Why Does New Testament Textual Criticism Matter? Refined Definitions and Fresh Directions

A. Numerous Manuscripts and Myriad Textual Variations in the New Testament Text

Complexity is one source of frustration that has led many students and scholars to keep textual criticism safely at arm’s length. Already 300 years ago, expressions of exasperation, motivated by fear for the faith, flowed into print after John Mill published his massive critical edition of the Greek New Testament in 1707. ¹ This edition, the first of its kind and the model for all editions to follow, consisted of a massive introduction and an extensive critical apparatus at the foot of its folio pages. It was a model, however, only in its format and not in the quality of its Greek text, for the running text at the top of the pages merely reproduced the abiding ecclesiastical text of the Church–the textus receptus–or the ‘received text’ that had solidified over some dozen preceding centuries. Though Mill died two weeks after publishing his work, some contemporaries heaped shame upon him for exposing to the Christian public some 30,000 variants among the New Testament manuscripts known at the time. For example, Daniel Whitby issued a folio volume of annotations to the New Testament in 1710, whose Preface and 110-page Appendix were in Latin, offering an examination of Mill’s variant readings, where he stated:

I GRIEVE therefore and am vexed that I have found so much in Mill’s Prolegomena which seems quite plainly to render the standard of faith insecure, or at best to give others too good a handle for doubting.

He added that, in comparison with the then-standard 1550 Greek text of Stephanus, ‘the variations [were] quadrupled by Mill after sweating for thirty years at the work’. And why was Whitby’s book written in Latin? That was the academic language of the time, and he wanted all to know that his book was meant for the learned, but he also wished to minimise advertising Mill’s perilous views to avoid spreading further doubt. ² Soon Mill’s edition, however, was well advertised by Anthony Collins, a Free-Thinker, who, in 1713, very publically referred his readers to Whitby’s Critique of Mill and then mentioned Mill’s 30,000 variant readings to demonstrate that the whole text of the Holy Scriptures had been rendered precarious and insecure. ³ Mill would have rejected both critiques, for the career-long labor of love sought nothing more or less than a better understanding of the process by which the Greek New Testament had been transmitted down to his own times.

Enter Richard Bentley in 1720, a preeminent classical scholar and Master of Trinity College at Cambridge, who defended Mill in words eloquently exonerating not only him, but also humanistic scholarship more broadly. Note Bentley’s often quoted assertion:

For surely those Various Readings existed before in the several Exemplars; Dr. Mill did not make and coin them, he only exhibited them to our View… .

The 30,000 Various Lections are allow’d then and confess’d: and if more Copies yet are collected, the Sum will still mount higher… .

It is good therefore … to have more Anchors than One; and another MS. to join with the first would give more Authority, as well as Security. Now chuse that Second where you will [and] there shall be a Thousand Variations from the First; and yet Half or More of the Faults shall still remain in both… [but with] some Copy preserving the True Reading in one place, and some in another. And yet the more Copies you call to assistance, the more do the Various Readings multiply upon you… . And this is fact, not only in the New Testament, but in all Antient Books whatever.

[And Bentley concluded:] ‘Tis a good Providence and a great Blessing, That so many Manuscripts of the New Testament are still among us; some procur’d from Egypt, others from Asia, others found in the Western Churches… .

Make your 30,000 as many more, … all the better to a knowing and serious Reader, who is thereby more richly furnish’d to select what he sees Genuine. ⁴

Bentley, of course, vastly underestimated the multiples of manuscripts and variants that would evolve, reaching some 300,000 variations by 1881 when Westcott and Hort published their landmark edition. ⁵ Today we have more than 5,500 different Greek manuscripts of the New Testament writings, and twice as many versional copies in all the languages of Christianity through Late Antiquity. ⁶ There is, however, no reliable estimate of the total number of variants found in our extant witnesses, though a wild guess would be somewhere between two fifths to perhaps three quarters of a million. Like Bentley, most of us view this profusion positively and call it an embarrassment of riches when compared with most other ancient writings. We like to say also that, comparatively, we have perhaps the highest proportion of early witnesses. Kurt Aland affirmed, as have many others, that conjectural emendation is unnecessary and that the original text assuredly can be distilled from this abundance of Greek and other witnesses. ⁷ Some of us, however, are not that sure about the easy distillation of the original text, an issue discussed below.

The staggering quantities of witnesses and their variations do add exponentially to textual criticism’s complexity, though, in actuality, textual critics probably have to deal with only a tiny fraction–about thirty thousand variants. Moreover, we treat variation-units, that is, places of variation in the text where each unit contains two to a half-dozen or more different readings. This permits a more realistic figure for the actual workload: The United Bible Societies’ Greek New Testament contains 2,040 variation-units. A sampling suggests that the Nestle-Aland Greek New Testament, with a more ample apparatus, treats about six times as many variation-units or some 12,000. This calculation could be far off the mark, but at least it shows that the problem is not nearly as large as first appears, though it also guarantees that more than enough variants are available to occupy textual critics for decades.

The sheer mass of material may have discouraged some from joining the struggle to make sense out of it all, but textual criticism from the Renaissance through modern times has developed methods to tame this ‘jungle’ of data, primarily through continued refinement of its methods, including the formulation of increasingly sophisticated quantitative measures of manuscript relationships. Since textual criticism matters, it is essential that these procedures are understood.

B. Assessing Method: The Criteria for the Priority of Readings

The long tradition that has brought us, virtually in a straight line, to the discipline as practised recently reaches back to John Mill and Richard Bentley and to their successors, J. A. Bengel, J. J. Wettstein, J. J. Griesbach, K. Lachmann, C. Tischendorf, S. P. Tregelles, F. H. A. Scrivener, B. F. Westcott and F. J. A. Hort, C. R. Gregory, F. G. Kenyon, K. Lake, M.-J. Lagrange, L. Vaganay, Eberhard and Erwin Nestle, and comes down to those more contemporary, Kurt and Barbara Aland, Ernest Colwell, Bruce Metzger, and most others that may come to mind. And what did these prominent figures have in common? First, they actively formulated principles or criteria for sorting through the readings in each variation-unit and isolating, as they would say, ‘the original reading.’ Second, they shared a common, singular goal, establishing the original text of the entire New Testament. Consider, first, the criteria employed then and now–with refinement–in assessing the variants in each unit of text.

Nothing in scholarship is simple, which applies ‘with a vengeance’ to the basic method in textual criticism, namely the embrangled criteria for the priority of readings. To be sure, these guidelines are largely subjective and often inconclusive, but they do constitute the methodological core of textual criticism, and might as well be faced head-on. Moreover, they are venerable, having evolved over the past three hundred years, though with intimations already as early as Origen (third century) and some carefully formulated items as early as Erasmus (1516). Since that time, most critical editions of the Greek New Testament have employed, implicitly or explicitly, some dozen to sixteen recognised criteria. Their history and evolution, the rationale for their validity, and the qualifications and cautions essential in their use cannot be explored extensively here, ⁸ but any such account would leave an impression of high complexity.

In recent years I have utilized a double term to describe these guidelines, namely, ‘Criteria/Probabilities’, which emphasises that they are not ‘canons’ or ‘rules’–something fixed that can be followed mechanically–but criteria in a proper sense. A criterion is a standard or test by which something can be judged, a measure of value. Hence, in textual criticism, as one seeks to discover the variant reading that is prior to all other variants in its variation-unit, the closest synonym for criterion would be ‘probability’. That is, what is the ‘measure’ or ‘estimate’–what is the probability–that a given variant was prior to the others? Answering this basic question becomes a spiral process, which involves testing each variant in a variation-unit with several criteria. Most textual critics agree that the leading, overall criterion is this: Can one variant in a variation-unit explain the rise of all others? If so, that variant has a high probability of being the earliest or prior reading and thereby represents (to use the preferable expression) the earliest attainable text ⁹ in that unit.

A second criterion that often can be decisive is known as the ‘harder’ reading criterion, and it suggests that the more difficult, rougher reading in a unit is more probably the prior reading (unless it is nonsense) because scribes and readers of manuscripts tend to smooth out rough grammar and to improve unclear expressions, rather than to create a problematic text.

Matthew 27:16-17 provides an example of these two criteria, a passage I have used often. Here the imprisoned insurrectionist has two varying names in the textual tradition: ‘Barabbas’ and ‘Jesus Barabbas’, and, perhaps counterintuitively, ‘Jesus Barabbas’ is the variant that explains the other and is thereby the earlier reading, for scribes would be strongly inclined to separate the sacred name, ‘Jesus’, from that of a criminal, but would have no inclination to add it to ‘Barabbas’ if the latter were the earlier reading. In this case, ‘Jesus Barabbas’ is also the ‘harder reading,’ that is, disturbing to Christian scribes in its linking of Jesus’ name to that of an evil person, thus motivating such scribes to remove it. Hence, based on these two criteria, the insurrectionist’s name was, indeed, ‘Jesus Barabbas’ (see further below). ¹⁰

These two criteria, out of nearly a dozen categorised as ‘Internal Evidence’, assess what authors were most likely to write and what scribes and readers were likely to transcribe when manuscripts were being copied or used. Among the remaining internal criteria, the probable prior reading in a variation-unit would be the one that conforms to the author’s recognisable vocabulary and style; or to the author’s recognised theology; or to Semitic forms of expression (since the New Testament authors, being either Jewish or familiar with the Septuagint/Greek Old Testament, likely would reflect Semitic style).

In 1 Cor. 13:3, the choice between ‘… that I may boast’ and ‘… that I should be burned’ cannot depend upon one criterion, but, in support of the former reading, the verb ‘to boast’ occurs thirty-four times and is a frequent topic in Paul’s genuine letters, but ‘to burn’ occurs only in this textual variant. ¹¹ As to the argument from an author’s theology, in John 1:34 the reading ‘the Son of God’ is in line with the Fourth Gospel’s theological terminology, over against ‘the chosen one of God’. Luke 2:14 illustrates how a Semitic expression comparable to ‘peace among people of [God’s] good pleasure’, increases the probability of that variant as prior to ‘peace, good will among people’.

Continuing the list of internal criteria, the prior variant probably would be the reading that does not conform to parallel passages found in the Gospels or other Christian texts, or in the Old Testament, or in routine liturgical passages. Why? Because harmonisation of the text that a scribe was copying with similar, familiar texts became a common phenomenon–occurring intentionally if a scribe felt something had been omitted, or almost unconsciously when commonly used formulations came to mind.

Most common, for obvious reasons, is the harmonisation of one Gospel text to that of another Gospel. In the Mark 3:5 account of Jesus healing on the Sabbath, Jesus is said to have looked at the Pharisees ‘with anger’ (no textual variants there), though in the Lucan parallel (Luke 6:10) many witnesses attest to either the absence or presence of this phrase. This would suggest that scribes have incorporated into Luke the unusual reference to Jesus’ anger, but harmonisation can go either way, so additional criteria must be considered. For example, several instances of Jesus’ anger appear in Mark, but not in Luke (or Matthew). ¹²

In Matt 2:31, choosing the prior reading between ‘Lord, have mercy upon us’ and ‘Have mercy upon us, Lord’ is settled in favor of the latter since it does not follow the usual liturgical order of words: scribes consciously or unconsciously would alter it to fit the normal oft-recited order.

Perhaps more obvious are several criteria that represent ‘External Evidence’, which assess supporting manuscripts with respect to their age, quality, geographical distribution, and groupings. Most prominent in the search for the earliest attainable text would be a variant’s support by the earliest manuscripts, patristic citations, or versions (or by later witnesses assuredly preserving early texts).

Examples are legion, as evident from perusing the pages of the Commentary on the United Bible Societies’ Greek New Testament and discovering the frequency of phrases such as the ‘overwhelming external attestation’, ‘superior external evidence’, or ‘early and widely diversified witnesses’, which describe the basis of the editors’ decisions. For instance, according to John 1:28, was John (the Baptist) baptising in Bethany or Bethabara? Though the latter has some twenty witnesses, the earlier and much stronger attestation supports Bethany.

Or, probability that a reading has priority in its unit would increase if that variant is supported by the ‘best’ witnesses (based on repeated cases of a manuscript or other witnesses possessing prior readings, which, by the way, results in significant overlap with the earliest witnesses), or has wide geographical support, or has the support of one or more groups of witnesses known to be early or of recognised ‘best quality’.

In Mark 1:41, one variant has Jesus compassionate toward a leper, the other asserts that Jesus was angry. The support for the latter is highly restricted, and most, therefore, would describe the attestation for ‘compassionate’ as early, ‘best’, and diverse.

It is essential to note that in virtually all the examples above–as in the case of most text-critical decisions–additional criteria come into play as the variations are evaluated over against one another. In the last example, Jesus as ‘angry’, though not well attested, likely would be the harder reading, so that other factors would be considered to assess probability. Compare our first example, ‘Jesus Barabbas’, which by far also has the weaker attestation, while ‘Barabbas’ is supported by the earliest and ‘best’ witnesses, and by those most widely distributed geographically and in recognised groups. Normally that would accredit ‘Barabbas’ as the name originally written. Why, then, would most select ‘Jesus Barabbas’ as the prior reading? Because, in this case, two other criteria (the reading is prior that best explains the other(s), as is the ‘harder reading’) ‘trump’ all other relevant considerations.

Clearly, such complications accompany the use of the criteria, for normally some will be relevant to one case, others will not; often they will conflict with one another; or more than one reading may meet the priority test, etc. Thus, they do not function as ‘laws’ or firm ‘rules’ that can be applied mechanically. For example, a reading may be identified as the ‘hardest’ reading, yet it may not conform to that author’s style or theology. Or one variant may appear to explain the others, but at the same time may lack support from early manuscripts, or may evidence harmonisation. At such points the time-honored text-critical procedure of ‘the balance of probabilities’ enters the assessment, and one must circle back, reapply and rethink the criteria, place the emerging results on a balance scale, and move toward the direction in which the scale tips. All credible criteria–all probabilities–are on the table as the textual critic works, and it might be said, at the risk of inducing skepticism, that often the final question is, ‘Which probability is more probable?’ Yet, as all will agree, the method works well in the vast majority of cases.

One crucial step, however, may be lacking: weighing the various contexts of the variation-unit. This becomes an exegetical task, which is already explicit in some criteria and implicit in others, for the question becomes: Which variant fits best with the immediate or broader context?–with the phrase, the sentence, the paragraph, the writing, or the collection of writings in which the variation-unit is found? Sometimes one must go farther: Which variant makes sense in the broader contexts of early Christianity–that is, in Judaism or in Roman culture? My students often will hear the refrain: ‘The exegete is the final arbiter in textual criticism’.

Describing how these criteria function may recall the much-quoted wish of the United States President, Harry Truman (1945-1953), about his economic advisers: ‘I wish I had a one-armed economist’. When asked why, he answered: ‘Because they are always saying, “On the one hand…”, but then “On the other hand…”.’ We might echo the same wish, for this list of criteria suggests that we have a multiple-armed textual critic at work.

When these criteria/probabilities have been applied to the thousands of variation-units in the New Testament and its earliest attainable text has been constructed, admittedly the result will not be a text found in any extant manuscript or in any manuscript that ever existed, because the method produces an eclectic text consisting of innumerable priority readings in the array of variation-units. Hence, the earliest attainable text in one unit might be traced assuredly to the fourth century, in another to the third, or rarely to the second century (because only some four papyrus manuscripts can be dated that early), though citations in patristic writings can establish early dates for numerous variants. This result may appear to be minimal, yet apparently there are relatively more New Testament manuscripts closer to the time when its writings originated than found in other bodies of ancient literature, thereby enhancing the credibility of the method for reaching well back into the transmission process.

Why does textual criticism matter? Because scholars, students, clergy, and others wish to access the earliest attainable text of the churches’ authoritative writings as they engage in interpretation and exegesis. Textual criticism, however, is important beyond achieving the earliest recoverable text, for it remains to explore what insights into early Christianities might be gleaned from the textual variants that were cast aside and not selected for the earliest text-form (to be discussed below). These methods, though complex, provide credible means for making those determinations over a vast portion of the text.

C. The Traditional Goal of Textual Criticism

Throughout the modern history of New Testament textual criticism, common methods were accompanied by a common traditional goal. For the distinguished scholars in the earlier list, the purpose of textual criticism centered in the removal of scribal error so as to reveal and to restore the original text. F. J. A. Hort, for example, stated in 1882:

[T]extual criticism is always negative, because its final aim is virtually nothing more than the detection and rejection of error… . Where there is variation, there must be error in at least all variants but one, and the primary work of textual criticism is merely to discriminate the erroneous variants from the true. ¹³

Kurt and Barbara Aland express it more succinctly in their list of ‘Twelve Basic Rules for Textual Criticism’. The first is ‘Only one reading can be original’. ¹⁴ Bruce Metzger’s manual, The Text of the New Testament, for more than four decades has carried the subtitle, Its Transmission, Corruption, and Restoration, emphasizing that the goal is to restore the text after removing the corruptions that have accrued during transmission: the textual critic must ‘rectify the errors’. ¹⁵

What are the implications of these seemingly sensible statements? Foremost is their singular focus on identifying the one and only ‘true’ reading, implying that all the other variants in a given variation-unit are erroneous, and thereby essentially worthless–to be discarded like chaff in the wind. The further implication, varying in degrees from critic to critic, is that removing transmission errors is a rather simple matter and that, therefore, restoring the original text also is accomplished with relative ease. Already we have sampled some of these not-so-simple matters, and others are to follow.

The traditional view of textual criticism, in addition, often implies a neutral and detached understanding of manuscripts and variants. Manuscripts, implicitly, often are viewed as dead entities or at best passive artefacts that are mere repositories for their essentially inert variants. That is, variants are treated merely as discrete data that can be separated summarily from their manuscripts and isolated from their material contexts. Of course, the date or century in which a manuscript was copied is noted and becomes important in the text-critical process. Yet, not much attention has been paid to the context of a manuscript’s period–to its setting and to what was transpiring in the churches of the time. Closely linked is a manuscript’s provenance, though this is known with certainty in relatively few cases, but even then the manuscript’s theological and socio-cultural context is explored infrequently. A manuscript’s variants, of course, are given full attention, but most often only to identify those that are judged to belong to the original text, again with little attention to what information the rejected variants might offer. That is, textual decisions are treated as binary–in or out–and once rejected, variants are relegated to the netherworld at the foot of the pages in a critical edition, where they become, at best, second-class citizens, mostly overlooked and often without a further life. Some of us think we can do better than to deal callously with disconnected and dead variants abstracted from lifeless, artefactual manuscripts.

Clearly, then, textual criticism is a complex discipline, involving a small amount of science and a large measure of art, and is replete with challenges–I would say ‘fascinating challenges’. Critical analysis and reasoned decisions are required while working with both objective data and subjective judgments. For many, its meticulous scholarship has the attractions of detective work: studying ancient manuscripts, seeking their interrelationships, comparing their myriad variations and hunting out the earliest, and constructing a text that reaches back as far as possible. This prospect has captivated many scholars over the centuries, who ‘love the chase’ and take pleasure in arriving at satisfying solutions. Others, however, will find the complications frustrating and the endless details tedious, and they wonder, Where is the excitement? For them we hasten to point out the new elements that have appeared, with considerable noticeability and some notoriety, in the past twenty years.

A. Transition to New Approaches: The Multivalence of the Term ‘Original Text’

During an International Society of Biblical Literature meeting in Finland, a trip to St. Petersburg was offered participants, including a visit to the Hermitage museum. As I entered, an English-speaking guide, alerted to this North American group, pointed to a painting by an artist of great fame and said to me: ‘That is a genuine original’. Whether he meant that the painting was not retouched, or had been found under an over-painting, or had inspired later forgeries was not clear, but the comment confirmed my growing conviction that a concept of multiple originals was not unknown. Indeed, in textual criticism, ‘original’ has multiple usages or dimensions. ¹⁶ For example, the text of a New Testament writing might be designated an ‘autographic original’, though it is not particularly useful to discuss autographs of our New Testament writings. Another text might be called an ‘authorial original’, and that might be better, but the terms ‘author’, ‘authorial’, and ‘authorial intention’ have been problematized in recent literary-critical discussion. Yet, these two uses of ‘original’ would attempt to describe what an author wrote, apart from later rewritings. ‘Canonical original’ might describe the form in which a writing or a collection of writings (e.g., the Gospels or the Pauline letters) gained some official recognition as authoritative. During transmission, harmonisations or liturgical material, for instance, might have been introduced into the texts. Therefore, the term ‘interpretive original’ might be used to describe any such reformulation of a writing as it was altered for use in the worship or teaching, and ‘predecessor original’ for any text-form that has been superseded. Obviously, a canonical text-form also would likely be an interpretive text-form and might become a predecessor text-form.

More specifically, one of the two distinctive forms of the text of Acts is an interpretive text-form and the other a predecessor text-form, even if–as I think unlikely–the same author wrote both. So, an interpretive original replaces the prior original upon which it has imposed its fresh interpretation. It is universally affirmed that every edited or revised text and every translation of a text is in some way a new interpretation and, therefore, a new ‘original.’ Think of the numerous versions of the New Testament translated into some twenty languages and dialects! ¹⁷ Indeed, one might push this notion to its limit and argue that, in a real sense, every intentional, meaningful textual alteration to a text creates a new text-form, a new ‘original’.

As a matter of fact, this line of inquiry becomes more poignant at the level of individual variants–and more relevant to textual criticism. The so-called Lord’s Prayer has been transmitted to us in a shorter Lucan form and in a longer Matthean version, with six main variant forms in the manuscript tradition. It is easy enough to argue that the shorter Lucan form is ‘original’. But the use of that prayer in early churches, as known from extant manuscripts, reveals that worshipers in various localities, hearing the Lucan passage, would hear it in multiple forms, often with some of the Matthean features worked in. Some Christians, then, would hear God addressed as ‘in heaven’, and would pray that his ‘will be done’; some would add that it be done ‘on earth as in heaven’; some would end the prayer with ‘Do not lead us into temptation’, with or without ‘but deliver us from the evil one’. None of these variants was part of the early Lucan form. Others using Luke would pray, ‘Let your Holy Spirit come upon us and cleanse us’, but, according to extant manuscripts, no Christians hearing either Luke or Matthew in the earliest few centuries would pray, ‘For thine is the kingdom, and the power, and the glory forever and ever’, nor the words sometimes added after ‘glory’, namely, ‘of the Father and of the Son and of the Holy Spirit’. Yet, if the famous Codex Bezae, written about 400, was read, worshipers would hear the same prayer in almost identical words whether the lector used Matthew or Luke, for the scribe of Codex Bezae (or a predecessor) brought the two passages into conformity–so both versions of the prayer here became interpretive ‘originals’.

Now, the pertinent question is this: For the innumerable Christians hearing and reciting the Lord’s prayer untold millions of times cumulatively, what was authoritative and original? Unless the hearer were an Origen (who knew multiple manuscripts in the early third century) what they heard was original for them–and also canonical. So how many text-forms of the Lord’s Prayer were original in this sense, and how many originals were there? As already noted, theoretically there are as many originals as there are meaningful variants. Clearly, in our example, manuscripts with variants that conform the Lucan text to that of Matthew are interpretive originals–for they have reformulated the predecessor original text in Luke, and thereby have become new originals. And the late additions to Matthew–the dignified and majestic Doxology and the Trinitarian affirmation–are similarly interpretive, and obviously influenced by liturgical interests. Once these variant forms have been heard and recited repeatedly by Christians, as David Parker put it, ‘they will be a part of the way in which we read and interpret the Lord’s Prayer [and] we shall not be able to erase them from our minds, and to read a single original text as though the others had never existed’. ¹⁸

So, can anyone assert that rejected variants simply can be discarded?

Textual critics for some generations, uneasy about the term ‘original’, have placed it in quotation marks, and numerous handbooks have circumvented the issues it raises by stating the goal of textual criticism as seeking, not ‘the original text,’ but ‘the most likely original text’. I and many others prefer the more honest and useful statement: Textual critics seek to establish ‘the earliest attainable text’. Bentley in 1720 aspired to this goal, though he never completed his edition, and Lachmann in 1831 sought the text of the fourth century rather than any presumed original. Perhaps they were wiser than we have been, and what was good enough for Bentley and Lachmann is good enough for me, except that a profusion of early papyrus manuscripts now available permits us to move farther back in time–and with vastly increased confidence. Yet we cannot move much farther back than the third or the end of the second century. ¹⁹ For these reasons and others, the notion of ‘original text’ has shown itself to be both elusive and illusive, and we are compelled to face the fact that the ‘original text’ has exploded into a highly complex multivalent entity. It is the Holy Grail of textual criticism–conceivable, but likely unattainable. It remains a stretch, then, for Westcott and Hort to have called their 1881 edition The New Testament in the Original Greek (especially when only six papyri were known at the time–and they cited none; today we have 124 different papyri). The same judgment must be rendered on Kurt Aland’s long-standing assertion that the Nestle-Aland texts in the 26^th and 27^th editions represent the original text (and he did not use quotation marks!). ²⁰

On the other hand, to seek the earliest attainable text is a proper and reasonable goal, though alone it can be only a partial goal. In 2007 Eugene Botha offered a blunt assessment of the state of New Testament textual criticism entitled, ‘New Testament Textual Criticism Is Dead! Long Live New Testament Textual Criticism!’ ²¹ In what respect is textual criticism dead? In its traditional goal and in the increasing complexity of its methods. Barbara Aland’s 2005 presidential address at the Society for New Testament Studies ²² made it clear, says Botha, that ‘the basic assumption was still in place: textual criticism is a quest (still elusive) for the original text of the New Testament’. Botha then describes hearing this address as a ‘curious experience’ and, indeed, as a ‘revelation’, for he realized that textual criticism ‘is a discipline which has not seen fundamental changes in decades, and which the vast majority of New Testament scholarship can and are doing without’. When the people shout, ‘The king is dead!’, they immediately recognize the new king by declaring, ‘Long live the king!’ and they move quickly beyond the deceased monarch to embrace a new era. And what replaces the traditional objective and methods of textual criticism? Botha, without using the term, ventures to describe ‘Narrative Textual Criticism’, as ‘a re-imagining and re-discovery of the role of textual criticism’. ²³ My aim, and those of several others, has been similar–to breathe some life back into the process and to enlarge the vision of textual critics and of those who should profit from their labors.

B. Contemporary Narrative Textual Criticism: It’s All about Variants and It’s All about Context

If the goal of seeking the earliest attainable text is a proper, but only a partial goal, what is partial about it? It is the discipline’s almost exclusive focus on eliminating error so as to expose the single true and original text, while discarding the rejected variants along the way.

My favorite analogy here is from film-making prior to the digital camera. Movies were made by taking and retaking scenes, with the producer periodically yelling, ‘Cut!,’ and then reshaping the dialog or action for another ‘take’. When myriad scenes had been shot, the reels of film went to the cutting room, where the final editors fashioned their version of the film for distribution. So far the process is analogous to traditional textual criticism–you can walk out of the cutting room with presumably the best attainable product that had a compelling story to tell. And that was the end of it!

But now, a growing number of textual critics, mostly in North America and the United Kingdom, would react this way: We would rush back into that cutting room, sweep up the snippets from the floor, and hasten to examine them for their own stories, their own experiences, their own voices.

For instance, we would rescue from the floor an alternate text of Acts, the various endings of Mark, the Pericope adulterae, the liturgical treasures from the Lord’s Prayer, and hundreds of other narratives that linear, single-agenda scholars would leave behind. After all–to repeat once again–not every rejected variant represents an error. To the contrary, numerous meaningful variants present alternate narratives, often revealing different historical, doctrinal, liturgical, or ethical interpretations or practices arising out of real life. The New Testament text is a living text, and the meaningful variants can reveal the life, the liveliness, and the dynamism that existed in the churches, producing narratives parallel to those present in the selected text at the top of the pages in our critical editions. How, then, do we dare relegate these variants to Sheol? For untold numbers of Christians, these discarded snippets were part of their ‘original’ and sacred canonical Scripture and, what is more important, they were part of and evolved from their real life-experiences in the churches.

In 1952 Ernest Colwell startled the scholars who held that variants largely resulted from carelessness because the writings destined to become the New Testament were not yet considered sacred or canonical. But Colwell stood this notion on its head when speaking of meaningful variants: ‘most variations … were made deliberately [and] … were created for theological or dogmatic reasons’. Then came his most telling claim:

It was because [the New Testament writings] were the religious treasure of the church that they were changed… . The paradox is that the variations came into existence because these were religious books, sacred books, canonical books. The devout scribe felt compelled to correct misstatements which he found in the manuscript he was copying’. ²⁴

Now we jump forward to 1993, when Bart Ehrman created a flash point with his Orthodox Corruption of Scripture, a work now widely enough known to require only a sentence summary: During the Christological controversies of the first three centuries, ‘proto-orthodox scribes [as he called them] sometimes changed their scriptural texts to make them say what they were already known to mean’. ²⁵ Hence, they ‘corrupted’ their texts to maintain ‘correct’ doctrine, and Ehrman’s volume treated some 215 variation-units throughout the New Testament that reveal dynamic Christian communities working diligently to sort out their Christological views, all the while adding to the store of variants with fresh interpretations.

Earlier textual critics had been willing to attribute such arrogant alteration of texts only to heretics, who were accused frequently of twisting a text to favor their views. Ehrman, however, boldly and correctly turned this on its head, for it became clear that the text had been altered by the orthodox in their own interests–not by the ‘bad guys’, but by the ‘good guys’. No one will agree with every example Ehrman provides, but a real-life narrative lies behind innumerable variants, opening a window–to use his phrase–on the social history of early Christianity. Examples are ready at hand: In Mark 1:1, the variant words after ‘Jesus Christ’, namely ‘the Son of God’, and the one-word substitution in Luke 2:33 that changes ‘his [that is, Jesus’] father and his mother’ to ‘Joseph and his mother’ are anti-adoptionist alterations in favor of orthodox dogma. Ehrman argues that the two verses, Luke 22:43-44, which introduce Jesus’ ‘bloody sweat’ in Gethsemane, were inserted in the text with anti-docetic motivations. The same motivation is seen in the insertion of the verse-long variant, Luke 24:12, where the linen cloths in the empty tomb are seen as evidence of the physical reality of the resurrection. ²⁶

So, can anyone say that textual variants have not affected–or been affected by–the doctrine of early Christianity?

A second flash point came four years later, in 1997, when David Parker published The Living Text of the Gospels, ²⁷ a book unsettling for some, but opening new vistas for textual criticism. Parker’s now classic example, worth repetition, is a chapter on ‘The Sayings on Marriage and Divorce’ in the Synoptic Gospels, whose four passages yield twenty-some major textual variants. Some variants reflect Jewish, others Roman provisions for divorce; some prohibit divorce but not remarriage, while others allow remarriage but not divorce; some variants describe adultery as remarriage, others as divorce and remarriage or marrying a divorced man; and some variants portray Jesus viewing divorcing one’s wife as cruelty—thereby treating her as if she were an adulteress, though she was not—which might affirm her right to remain single, yet without implicating adultery by the divorcing man. Sometimes the divorcing man commits adultery, sometimes not; sometimes the divorced or divorcing woman commits adultery, at times also if she remarries, and, finally, sometimes a man marrying a divorced woman commits adultery. ²⁸ Parker states that this tangle of variants will ‘show that the recovery of a single original saying of Jesus is impossible’, and that one variant cannot be claimed as more original than the others, for these many variants represent ‘a collection of interpretative rewritings of a tradition.’ ²⁹

‘What a mess!’ one might conclude, but rather what these sayings embody are remarkable opportunities for a glimpse into the real life-situations of the early churches. How did these variants arise? Undoubtedly out of dynamic experiences with the emotionally-charged issues of divorce and remarriage, and what remains is a rather detailed account of a manifold problem and numerous proposed or practiced solutions. What exactly did the evangelists write? We cannot know. What did Jesus say on the subject? Layers and layers of tradition and transmission separate us from that. But a window is opened for us to observe and to re-live with early Christians the pathos and the agonizing, intractable ethical dilemmas that they faced. We can feel their pain.

So, can anyone say that textual criticism is mechanical, detached from life, and has no feelings?

Parker, as an Anglican clergyman, makes a final point: In situations like this, with no single original text accessible, no simple resolution at hand, but with multiple variants leaves open multiple options today as well: ‘The people of God have to make up their own minds. There is no authoritative text to provide a short-cut’. ³⁰

So, can anyone say that textual criticism does not conjoin us with our early Christian forebears in their ethical and moral concerns?

A final example of narrative textual criticism, among hundreds, is the so-called ‘longer ending’ of Mark in 16:9-20, which, at 171 words, happens to be the longest single textual variant in the Greek New Testament (rivaled only by the Pericope adulterae, John 7:53–8:11, at 169 words). This appendage to Mark, though known to Irenaeus in the second century, has no manuscript support until the fifth century; hence, both Codices Vaticanus and Sinaiticus (mid-fourth century) lack it, and the pericope can in no way be reckoned as part of the earliest attainable text of Mark.

During thirty years of teaching, hundreds of pre-medical and School of Nursing students enrolled in my large course on Christian origins, and I made sure they knew this passage, what it contained, and its potential for tragedy. The alleged words of the risen Christ include this insidious portion (Mark 16:17-18):

And these signs will accompany those who believe: In my name they will cast out demons; they will speak in new tongues; [18] they will pick up serpents, and if they drink any deadly thing, it will not hurt them; they will lay their hands on the sick, and they will recover.

Our University Hospital and the acclaimed Cleveland Clinic just down the street attracted clientele from a wide geographic area, including families from Appalachia–a stronghold of American ultra-fundamentalism. Doctors and nurses, of course, are not likely to persuade such a fundamentalist that Jesus never said these words or that they are not authentic to their King James Bible. Yet, I thought it mandatory to stress the inauthenticity of this passage if only one child could be saved whose parents decided to leave when a surgeon told them their child might die without an operation, and when they responded that the good folk in their church would lay hands on the child and she would recover. I thought that knowledge about this passage might deter some over-zealous preacher from handling a poisonous snake or drinking arsenic during a worship service. Every year I found ample newspaper clippings of the tragic results of these ‘acts of faith’ occurring a couple of hundred miles south.

So, can anyone say that textual criticism is irrelevant to our own times and has no human sensitivities?