The question here is divided into two because it is roughly the same question, treated in terms of either variants in our text or else variant translations, namely: Whether textual variants constitute or evidence errors in the autographa; or make the concept of the autographa meaningless; or whether variant translations do so.

We will recall the definitional objection with respect to the autographa. The original writing of each book of the Bible is not extant, and this was the case very early. Notice that the Bible never mentions the whereabouts of original copies (with the exception of the copies of the law to be stored in the ark), and seems quite unconcerned about them. 

How, though, do we understand the words of the Westminster Confession of Faith, that “by his [God’s] singular care and providence [the Scriptures are] kept pure in all ages”? It is one thing to say that this refers only to the word of God and not to manuscript copies per se; and then to follow by noting that the copies are the word of God insofar as they say what the word is. But it is in this “insofar” that one easily loses grip on where the inerrant word is.

I think an example from the history within the Bible might help here. 

Let us assume that the book referred to in 2 Kings 22:8-13; 2 Chronicles 34:14-21 was Deuteronomy. It doesn’t matter much except to keep our example as simple as possible. Here is the question: Did this Book of the Law cease to be God’s word during the time in between its loss and the moment of King Josiah re-discovering it? Hopefully we can at least see the relevance of the question. 

I would like the one who ties inerrancy strictly to the total-constant presence of that perfect word to help me understand how our modern scenario is fundamentally different from that of Josiah’s situation. I will not have much to say about translations because, in the final analysis, the person who is obsessing over that as a matter of inerrancy needs to graduate from an essentially nominalistic and hyper-inductive way of viewing word-signs as ends in themselves.

On the other hand, someone might finally get the point and then object, “But why would God inspire such a perfect word in itself but then not guarantee that His whole church would retain the whole thing for herself?” That may be a very good question; but can we at least see that it is a distinct question? The way to get our handle back on the “where” of the word of God is not to blur that distinction again, or disregard it in the first place, but to make a study of that discipline which brings the fullest picture of the whole canon to us. 

Textual Criticism and Textual Variants 101

Textual criticism is simply the science that seeks to determine the most reliable reading of a text. Although, as Paul Wegner points out, it must be considered a science and an art: 

“It is a science because specific rules govern the evaluation of various types of copyist errors and readings, but it is also an art because these rules cannot be rigidly applied in every situation. Intuition and common sense must guide the process of determining the most plausible reading. Informed judgments about a text depend upon one’s familiarity with copyist errors, manuscripts, versions, and their authors.”1

The subject matter covers over 5,800 Greek MSS (manuscripts) from the early centuries, not to mention others in Latin (over 10,000), Coptic, and Syriac (over 350). It is true that most of our manuscripts come from the second millennium AD, and many do not include the whole New Testament. The oldest known manuscript of the complete New Testament is Codex Sinaiticus which is dated to the mid-4th century. However, Daniel Wallace is helpful here is reply to the objection that only 838 of those manuscripts are from before the end of the first millennium:

“To argue that New Testament manuscripts from the early centuries are not very plentiful is only true in relation to later New Testament manuscripts—not to anything else in the ancient world.”2

Further, the critics neglect to mention “that these later manuscripts add less than 2 percent of the material to the text.”3

Beyond that, more than 30,0004 quotations of the New Testament by the church fathers have been tabulated and used to discover a pattern of the truest text. The statement so often quoted by Christian apologists about this last point was co-authored by Bruce Metzger and none other than Bart Ehrman. 

“If all other sources for our knowledge of the text of the New Testament were destroyed, [the patristic quotations] would be sufficient alone for the reconstruction of practically the entire New Testament.”5

A textual variant is “any place among the manuscripts in which there is variation in wording, including word order, omission or addition of words, even spelling differences.”6

What matters in this field of study is the nature of the variants, not the number. When we begin to see examples, we will see why that is. Now in any such case, the Greek New Testament will have a footnote directing the reader to the “apparatus” on the bottom of the page. This is what is called a “text-critical” issue, meaning that there are different manuscript traditions, but that the best manuscripts are what is reflected in the Bibles we have.

Moving on to methodology, we may picture the text critical scholar as tracing upward on a pyramid—backward toward the original. Now if we think of inerrancy as a scientific hypothesis, then its facticity ought to be a function of the prediction of fewer errors the more one moves back to the original. The claim of inerrantists is very simple: This is exactly what we find.  

Causes and Kinds of Variants  

Naturally in times before the printing press and more advanced machines for disseminating textual information, while the proclivity toward error may not have been higher, the means for standard measurement were obviously unavailable by comparison. It has always been known that the monks who labored in the scriptorium did so under more primitive conditions, whether that be from dim lighting, writing instruments, or the quality of the manuscript from which they copied. Very often, due to the high cost of writing materials, the ending to sentences were crammed or bled into a margin. Other times, a previous scribe would include a note in a margin too close to the text. The latter was called a “scribal gloss,” and these two were not always easy to differentiate.

There are similar looking and similar sounding (homophony) letters or words. The latter was a factor when more mass copying was done in a room with a reader dictating to many scribes. To cite a few examples given by Wegner, with the former, a Hebrew dalet (ד) can look like a resh (ר), and a final mem (ם) like a samech (ס). As to words, the Hebrew words for “all,” kol (כֹּל), and “voice,” qol (קוֹל), sound identical. There is also the omitting of a word (haplography), the duplicating of one (dittography), or the reversing the order of two consecutive words (metathesis).

Another phenomenon called fusion, in which two words are improperly joined together, can introduce a new concept where there is none. The example Wegner uses is of Leviticus 16:8 in which some versions leave Azazel untranslated. “Medieval rabbis identified Azazel as a hairy desert demon.” But, Wegner says, “why would Aaron give one of the goats to a demon? It makes more sense to divide the name in two … ‘for the goat of departure [or ‘going away’],’ referring to the goat that is led off into the desert. This reading is supported by the LXX and Vulgate.”7 The opposite of this is fission, namely, one word divided into two.

This is not to say that there were not also intentional changes by scribes. For the most part, a combination of reverence and community oversight prevented such deliberation; but that was not always the case. Scribes would occasionally update archaic terminology with a contemporary equivalent, or else attempt to clean up any spelling or grammatical features that were seen to be amiss. This could reflect their own ignorance so that the error was theirs. 

Now as far as how this science works with this data, Wallace says that,

“The evidence is broken down into two categories, broadly speaking: external evidence—that is, what the manuscripts, ancient versions, and patristic quotations of the New Testament read; and internal evidence—what scribes were likely to have done (such as harmonize passages) and what the author was likely to have done. External evidence and internal evidence are usually on the same side—that is, both of them normally point to the same reading as authentic. When this is the case, the decision becomes easy.”8

Obviously that is not always the case. Wallace illustrates with Philippians 1:14 and Matthew 27:16-17. In the former, no theological matter is at stake. In the latter, some early MSS have “Jesus Barabbas”

λεγόμενον Βαραββᾶν

λεγόμενον [Ἰησοῦν] Βαραββᾶν. 

While it may lead to speculation as to meaning—e.g., a “son of father” being substituted for the real Son of the Father?—the appeal to the “pious scribe,” omitting some things that seemed to involve an irreverent reading, was more reasonable than an addition. So that Jesus’ name would not be associated with Barabbas is more likely than that His name was added because it made more informational sense. 

The principle of lectio difficilior potior tells us that the more difficult reading is the stronger. The reason is that a scribe encountering a grammatical or syntactic ambiguity, particularly where he senses something doctrinally embarrassing, is more likely to smooth out the rough edges with something he perceives to be a simpler solution. Perhaps a more orthodox solution. Closely associated with this principle is that of lectio brevior which says that when a scribe amended a text it is much more likely that he did so by addition than by subtraction. Thus, if there is a clear difference, the shorter reading is to be preferred. In all such principles, the aim of this discipline is “to choose the reading that seems to give rise to the other(s).”9

Criteria for Meaningful Variants

There is a kind of psychological warfare played upon the minds of Christians in our first contact with the idea of variants. There is a saturation bombing of numbers. Bart Ehrman plays this game very well, though he admits in one book how deceptive it is. Let us begin with the bomb: approximately 400,000 known textual variants. Actually, when you include spelling errors, there are far more. How does this strike us? Daniel Wallace gives some helpful context:

“In the past century, the estimated number of textual differences has risen steadily, largely for two reasons: the number of manuscripts known to exist has significantly increased, and the painstaking analysis of individual New Testament books has been published, displaying more variants than were previously known. The best estimate today is that there are as many as 400,000 textual variants among the manuscripts. Yet the New Testament has less than 140,000 words in it. This means that for every word in the New Testament there are, on average, almost three variations.”10

Furthermore,

“if a single manuscript deviates from all other manuscripts in a given place, its reading counts as a single textual variant, and if a thousand manuscripts agree with each other but deviate from all others in one place, this wording too counts as a single textual variant.”11

But for all of that, the number still has a psychological hold on us. What then do text-critical scholars do to make sense of this? First, the more MSS we have the more variants we will have. On the other hand, as another scholar argued, “the more manuscripts we have the more we can compare their readings and trace their relationships, ultimately enabling us to have greater confidence about the wording of the original.”12

There are two key words here: viable and meaningful. “A variant is viable only if the variant has a good possibility of being part of the original wording. A variant is meaningful only if it changes the meaning of the text.”

Scholars have suggested a tetrad of kinds (the chart above). 

If we do the math, we see that less than 4,000 variants of the 400,000 total variants are both viable and meaningful, or less than 1%. One might conclude that this is still very high. 

In terms of the critical Hebrew text, “approximately one textual note appears for every ten words; thus 90 percent of the text is without significant variation.”13

Examples of Variants 

We will use the structure of the tetrad just introduced.

Neither viable, nor meaningful. Most of the variants fall into this category. For example, differences in spelling make up 70 percent of all textual variants. Many of these are obvious “nonsense readings” as even the critic will acknowledge. This is the case of the eighth century MS, codex Regius. Whereas John 1:30 should read: Ὀπίσω μου ἔρχεται ἀνὴρ, codex Regius14 instead makes the last word ἀὴρ. One can see how the scribe neglected the nu—but what does that make it? Now it means “air.” After me comes air? Of course not. Even a skeptic will acknowledge this for what it is. 

Viable, but not meaningful. One NT MSS spell John’s name two different ways in Greek: Ιωννης and Ιωνης. Obviously enough, once in the context of John’s letter, the reader knows exactly the “John” to which the two spellings refer. As Wallace remarked, “Whether John’s name was spelled in the Greek New Testament with one nu or two may remain a mystery. But John’s name is never spelled M-a-r-y.”15

Meaningful, but not viable. Luke 6:22 in the earliest MSS says: “Blessed are you when people hate you and when they exclude you and revile you and spurn your name as evil, on account of the Son of Man.” But in Codex 2882 (11th c.), it says, “Blessed are you when people hate you and when they exclude you and revile you and spurn your name as evil.” In the second case, Jesus is blessing anyone who is hated and mistreated for any reason. However, since it only occurs in one late manuscript, the relevant variant is not considered viable. 1 Thessalonians 2:9 features “the gospel of Christ” in one later medieval MS, whereas all other known MSS have “the gospel of God” in that place. 

Viable and meaningful. 1 John 1:4 says “And we are writing these things so that our joy may be complete” or “And we are writing these things so that your joy may be complete.” Some MSS says ἡμῶν and others ὑμῶν. These possessive pronouns differ by one letter. Two readings of 1 Thessalonians 2:7 give us a good example of this most difficult category and how simple it can be to resolve. Here we have a plain spelling error somewhat early in the tradition—but which one is the error? It comes down to the Greek words ἤπιοι and νήπιοι. The first means “gentle” and the second means “little children.” Either Paul said, “We became gentle among you” or he said, “We became little children among you.” It should be plain that context will resolve the vast majority of such cases. One more example will show that the word “meaningful” in this “viable and meaningful” formula does not refer to a significance of doctrinal ambiguity. In some MSS of Romans 5:1 it says “We have peace” while others say “Let us have peace.” The reason is that two forms of a Greek word differ only in the middle vowel (ἔχομεν or ἔχωμεν)—being either in the indicative or subjunctive mood. But does this altar the doctrine? Not at all. It is either telling us how justification is related to peace with God, or else it is Paul exhorting us to live in light of it. 

Translations and Inerrancy 

Another very uninformed sort of objection against inerrancy falls under this category. We may call it the “telephone game objection,” or the “translations of translations of translations” objection. In its most common form it says, “You do realize, don’t you, that each translation of the Bible is a copy of a copy of a copy,” and so forth. What such a person usually means is that it is a translation of a translation of a translation. This latter way of putting it is most ignorant. All of the translations that are worth talking about went back to the most ancient sources and compared all of the available MSS. As we saw, the Vulgate accessed the TNK directly for its Old Testament. And the process only became more sophisticated with each chapter in the history of translations. 

What often obscures this subject is the debate between King James Version Only advocates and their critics. Essentially, while the KJV drew from the more plentiful body of Byzantine texts of the later Middle Ages (the “Majority Text”), the work of Wescott and Hort (the “Critical Text”) drew from the smaller but much earlier Greek manuscripts (or MSS). The Textus Receptus (Latin for “Received Text”) is the Greek New Testament derived from later manuscripts. The suggestion is that this was the basic text accepted by the consensus of Christian churches throughout the Middle Ages. It is closely associated with (though not identical to) the “Majority Text.” 

Each side will make its case for why their body of MSS is more reliable. The earlier Greek MSS were discovered and analyzed a few centuries after the KJV translation. The facts that they were worked on at the height of the Modernist Controversy, and that two scholars (Wescott and Hort) who were not exactly orthodox in their own doctrine, contributed a most famous Greek revision, can all give rise in plausibility for the KJVOist’s concerns. It will also be said that Wescott and Hort threw out 80-95% of extant Greek MSS in favor of 5-20% with a main stress on B and Aleph.16

At the extreme are those who seem to think that the King’s English is the only inspired word of God. We need to consider a very common sense question at this point: If God inspired the KJV in such a way that its word-for-word English propositions are the essence of divine truth, then in what sense may the Bible be translated into other languages: Arabic, Spanish, Chinese, Russian, or the many African dialects? Even the KJV went through three changes dealing with things as small as spelling differences; yet this would seem to compromise the perfection of God’s word if such a wooden “word-for-word” transmission was what it means to have God’s word. Moreover, the 1611 version of the King James included the Apocrypha which is not considered inspired Scripture by the Reformed confessional tradition. But if this original version has divine authority, then so does the Apocrypha. Thus the KJVO advocate must choose between Apocrypha and the Reformed Confessions.

At lesser extremes, there is still a struggle by those in this movement to see how different translations can all be the same word of God. Surely the differences make for different truths, one might reason. However, if the differences between Greek texts were doctrinally significant, you would expect rival theologies and sects to grow out of distinctive readings of those texts. But compare the positions of advocates of each view and that wide of a divergence in doctrine cannot be found. 

Leaving beside the KJV issue, let us take any three or four of the best English translations into our hands. The differences are insignificant, and where they are, the path to the original text is easy to trace. 

Most importantly, harkening back to the lesson of Josiah and the lost and found book of the law, what exactly is the essence of the word of God? Is it what was written by the inspired author or is it what is copied by the copyists? Now that was a test. If in your mind, your next thought what, “Wait, then the word of God is as lost as the autographs and we don’t have any of the word!” then I might suggest listening again:

The inerrant Word of God—is it what was written by the inspired author or is it what is copied by the copyists? 

If you say, “Well, they can be the same in most cases, and can be discovered where they are not the same,” then you are on the right track. That’s very much in the spirit of that statement in the Westminster Confession,

“The Old Testament in Hebrew (which was the native language of the people of God of old), and the New Testament in Greek (which at the time of the writing of it was most generally known to the nations), being immediately inspired by God, and by his singular care and providence kept pure in all ages, are therefore authentical” (I.8).

Just as we saw with the bracket texts, so it is here. If “Text A” is not the canon of Scripture (Text B-X), then it is not the inerrant word. So it cannot be an argument against the inerrant word to not know which text (if any) should be in the “Space A”. That is a canon issue—not an inerrancy issue.  

At this point someone may accept the argument, but may begin to wonder what the difference is between this distinction between words and Word, on the one hand, and the distinction between words and the Word made by Barth. It is an understandable misgiving at first. However, a few important differences may be observed.

First, Barth divorced God’s meaning and the human author’s meaning, so that the latter did not in fact mean God’s inerrant word. Classical realist inerrancy maintains that all that the inspired authors wrote means what God meant as His own word.

Second, the Barthian principle separation on a “results” level is not between autographa and copies, but between the initial revelatory event and the authors’ later recollection. Classical realist inerrancy affirms the former division and denies the latter division. 

Third, the Barthian division in the Bible’s content is between the kind of subject matter that is revealed and the kind of subject matter that was not. Classical realist inerrancy maintains that every subject matter that the Bible addresses is God’s inerrant word.

__________________

1. Paul D. Wegner, “Has the Old Testament Text Been Hopelessly Corrupted?” in Stephen B. Cowan & Terry L. Wilder, ed., In Defense of the Bible: A Comprehensive Apologetic for the Authority of Scripture (Nashville: B&H Academic, 2013), 120.

2. Daniel Wallace, “Has the New Testament Text Been Hopelessly Corrupted?” in Cowan & Wilder, ed., In Defense of the Bible, 147.

3. Wallace, “Has the New Testament Text Been Hopelessly Corrupted?” 148.

4. While Wegner counts 30,000 (likely restricting himself to “the Patristics” proper), Wallace cites “more than 1 million quotations” (likely extending this throughout the entire first millennium) – “Has the New Testament Text Been Hopelessly Corrupted?” 146. In a lecture entitled, “Is What We Have Now What They Wrote Then?” at the Text & Canon Institute, Wallace pointed out that the idea that all but a few verses of the New Testament are replicated in the patristics of the first three centuries “is a myth” and that, to our great shame, it took a Muslim apologist to point this out. However, it is still the case that, “We have virtually the entire New Testament multiplied several times over.”

5. Bruce M. Metzger and Bart D. Ehrman, The Text of the New Testament: Its Transmission, Corruption, and Restoration, 4th ed. (New York: Oxford University Press, 2005), 126.

6. Wallace, “Has the New Testament Text Been Hopelessly Corrupted?” 143.

7. Wegner, “Has the Old Testament Text Been Hopelessly Corrupted?” 128.

8. Wallace, “Has the New Testament Text Been Hopelessly Corrupted?” 158.

9. Wallace, “Has the New Testament Text Been Hopelessly Corrupted?” 159.

10. Wallace, “Has the New Testament Text Been Hopelessly Corrupted?” 143-44.

11. Wallace, “Has the New Testament Text Been Hopelessly Corrupted?” 144.

12. Richard Bentley, cited in Wallace, “Has the New Testament Text Been Hopelessly Corrupted?” 145-46.

13. Wegner, “Has the Old Testament Text Been Hopelessly Corrupted?” 133.

14. Also known as Codex L.

15. Wallace, “Has the New Testament Text Been Hopelessly Corrupted?” 157.

16. Codex Vaticanus (referred to as “B” within the science of textual criticism) is one of the four main uncial codices of the first few centuries. The name derives from its location of conservation, namely in the Vatican Library. Erasmus began the work of relating it to the Latin Vulgate in the early sixteenth century. Codex Sinaiticus (referred to as “א,” or “Aleph,” within the science of textual criticism) is one of four main uncial codices of the first few centuries. The name derives from its location of discovery by Tischendorf at Saint Catherine’s Monastery in the Sinai Peninsula in the late 19th century.