201cbody\u201d) likewise has a sufficiently broad range that it can translate either \u03c0\u03c4\u1ff6\u03bc\u03b1 (pt\u014dma) or \u03c3\u1ff6\u03bc\u03b1 (s\u014dma). The Syp reading \u0726\u0713\u072a\u0717 (phagreh, noun with possessive) might thus also support any of the three Greek readings.
\nEven in the case of the Old Syriac, however, certain aspects of translation technique (for example, the transliteration of proper names) offer reliable evidence even for small-scale features of the Greek source text. Further studies of translation technique are needed. Subsequent revisions of the Syriac New Testament sought increasingly formal parallels with the Greek source text. For example, while the Old Syriac regularly rearranges the word order of its Greek source, the Peshitta does so less often, and the Harklean attempts maximal correspondence, even to the detriment of comprehensibility. Numerous other characteristics reflect a similar pattern of increasingly formal relationship between Syriac and Greek.<\/p>\n
Figure 14.6. London, BL Or. MS 5707 (Gregory-Aland 086; fa 6) is a sixth-century Greek-Coptic bilingual manuscript of John\u2019s Gospel with Greek in the left column of each page and Fayumic Coptic in the right. It was later erased and over-written. Shown here is John 3:5\u20137 in Fayumic partially visible underneath a later Arabic text<\/p>\n
On the whole, New Testament textual critics have focused on reconstructing the second- or third-century Greek source text of the Old Syriac; they have been comparatively less interested in the later Greek texts toward which later Syriac versions were revised. It is this early source text, however, that is hardest to reconstruct. Continued research, both on the transmission history of the Syriac New Testament in its various versions and on Syriac sources that cite the New Testament, is still needed.
\nCoptic. Coptic is a stage of the Egyptian language that flourished in the third through tenth centuries. Egypt\u2019s arid climate preserved numerous Greek and Coptic manuscripts on papyrus and parchment in city dumps from late antiquity. For all Coptic dialects except Bohairic (which continues to be used as a liturgical language by the Coptic church), this is the primary source of manuscript evidence.
\nPaleographic dating of Coptic manuscripts is notoriously difficult, but physical evidence for the New Testament begins in the third century at the earliest. Significant manuscript finds datable to the fourth century suggest that the bulk of initial translation was completed in the later third and fourth centuries, although revision toward the Greek New Testament continued.
\nWhile we know nothing about the Coptic translators or the initial contexts for translation, it is plausible that the development of Coptic writing systems and the translation of early Christian texts accompanied the rise of Egyptian monasticism.
\nCoptic refers to a continuum of related dialects distributed along the length of the Nile; these dialects had slightly different writing systems. New Testament texts were translated directly into a number of dialects. In other cases, New Testament texts were not freshly translated from Greek but instead reflect adaptation of translations in a neighboring Coptic dialect (dialectal transposition). In some cases, debate continues about whether direct translation or dialectal transposition better explains the evidence. Even in the case of direct translation, interference from existing translations in other dialects is possible. In cases of dialectal transposition, moreover, revision toward the Greek and bilingual use alongside it continued throughout late antiquity.
\nSahidic. The Sahidic dialect (sa) was the most influential in late ancient Egyptian Christianity. The entire New Testament is preserved in Sahidic, although some books are attested better than others. As one might expect, manuscript evidence is strongest for the Gospels. Compared with Bohairic, there are relatively few Sahidic manuscripts. The manuscripts derive primarily from excavations of late ancient trash heaps. While a comprehensive list does not exist, Christian Askeland suggests that there are between 350 and 700 manuscripts of the Sahidic New Testament. As a result of significant manuscript finds during the twentieth century, new critical editions are needed for parts of the Sahidic New Testament, although recent editions of the Gospels, Acts, the Catholic Epistles, and the Apocalypse have moved the state of scholarship forward.
\nBohairic. Following the Arabicization of Egypt in the ninth and tenth centuries, the Bohairic dialect (bo) of Coptic continued as a liturgical language. As a result, the Bohairic New Testament has far more manuscript attestation than the other Coptic versions. (As in the case of Sahidic, however, there is no comprehensive list of manuscripts.) Unlike other dialects of Coptic, most Bohairic evidence dates from the eleventh century and after, although a number of manuscripts exist from as early as the fifth and sixth centuries. In part this is because Bohairic was initially the dialect of the Nile Delta, where the soil was too damp to preserve manuscript remains. The Bohairic tradition reflects direct translation from Greek, and it is likely that the later Bohairic tradition was further revised toward both Greek and Arabic. The only full edition of the Bohairic is that of Horner (1898\u20131905). Especially in light of twentieth-century manuscript discoveries, further editorial work is needed.
\nOther dialects. In addition to translations of the New Testament in Sahidic and Bohairic, parts of the New Testament are preserved in \u201cminor\u201d Coptic dialects, including Akhmimic (ac), Lycopolitan (ly), Middle Egyptian (mae), and Fayumic (fa). Manuscripts of these minor versions come from archaeological excavations of late ancient papyrus and parchment, and they date between the third and eighth centuries. Finds in the Akhmimic dialect include fragments of Matthew, Luke, John, Romans, Galatians, James, and Jude. This version probably reflects dialectal transposition from Sahidic rather than a fresh translation. The Lycopolitan dialect offers two manuscripts of John. While these resemble the translation of John in the closely related Sahidic dialect, the Lycopolitan does not seem to be a dialectal transposition. Middle Egyptian preserves significant portions of Acts and two different versions of Matthew. While it has been suggested that one or both translations of Matthew reflect different Greek source texts than are otherwise known, these arguments have not been widely persuasive. The Middle Egyptian version of Acts may be related to a text like that of Codex Bezae (05). The Fayumic exists only in relatively small fragments. Fayumic is closely related to its northern neighbor Bohairic, and some portions of the New Testament in Fayumic may reflect transposition from Bohairic rather than fresh translation. The Fayumic New Testament certainly experienced direct contact with Greek, however, as is attested in the Greek-Fayumic bilingual manuscript 086, the Coptic portion of which is known by the siglum \u201cfa 6.\u201d
\nCoptic translations and textual criticism of the Greek New Testament. Coptic translations play an important role in the textual criticism of the Greek New Testament, but caution is again required. As with Latin and Syriac, Coptic translations of the Greek New Testament are limited in reflecting their source text. Coptic has a verbal structure quite different from Greek, which makes one-to-one correspondence with Greek moods and tenses problematic. Coptic generally does not allow us to identify the Greek preposition + case construction used in the source text; for the same reasons, it is often difficult to use the Coptic text in support of a Greek verb with a prepositional prefix. Coptic, like Syriac, does not have grammatical case and relies on word order and prepositions to relate the components of a sentence. For example, the Greek witnesses to Matthew 4:3 attest conflicting locations for \u03b1\u1f50\u03c4\u1ff7 (aut\u014d, \u201cto him\u201d). Sahidic of Matthew 4:3 reads \u2c81\u03e5\u03ef\u2ca1\u2c89\u03e5\u2c9f\u2ca9\u2c9f\u2c89\u2c93 \u2c89\u2ca3\u2c9f\u03e5 (aftipefouoei erof, \u201che-approached to-him\u201d), placing the equivalent of Greek \u03b1\u1f50\u03c4\u1ff7 directly after the verb. This, however, does not support any particular location for \u03b1\u1f50\u03c4\u1ff7, since this is simply the standard grammatical location for this element of the sentence. These are only some of the many difficulties in reconstructing the Greek source text.
\nWhile Coptic is grammatically and syntactically different from Greek, Coptic dialects absorbed a significant number of Greek loanwords. (Syriac and Latin, too, have Greek loanwords, although a comparatively lower proportion of them.) Sometimes these carry the same nuances in Greek and Coptic (functioning as true loanwords); at other points they exhibit semantic drift or stretch. This means that often a Greek loanword in Coptic translation provides evidence for that word in Greek source text, but not always; sometimes a different, more familiar Greek loanword is used instead.
\nAs in Latin and Syriac, we are on firmest ground when citing Coptic translations for the presence or absence of large variation units (such as the pericope adulterae, Jn 7:53\u20138:11, which is absent from all Coptic translations except some late Bohairic manuscripts). In the case of smaller variations, the Coptic versions often simply do not provide a reliable basis. Careful studies of translation technique, however, continue to reveal particular patterns of correspondence between Greek and Coptic. More methodologically problematic for the use of Coptic translations in textual criticism of the Greek New Testament is a lack of adequate critical editions, especially in Bohairic.<\/p>\n
CONCLUSION: SCRIPTURE IN TRANSLATION<\/p>\n
In this chapter I have surveyed the main features of the New Testament as translated into Latin, Syriac, and Coptic. Although we do not have reliable total numbers of manuscripts in any of the early versions\u2014and the total number of manuscripts is probably not as large as it often claimed\u2014the early versions are nonetheless attested by a rich manuscript tradition that attests to the care and devotion of Christian communities who read the New Testament in translation.
\nThroughout the chapter I have suggested that early translations are not as useful as we might assume for reconstructing small-scale differences in their Greek base text. Further studies of translation technique may, of course, aid us in a more precise understanding of how these various translations reflect the Greek New Testament.
\nYet we must remember that we are working toward different ends from those communities that translated, revised, and transmitted these early translations of the New Testament. By the fourth century, the majority of Christians no longer encountered the New Testament in Greek but in an ever-growing number of other languages. Early translations were not created primarily to aid ancient biblical scholars in reconstructing a Greek text but to serve as Scripture. Translations are not simply witnesses for reconstructing the text of the Greek New Testament. They are witnesses to Christian theology and practice in a more robust way, as a window into how the early church studied, prayed, and worshiped.<\/p>\n
Key Takeaways
\n\u25ba In the second to fourth centuries, the New Testament was translated from Greek into other languages such as Latin, Syriac, and Coptic. Many other translations followed later.
\n\u25ba Manuscripts of the New Testament in these versions sometimes shed light on the Greek text used by the translators, but using the translation for textual criticism of the Greek New Testament is difficult and often too fraught with problems to be useful in the details.
\n\u25ba There are probably not ten thousand Latin manuscripts, and claiming twenty-five thousand New Testament manuscripts beyond the Greek is an exaggeration too big to keep using. It is better to say there are a few thousand versional manuscripts and leave it at that.
\n\u25ba The versions are not simply of value for reconstructing the Greek text; they are immensely valuable in their own right as the Scriptures for the communities that produced them. In some cases, communities continue to use them as their only Bible.<\/p>\n
CHAPTER FIFTEEN<\/p>\n
MYTHS ABOUT MODERN TRANSLATIONS<\/p>\n
VARIANTS, VERDICTS, AND VERSIONS<\/p>\n
Edgar Battad Ebojo<\/p>\n
I DID IT MY WAY: THE TRANSMISSION OF THE NEW TESTAMENT TEXT AND THE FACE OF MODERN SCRIPTURE TRANSLATION<\/p>\n
In its December 31, 2017, translation report, the United Bible Societies highlighted that the New Testament has now been translated into 2,189 languages worldwide. For evangelical Christians, this is something to praise God for, as this means that more and more people have been provided access to the transforming Word of God in their heart languages. The other side of the coin is getting the assurance that these New Testament copies are a faithful replication of the message of the original. But how can this be achieved given that there are now many competing Greek text editions readily available in the market? How should Bible translators choose from these? Or do Bible translators even use these Greek text editions in the first place? Ultimately, how should Bible users, especially Christians who have a very high view of the Scriptures, treat textual decisions undertaken by translators, which may or may not agree with their theological convictions?
\nMy goal here is to sketch a broad landscape of how New Testament translators worldwide perform their task, specifically when confronted with textual variations reflected in the critical apparatus of a given printed Greek text. Furthermore, since most of those who use the New Testament today read it from a particular translation, understanding the relationship that exists between textual criticism and Bible translation is a must. As such, part of my project here is to situate the theories and praxis of textual criticism in a more pragmatic scenario\u2014in conversation with Bible translation\u2014and see how textual criticism, as an interested discipline, is affecting (or not) those who are using the text that textual critics have reconstructed.
\nTextual criticism and Bible translation are not unrelated. In fact, Bible translation can become the most vital expression of textual criticism. For one, many of the documentary sources examined by textual critics are themselves translations in languages such as Coptic, Latin, Syriac, and so on, intended to faithfully transmit the message of their Greek source texts. Lest we become too dogmatic about a purist approach, it is worth remembering that many, if not most, of the early Christians read their New Testament in a language translation that they understood best. As such, the search for the \u201coriginal text\u201d (or the Ausgangstext), however elusive it might seem, will always include a thorough examination of these ancient translations. The construction of a list of New Testament variants, as represented in the critical apparatuses of New Testament Greek text editions or even in more exhaustive online databases, will always take into account a comparative analysis with these ancient translations. Hence, the value of the ancient versions and translations to the text-critical discipline is indispensable in this regard.
\nSecond, while there are now comparatively more evangelicals inclined to appreciate better the intricacies of textual criticism in the source language of the New Testament, most evangelicals still encounter text-critical issues via modern translations available to them. This hiatus needs to be bridged if textual critics are to make the discipline relevant to a broader spectrum of users. It is, therefore, incumbent on evangelical textual critics to take evangelicals to another level\u2014even if they use only modern translations. This is a clarion call for relevance. But this raises the question: Can we trust the work of modern translators?<\/p>\n
JUST DO IT: THEORIES AND PRAXES OF MODERN SCRIPTURE TRANSLATION<\/p>\n
There are essentially two types of translation approaches employed in New Testament translations projects worldwide, with varying degrees of theoretical and methodological affinities and dissimilarities: form-based and meaning-based translations. From these two approaches Bible translators choose to prepare their drafts, and their choice will determine greatly whether their drafts closely reflect the form or the meaning of their source Greek text.
\nThis, of course, assumes that Bible translators really use Greek text editions when translating. But do they always? This is a very important question that touches the very core of a translation team, and the answer lies in the sociocultural makeup of the community where the translation is being undertaken.
\nThe English-speaking world (as well as French, Spanish, and German-speaking) is blessed to have translators who are well equipped in the original languages of the Bible and the facility to understand and analyze them. This is not the case in most parts of the world, however. Many translation projects worldwide, especially in the Global South, employ as translators native speakers who do not have direct access to or the expertise in the original biblical languages. This may come as a surprise, but this is the reality on the ground. However, many evangelicals would concur that the absence of local \u201cBible experts\u201d must not deter the translation of the Scriptures into the heart languages of the people in need of encountering God. Scripture poverty should not be allowed to perpetuate because the community cannot offer qualified local personnel; there must be ways to address this. How do Bible publishers bridge this gap to ensure the faithful transmission of the New Testament text?
\nTo address this grassroots reality, many Bible translation teams employ the base-model approach to translation, wherein intermediate translations (usually from the languages of wider communication) are used to check and hopefully satisfy the requirements of form, structure, and meaning of the source text in the target language to produce translations with optimum intelligibility and acceptability. Furthermore, in these translation contexts the absence of local \u201cBible experts\u201d is addressed through the appointment of a translation consultant assigned to a project. The translation consultant, who is usually a nonnative speaker, is a professionally qualified member of the team who can give translators advice on how best to deal with translation issues and problems, including text-critical concerns.
\nA third safeguard in the translation process is the availability of translation tools. One of these tools is the Translator\u2019s Handbook series, a kind of translation commentary with special focus on exegesis and linguistic analyses of each book of the Bible, with various examples from around the world on how a particular passage might be translated best. This handbook series is also a good resource for discussion of text-critical problems from a decidedly translation perspective.
\nA handy companion to this handbook series is Roger Omanson\u2019s A Textual Guide to the Greek New Testament. If students (and experts) of textual criticism find help in Bruce Metzger\u2019s Textual Commentary, New Testament translation teams are in the same way informed of the intricacies of textual variations through Omanson\u2019s book. While Metzger\u2019s Textual Commentary is the basis of Omanson\u2019s Textual Guide, it dispenses with the technical jargon of the latter and is easier for readers whose primary language is not English.
\nThese three component practices have proven their worth in most New Testament translation projects worldwide and need not be elaborated here except to note that, even with these safeguards, translators still cannot completely escape questions of textual integrity. This reality is betrayed by the footnotes produced alongside the published texts.<\/p>\n
TAKE ME AS I AM: GREEK EDITIONS AND THEIR ROLE IN TRANSLATION<\/p>\n
Choosing the Greek textual base for translation projects. The presence or absence of a text-critical footnote in any modern translation of the New Testament is largely determined by the textual base that the translators and\/or the publisher decide to use up front. In many cases, publishers decide on the textual base to satisfy either their personal preferences or the expressed preferences of their target users.
\nThere are now more New Testament Greek editions available to students of New Testament and to Bible translators than ever before. The most recent is the Tyndale House edition, produced under the oversight of Dirk Jongkind and Peter J. Williams. Also fairly recent is Michael Holmes\u2019s SBL Greek New Testament, the apparatus of which shows comparative readings of four major editions of the Greek New Testament.
\nThe underlying Greek text for the New English Bible (and also REB) was R. V. G. Tasker\u2019s The Greek New Testament. The subsequent revisers of the 1978 NIV ostensibly used the text now found in A Reader\u2019s Greek New Testament, first published by Zondervan in 2003 and now on its third edition.
\nOn the other hand, those who prefer something in the KJV tradition use the editions of the textus receptus published and distributed most widely by the Trinitarian Bible Society. Other KJV adherents consult Robinson and Pierpont\u2019s The New Testament in the Original Greek as well as Hodges and Farstad\u2019s Majority Text edition.
\nBut while these Greek text editions are available and have been market-positioned based on the strengths their publishers feature, it is yet to be seen how translators in the field will respond and use them in their translation projects. The jury is still out.
\nGreek text editions of the United Bible Societies. The Greek edition most widely used as base for translating the New Testament is the shared text of the United Bible Societies\u2019 Greek New Testament (UBS) and the Nestle-Aland Novum Testamentum Graece (NA), now in their fifth and twenty-eighth editions, respectively. These two editions, both published by the German Bible Society on behalf of the UBS, are a product of an interconfessional editorial committee, essentially depicting the interconfessional ethos of the UBS.
\nThis shared text is extensively distributed in both its printed and electronic formats. Without reducing the serious scholarship behind its conception and production, the prominence this common text receives derives largely from the global presence of national Bible societies affiliated with the UBS in more than two hundred countries. Unless there is a strong clamor from the churches they serve, the default is always the UBS-NA text. The reason for this, unknown to many, is a document jointly agreed on in 1968 (revised in 1987) by the UBS and the Vatican\u2019s Secretariat for Promoting Christian Unity. This historic document is known in the Bible Society circle as the Guidelines for Interconfessional Cooperation in Translating the Bible; it states in 1.1.1.1, \u201cFor joint translation programs, teams should base their work on the critical edition of the Greek New Testament published by the United Bible Societies, which is itself a joint effort of scholars representing Roman Catholic and other Christian constituencies.\u201d
\nThe document not only specifies UBS\u2019s recommended textual base; it also intimates the sensitive challenges translators face in rendering textual judgment, especially when there is already a long tradition of Bible translation in a particular language community, as is the case for English, German, French, Spanish, and within the Orthodox circles. In such a context, Bible translation is deliberately conservative when it comes to the issue of a textual base. This is because the translator is both an intermediary between communities (i.e., his publisher-employer vis-\u00e0-vis his community) and a member of that speech community, expected to be an objective representative of the academic community but at the same time expected to represent also the interests of his faith community. Received tradition, therefore, plays a game-changing role in this context. This sensitive nature of translation is best expressed in this statement in the guidelines:<\/p>\n
Though a critical text [i.e., UBS] must form the basis of any adequate translation, it is recognized that in some situations certain constituencies may require that some passages of the New Testament found in the Byzantine tradition (as largely represented by the Textus Receptus) should be noted in the translation. When this is the case, such material may appear in footnotes with an appropriate marker in the text. The extent of textual adjustment will depend, of course, upon the local situation, and will need to be covered carefully by clear and detailed principles.<\/p>\n
This statement touches on one of the central goals of modern translation: acceptability. Bible translators are very sensitive to this aspect of the work.
\nThe guidelines further state, \u201cTranslators should normally follow this text for readings rated as A or B in The Greek New Testament but may choose other well attested readings when the text has a C rating.\u201d The letters A, B, C (and D) refer to the rated apparatus of UBS, which has been an integral feature of this Greek text edition since its conception, indicating the relative degree of (un-)certainty of the text adopted by the UBS Editorial Committee vis-\u00e0-vis the other existing variants. UBS is certainly not the originator of this device, but it fulfils certain functions that translators require, lest they get utterly lost in the forest of textual minutiae.
\nThis guideline recognizes the reality in the field that many communities do not have native speakers well equipped to handle the original languages of the Bible, much less to decide confidently on questions of textual integrity. Hence, a rated apparatus, with some prescriptive instructions, comes in handy. The presence of translation consultants is then reinforced by the rated apparatus in assisting native translators to make informed decisions when confronted with textual questions.
\nAlthough there has been intense critique of this system, its value for Bible translators in the field cannot be underestimated. Translators often do not have direct access to textual scholarship in the same way that textual critics (mostly in North America and Europe) do. Despite this, Bible translators are committed to ensuring the accurate and faithful but also natural and meaningful transmission of the biblical text, based on the mutually agreed-on textual base. Admittedly, this decision is deliberate and does not always depend on the best methodical criteria for the choice of reading that textual critics might use.
\nExcluding the included: when textual tradition meets church tradition. Note must also be made of the single and double brackets in the UBS-NA and other editions, which present certain translation challenges too. Strictly speaking, brackets are part of the text rather than the apparatus, and therefore the enclosed text becomes part of the domain of the translator as well. Bracketing, in the context of New Testament textual criticism, is a necessary editorial device used when (1) it proves difficult to render a clear decision on the question of \u201coriginal reading\u201d and\/or (2) an editorial committee cannot reach a satisfactory consensus.
\nUBS and NA explain that single-bracketed entries \u201cmay be regarded as part of the text \u2026 but cannot be taken as completely certain.\u201d Double-bracketed passages, on the other hand, are \u201cknown not to be part of the original text\u201d but are included in the text because of their importance to the history of the church. Double-bracketed passages present special challenges for translators. There are five double-bracketed passages in the UBS-NA text (Mt 16:9\u201320, 21, aka the shorter ending of Mark; Lk 22:43\u201344; 23:34a; Jn 7:53\u20138:11), but only two are usually marked in most translations.
\nAlmost all modern translations include John 7:53\u20138:11 in the main text, and it is printed immediately after John 7:52, with corresponding brackets and a marginal note. This shows the tenacity of this tradition in the global translation history. It is recognized by subsequent and contemporary church tradition as not only inherently part of the New Testament but also as part of the Johannine Gospel. This has nothing to do directly with the decision of the UBS-NA Editorial Committee but relates to the preference of the churches that have historically used translations\/versions (or even manuscripts!) that included the pericope adulterae at this point of the Gospel.
\nHere is an observation worth noting. No one seems to have seriously noted this fact of praxis in translation vis-\u00e0-vis the trend of research in textual criticism. If including this passage after John 7:52 has been part of the translation tradition for as long as one can remember, what pragmatic value is there to translators and the communities they represent in the note about other locations it has been placed in the New Testament (i.e., after Jn 7:36 or Jn 21:25, or after Lk 21:38 or Lk 24:53)?
\nThe conclusion of the Gospel of Mark presents another interesting translation case. While textual critics continue to present arguments in favor of inclusion or exclusion of passages after Mark 16:8, translators are interested in another matter\u2014which ending to print in the main text. While textual critics continue to produce interesting treatises on the original ending of this Gospel, translators (and the faith communities they represent) seem to have decided long ago that the Gospel of Mark must not end with Mark 16:8! Their question is which of the six narrative endings to conclude with, and the answer might just be found within the four corners of their churches.<\/p>\n
TO HAVE OR NOT TO HAVE FOOTNOTES? ALTERNATIVE RENDERINGS AND ALTERNATIVE READINGS<\/p>\n
Types of footnotes. There are aesthetic features in modern translations intended to appeal to Bible buyers, such as beautiful and colorful covers with titles in elegant and catchy type, with socially segmented product lines depending on the target audience; introductory materials to a given edition highlighting its marketing edge over the others; informative book introductions; columniation with varying degrees of sophistication; attractive section headings; impressively informative footnotes that may even threaten to upstage the main text; and others.
\nWhat many people do not realize is that several of these features have been part of the transmitted manuscripts. These include page numbers, which we see distinctly from codex manuscripts, written on the top middle area of a page (e.g., P46, 01, 02, 03, etc.); book titles rendered in calligraphic and exquisite rubrications at every beginning (and sometimes also ending) of a book; capacious columniation; chapter and verse divisions, which are not features of our earliest extant manuscripts, though their concepts are represented in some ways other than numbers; a cross-referencing system, which may have been the goal of Eusebius\u2019s Canons, at least for the books of the Gospels; and many others. The salient point to be underscored here is that these paratextual features were meant primarily to facilitate ease of reading of the Scriptures in a liturgical context.
\nOf these modern aesthetic features, footnotes are perhaps the best source for giving text-critical information because many footnotes provide an immediate way to increase the reader\u2019s appreciation for the earliest recoverable text. For our present purposes, it is enough to discuss only the textual footnotes, of which there are two types, generally distinguished by translators as \u201calternative rendering\u201d and \u201calternative reading\u201d footnotes.
\nAlternative rendering footnotes offer another way of interpreting a word or group of words because of inherent ambiguities in the source text or (seldom) because there is another way of expressing this in the receptor language. The manuscript tradition shows no (significant?) variation, but the text, as it is, can be translated in more than one possible way, which may or may not open up a whole new gamut of exegetical possibilities. Footnotes like this abound in the translations and are usually identifiable with the marker \u201cor\u201d (an abbreviation for the phrase \u201cor can be translated as \u2026\u201d).
\nOne example is sufficient for the moment: the phrase \u1f41 \u03b4\u1f72 \u03b4\u03af\u03ba\u03b1\u03b9\u03bf\u03c2 \u1f10\u03ba \u03c0\u03af\u03c3\u03c4\u03b5\u03c9\u03c2 \u03b6\u03ae\u03c3\u03b5\u03c4\u03b1\u03b9 (ho de dikaios ek piste\u014ds z\u0113setai) in Romans 1:17. Many evangelicals memorize this passage by heart by way of the KJV\u2019s \u201cThe just shall live by faith\u201d (also NKJV). In effect, this is also the rendition of many early twentieth-century translations, such as the ASV (except that \u03b4\u03af\u03ba\u03b1\u03b9\u03bf\u03c2 [dikaios] is translated as \u201crighteous\u201d rather than \u201cjust\u201d [Goodspeed and JB render it as \u201cupright\u201d]). Conversely, translators of the 1952 RSV rendered this, \u201cHe who through faith is righteous shall live,\u201d with a footnote, \u201cOr The righteous shall live by faith.\u201d
\nThis seems to be the first break from the KJV translation tradition, showing that this Habakkuk 2:4 quotation (also quoted in Gal 3:11; Heb 10:38) is open to two translation possibilities. The RSV\u2019s strategy (i.e., main text cum the alternative rendering note) is reflected by the TEV, CEV, and NABre, but many still reflect the KJV rendition plus a footnote of the alternative (CEB, CSB, ESV, HCSB, NASB, etc.), including the NRSV! Interestingly, some translations adopted one of the renditions without any footnote at all (NIV, NKJV).
\nIn instances such as this, the transmission of the text is not at issue, but the ability of the translators to present the broad spectrum of proposed interpretations across history is nonetheless put to the limits. Ultimately, translators still have to make a choice, which obviously will support one exegetical position and relegate the other as an alternative\u2014not an enviable position to be in.
\nThe second type of textual footnote is the alternative reading. Unlike the alternative rendering, the manuscript tradition itself is divided in this instance, offering two or more sensible and meaningful variants. And a choice of one over the other(s) may tip the balance of interpretation toward a particular line. The textual tradition, therefore, is at issue here. It is this kind of variation that characterizes many of the textual footnotes in our modern translations, as profiled in the critical apparatuses of many Greek text editions.
\nFootnotes as indicators of textual integrity. Not all variations noted in the Greek base text, by virtue of the nature of language and\/or the requirement of the receptor culture, need to be footnoted. Many are simply irrelevant to translation. For example, a variation unit may be linguistically relevant in the source language (or in the model text language such as English, Spanish, or French) but not relevant at all to the receptor language. Variation units that are orthographic in nature or those that have to do with word order may not always be relevant in some translation contexts, especially if they do not affect the meaning of the passage to be translated.
\nFurthermore, and perhaps more importantly in the context of Bible translation, the end-users\u2019 sociocultural context also plays a critical factor in the choice of which variation unit to include and reject. As such, a textual critic, who is an \u201coutsider\u201d in a translation scenario, should not expect to see all the variation units in, say, UBS reflected in a local translation; selectivity in choices will always exist. In fact, some communities in Asia-Pacific and Africa, for example, prefer their translations of the New Testament without footnotes at all, as the presence of a footnote is perceived to reduce the \u201csacredness\u201d of the translation; they believe that footnotes may encourage \u201cuncertainty\u201d about the firmness and immutability of God\u2019s Word. Of course, while this scenario may be theologically desirable, the history of New Testament transmission paints a more complex picture than this.
\nIntelligibility of the footnotes. But even footnotes that have long been present in the translation tradition are not always intelligible in the way they are structured in many English translations, which in turn are being carried over to other translations that used them as model texts. Let me use the disputed portion of Luke 23:34 to illustrate this point. \u201cAnd Jesus said, \u2018Father, forgive them because they do not know what they are doing\u2019 \u201d (\u1f41 \u03b4\u1f72 \u1f38\u03b7\u03c3\u03bf\u1fe6\u03c2 \u1f14\u03bb\u03b5\u03b3\u03b5\u03bd \u03c0\u03ac\u03c4\u03b5\u03c1, \u1f04\u03c6\u03b5\u03c2 \u03b1\u1f50\u03c4\u03bf\u1fd6\u03c2, \u03bf\u1f50 \u03b3\u1f70\u03c1 \u03bf\u1f34\u03b4\u03b1\u03c3\u03b9\u03bd \u03c4\u03af \u03c0\u03bf\u03b9\u03bf\u1fe6\u03c3\u03b9\u03bd), one of the so-called Seven Last Words of Jesus, is double-bracketed in UBS, and this decision is rated A.
\nMost English translations print the disputed text of Luke 23:34 in the main text, enclosed in single brackets (double brackets for ESV), plus a footnote calling into question its authenticity. One\u2019s curiosity is aroused, however, once footnotes are set side by side. The ASV notes, \u201cSome ancient authorities omit the sentence \u2026,\u201d while the RSV has \u201cOther ancient authorities omit.\u2026\u201d A beginning Bible reader will be at a loss as to what \u201cancient authorities\u201d means, whether it refers to a person in antiquity or people in the government. The use of the phrase \u201cancient authorities\u201d is clearly acceptable in text-critical circles, referring to the manuscript evidence, but it can be confusing to many in the translation context.
\nIt might seem that the term manuscripts is a better choice than authorities. Maybe, but not always. In contexts where the New Testament translation project is the first written document ever in that community (i.e., a highly oral culture), the concept of \u201cmanuscript\u201d as a written document or artifact will prove to be cognitively challenging. In some translation projects in Vietnam and in the Philippines, translators need a descriptive phrase in reference to manuscripts (that is, \u201cold books with old writings\u201d) to make it meaningful.<\/p>\n
ALL THE WRONG PLACES TO BE: SOME PARATEXTUAL FEATURES IN MANUSCRIPTS THAT HAVE AFFECTED THE FACE OF SCRIPTURE TRANSLATION<\/p>\n
There are variation units in Greek editions that have grown out of the paratextual domains of the manuscript tradition. These include indentation, paragraph segmentation, orthography, punctuation, itacism, nomina sacra, accents, and others. Some of these features can have implications for translations.
\nNomina sacra. Mark 1:1, in the manuscript tradition, attests the readings \u201cThe beginning of the good news of Jesus Christ, the Son of God\u201d and \u201cThe beginning of the good news of Jesus Christ.\u201d Here the variation was most likely caused by the series of abbreviations for names associated with the divine, or the so-called nomina sacra. The omission or the addition of the phrase \u201cSon of God\u201d could have been influenced by the optical sequence of nomina sacra in this sentence, involving six (or four) final upsilons in a series, all written in capital (majuscule) scripts, without any spaces in between words (scriptio continua), that is:<\/p>\n
\u0391\u03a1\u03a7\u0397\u03a4\u039f\u03a5\u0395\u03a5\u0391\u0393\u0393\u0395\u039b\u0399\u039f\u03a5\u0399\u0305\u03a5\u0305\u03a7\u0305\u03a5\u0305\u03a5\u0305\u03a5\u0305\u0398\u0305\u03a5\u0305<\/p>\n
Almost all the modern translations render the first reading, either putting the appositional phrase \u201cthe Son of God\u201d in single brackets (NABre) or simply marking it with a footnote about the other variant. Of interest is the rendition of the Lexham English Bible (LEB), which reflects the shortest reading: \u201cThe beginning of the gospel of Jesus Christ.\u201d
\nItacism. Variations in the source text caused by itacism (a linguistic phenomenon in Greek wherein two different letters had been pronounced similarly) may also affect translation, at least from an exegetical point of view. Romans 5:1 is a good case in point.
\nManuscript evidence is divided between the subjunctive \u1f14\u03c7\u03c9\u03bc\u03b5\u03bd (ech\u014dmen, \u201clet us have peace [with God]\u201d) and the indicative \u1f14\u03c7\u03bf\u03bc\u03b5\u03bd (echomen, \u201cwe have peace [with God]\u201d); most Greek text editions (Bengel, Tischendorf, UBS\/NA, SBL Greek New Testament, Robinson-Pierpont, Hodges-Farstad, etc.) reflect the indicative. The hermeneutical question is whether Christians have peace with God already on account of Christ\u2019s vicarious sacrifice on the cross or whether Christians must still actively do something to experience peace with God.
\nInsofar as English translations are concerned, most reflect the indicative reading (RSV, NRSV, NABre, TEV, CEV, NLT, etc.) over the subjunctive (RV, Philips). There is some variety in the translation tradition, however. Some simply print one reading as if there is no variation in the manuscript tradition. Conversely, NIV agrees with the majority translation, but NIV 1984\u2019s footnote (\u201cOr let us\u201d) is baffling, making it appear as though this is a simple case of alternative rendering and not a text-critical issue. HCSB\u2019s reading is even more surprising\u2014it has in the main text, \u201cTherefore, since we have been declared righteous by faith, we have peace with God,\u201d but the footnote interestingly says, \u201cOther mss read faith, let us have peace, which can also be translated faith, let us grasp the fact that we have peace,\u201d which conflates the two readings, producing a reading witnessed by no manuscript.
\nAccents. Full employment of accents in the manuscript tradition was a later development, although they are used sparingly in some of the earlier extant manuscripts. They were intended primarily as optical aids to Scripture readers. But some text-critical variations involving accents may also affect the translation of the New Testament, and some may even stir heated hermeneutical discussions on the gender question. Romans 16:7 is one such case, dealing with the gender of a person identified in the manuscript tradition as \u1f38\u03bf\u03c5\u03bd\u03b9\u03b1\u03bd (Iounian).
\nThe accusative \u1f38\u03bf\u03c5\u03bd\u03b9\u03b1\u03bd can be translated either as feminine (\u201cJunia\u201d) or masculine (\u201cJunias\u201d), and the choice is largely dependent on how this rare Greek name is accented. The manuscript tradition is divided between the acute accent (\u1f38\u03bf\u03c5\u03bd\u03af\u03b1\u03bd, Iounian, \u201cJunia\u201d), the unaccented (which can then be read as \u1f38\u03bf\u03c5\u03bd\u03b9\u1fb6\u03bd [Iounian, \u201cJunias\u201d; reflected by Tasker]), and the variant \u1f38\u03bf\u03c5\u03bb\u03af\u03b1\u03bd (Ioulian, \u201cJulia,\u201d supported primarily by P46; see Rom 16:15). Most Greek text editions prefer the first of these (Bengel, Westcott and Hort, UBS-NA, Holmes, Hodges-Farstad, Robinson-Pierpont, etc.).
\nThe current state of translation is equally interesting. The KJV reflects the feminine \u201cJunia.\u201d Interestingly, the RV rendered the masculine \u201cJunias\u201d even though Westcott and Hort reflected the feminine form \u1f38\u03bf\u03c5\u03bd\u03af\u03b1\u03bd. The ASV, RSV, NIV, JB, the Spanish Dios Habla Hoy, the German Lutherbibel, the French La Nouvelle Bible Segond, and the Traduction Oecumenique de la Bible also reflect \u201cJunias.\u201d On the other hand, GNT, NKJV, NRSV, CEV, NLT, ESV, and others reflect the feminine \u201cJunia.\u201d
\nThe footnote in the NLT is a bit misleading, as it states, \u201cJunia is a feminine name. Some late manuscripts accent the word so it reads Junias, a masculine name.\u201d This note suggests that the feminine reading results from the absence of accent, which is contrary to the manuscript evidence. The NET Bible, while supporting a feminine reading, has a long note containing an intriguing comment: \u201cIn Greek only a difference of accent distinguishes between Junias (male) and Junia (female). If it refers to a woman, it is possible (1) that she had the gift of apostleship (not the office).\u201d This type of note makes one wonder whether apostleship in the New Testament time was a matter of gifting when apostleship is associated with a woman, but an office when a masculine figure is in view.
\nThere is a growing voice that Iounian here is feminine (possibly the wife of Andronicus, who had made herself equally notable among the apostles, i.e., more notable than others in the apostolic band). This point is of fundamental importance because if a feminine reading of the name is right, and the subsequent clause, \u03bf\u1f35\u03c4\u03b9\u03bd\u03ad\u03c2 \u03b5\u1f30\u03c3\u03b9\u03bd \u1f10\u03c0\u03af\u03c3\u03b7\u03bc\u03bf\u03b9 \u1f10\u03bd \u03c4\u03bf\u1fd6\u03c2 \u1f00\u03c0\u03bf\u03c3\u03c4\u03cc\u03bb\u03bf\u03b9\u03c2 (hoitines eisin epis\u0113moi en tois apostolois), is translated as \u201cwho are prominent among the apostles,\u201d then Junia is the first woman apostle unequivocally mentioned in the New Testament\u2014a thought that thrusts this text-critical issue to the center of gender studies.<\/p>\n
VARIANTS, VERDICTS, AND VERSIONS: THE CONTINUING EVOLUTION OF THE NEW TESTAMENT TEXTS IN THE CONTEXT OF BIBLE TRANSLATION<\/p>\n
Bible translation is a viable avenue for conveying text-critical discoveries, albeit in a limited and deliberately selective way. As such, we must not expect modern translations to be the ultimate venue to exhaustively deal with text-critical concerns, the way textual critics do. The variables at play within the translation context are equally as dynamic as recovering the \u201coriginal text\u201d from ancient artifacts. In the context of evangelicalism, translation is fundamentally perceived as communicating clearly the biblical message and therefore must be judged along that primary intent.
\nThis challenges everyone to look at translation beyond the rigidities of the text but also at the idiosyncrasies of its transmission context, to take into account the various factors at work all at the same time. What modern translators can offer is the motivation and commitment to faithfully and meaningfully translate the sacred Scriptures of the early church for the use of modern churches, for their own nurture, growth, and maturity. Translation settings for this commitment are not always ideal, but all safety nets, within the capabilities and limited resources of Bible-translating agencies and the target communities they serve, are put in place to arrive at this goal. Yet at the end of the day, translators, especially those concerned with clarity and meaningfulness in the receptor language, will always be found to have \u201caltered\u201d in some ways the form of the source text, and sometimes even the perceived meaning of the source text, in the same way that scribes of old have intimated their preferences, theological or otherwise, through the various text forms they construed to be the Word of God for them.
\nIt will not escape our notice that translations do not always agree\u2014translational diversities abound. It is sobering to note, however, that this diversity of translations reflects also the rich diversities that existed in the early church. These diversities exist because translators do not work in a theological and sociocultural vacuum; their target audience is always in front of them when they prepare their drafts. In such contexts, new \u201creadings\u201d will indubitably emerge, conditioned by the linguistic and sociocultural requirements of the target communities for whom the translations are being prepared. Some of these readings may surprisingly be found to have the support of some least-known manuscripts of old; but perhaps more will not enjoy any manuscript support at all.
\nWe use various descriptive categories (e.g., omission, addition, substitution, harmonization) to describe the scribal alterations in the manuscript tradition. We may note that many of these alterations are not directionless, and not necessarily simply intended to distort God\u2019s Word either. It is more likely that these were deliberately and selectively undertaken because it made sense to the scribes and to the communities they represented. As such, if the transmission history witnesses to these deliberate alterations, it stands to reason that translators in the field also become deliberately selective in their choice of which reading to reflect as their main text and which reading to relegate as a marginal (alternative) reading in their printed Bible.
\nAccordingly, not all variants printed in the critical apparatuses of Greek text editions have any significant bearing for translations. Selectivity in the choices between variant readings and how alternative readings are represented is present in any translation, and the degree to which they differ is largely dictated by their target audience. While some variants have potential for their theological resonance, in many cases they are not significant for translation.
\nThis raises a whole new set of questions for editions of the Greek New Testament intended for translators in the field concerning whether variants that do not make any difference in actual translation work at all truly warrant inclusion in the critical apparatus. Many of these are still in the variation units of UBS. But the point is not only for UBS editions. This will continue to challenge any editor or editorial committee of any Greek New Testament, especially if they want to market their edited text as one prepared for Bible translators in the field.
\nTextual critics desiring to prepare a Greek text edition for Bible translators would do well to intently listen to Bible translators in the field, especially those who do not have direct access to the original biblical languages, to patiently probe and know from them how they are transmitting the \u201coriginal text\u201d through their translation drafts. The truth of the matter is that most modern translators outside the confines of North America and Europe work on their translation drafts daily, knowing that they are not accessing the \u201coriginal text\u201d but consciously working from model translations of scientifically reconstructed editions of the New Testament text. This is a humble recognition of the hard and painstaking labors and contributions of committed textual critics. But this similarly underscores the question of how the concept of the \u201coriginal text\u201d is to be appreciated afresh in changing, diverse contexts, in the same way that what we now call variations in the transmission history first emerged at one point in the history of the early Christians who wanted to make their received Scriptures relevant to their own contexts, without severing their connections from their spiritual forebears who passed on to them the Word of God. Indeed, verdicts on variants will certainly witness the continuing evolution of the New Testament texts through the versions of God\u2019s Word used today (and in the future) in different churches worldwide.<\/p>\n
MYTHS
\nAND
\nMISTAKES
\nIN
\nNEW TESTAMENT
\nTEXTUAL CRITICISM<\/p>\n
Myths and Mistakes in New Testament Textual Criti<\/p>\n","protected":false},"excerpt":{"rendered":"
FOREWORD Daniel B. Wallace We no longer live in a black-and-white world. We never did really, but those who are embroiled in debates about the Bible have often viewed things in such binary hues. These achromatic ideologies can be found on both sides of the theological aisle. Many who have abandoned the unreflective beliefs they … <\/p>\n