Now that we have covered many aspects of textual criticism [on the CPH blog], though certainly not all, I want to conclude by the “So what?” question that may well be in the reader’s mind. I’m aware that you may have even jumped first to this article to help you decide whether reading [many long posts] is worth the trouble.
The Two Big Questions
So then, so what, now that you have read [many of our blog posts] (I’m assuming that you have)? Let’s begin with two questions that may have led you to this and other blogs and books on the subject. Many students of textual criticism became interested in it because they wondered whether the Bible is trustworthy, in the sense that what we have today is what the original writers wrote, and whether variations in the ancient manuscripts could affect our theology. Since my seminary days as a student, and before that as a younger Bible reader in church, I have received basically the same answers to these two questions from my teachers (and we have touched on them previously): first, we can be confident that we have the Bible today in its original form in existing ancient manuscripts [as far as the translation is faithful to the original language critical texts1; Westcott and Hort 1881 and the Nestle Aland 28th edition of the Greek New Testament]; second, variant textual readings affect no major doctrine.
For a long time now, I have been on the answering end of these questions as an expert on the subject. I can offer these same answers honestly, but now that you have what we have discussed as context and background, I can be candid. For the first question, while I sincerely believe that we have sufficient material in the ancient manuscripts to reconstruct the original (i.e., the autograph), it is, in fact, a belief or a matter of faith. As we have noted earlier, many conservative textual critics refuse to accept a reading that is not supported by at least one credible Greek manuscript. At the same time, it is also a fact that the latest edition of NA text (NA28) has one reading that is not found in any extant Greek manuscript: the “not” in 2 Pet. 3:10 (discussed in connection with the CBGM). It is found in some important ancient versions, but the NA (and GNT) editors have essentially testified that the word is missing from all ancient Greek manuscripts. At the same time, they would say that the word was in the original and was simply lost at a very early stage, but not before it was copied to the ancient versions.
This immediately takes us to one reason that TC matters: the experts disagree about the wording of the text in various places. I am not referring to routine disagreements over the KJV or NKJV against translations based on Alexandrian manuscripts or over-literal vs. thought-for-thought translations, though those disagreements are very important. Instead, I am referring to such differences as we find in 2 Pet. 3:10, where determining the original reading presents the challenge of evaluating readings by internal and external criteria, as we have discussed. I doubt that every textual variation of this kind will be a matter of concern, but many will.
To be true to their consciences, textual critics who reject readings not supported by a Greek manuscript would be obligated to reject the NA28 reading in 2 Pet. 3:10, and the same could be said of any Bible student who takes the same position. For now and the foreseeable future, this is the only reading of its kind in the NA/GNT editions. However, as we saw in our discussion of the problem, the decision ultimately rests on the “harder reading” principle and the provision that the harder reading can be rejected if it is judged “impossible.” Many readings in the NT are judged by this principle and are, therefore, subject to opposing verdicts by textual critics. Whenever this happens, readers must somehow reach their own decisions. Indeed, critics and editors of the Greek text will sometimes disagree on readings due to other criteria as well. Thus the stability of the text as it appears in any printed Bible is something of an illusion. New editions are always coming out, and some of the changes may be due to differences in chosen variant readings.
What are readers who are not Bible scholars to do about this? If you are among this largest group, you have taken the first step by reading this [blog] (of course, I would say the same thing if you read another [blog or book] on TC instead). You now know the issues and the criteria used for making decisions about readings. Many no doubt will prefer to trust a particular Bible translation and their pastors to provide additional guidance for textual problems. Whether they fully realize it or not, what they are saying is, “We’re counting on you, Mr./Ms. Translator, or you, Pastor, to make the decision for us as to what the Bible actually says.” But I can tell you that the translators and pastors who are worthy of your trust will reply that they are committed to doing their best, and that is the best they can do. We who do this work all have varying levels of confidence over the textual decisions that we make, and some decisions are very difficult. For these and many other decisions, I feel what medical doctors must feel when they are discussing alternative treatments with critically ill patients. We can make recommendations if we have at least a shred of evidence to back them up; but in the end, it is the patient’s, or the reader’s, decision. I feel compelled to add, too, that very often a pastor is not going to have sufficient training and experience to be of real assistance. There is only so much that can be covered even in three or four years of seminary, and the work of pastoring usually does not allow time for study and practice in TC.
If you have not yet chosen a particular translation as your personal Bible, you may think that I have just made a decisive argument for the KJV, or at least the NKJV. We know that the text of the KJV will not change (barring misprints), so you will not have to worry about variant readings. The issue for the KJV is whether it is–as its most dedicated advocates maintain–the autograph. If it is not (as I would maintain), then it needs many changes.
It’s Up to the Reader to Choose
Now, let’s summarize this first point: TC is important because textual critics disagree with variant readings. As a result, readers ultimately have to decide for themselves how the Bible should read in these places, and they need some familiarity with TC. I should add that they also need access to variant readings, at least the more important ones. Most good translations provide notes about these variants, at least in the editions intended for study. More information can be found in good commentaries, and at a minimum, pastors should be familiar enough with the original languages to be able to identify variant readings.
EDWARD D. ANDREWS: Because textual scholars operate on rules and principles that must be applied in a balanced manner and textual scholars are coming at it from different mindsets and purposes, the choices that are made can vary. This is true also when we consider literal Bible translations (ASV, LEB, NASB, UASV), semi-literal translations (ESV, CSB), semi-interpretive translations (NIV, NRSV), interpretive translations (NLT, TEV, CEV), and even paraphrased Bibles (TLB, MSG, PHILLIPS). So, the Bible reader needs to be informed about different Bible translation philosophies. By default, literal Bible translations are more accurate, more of what God said rather than what the translator interprets the words as what God said. If you, the reader, want to have what God said by way of his authors and then to decide for yourself what the author meant by the words that he used, then you want a literal or semi-literal Bible. If you want to know what the original reading was and to be sure that the textual scholar made the correct choice, then you need to have a basic understanding of New Testament textual studies.
Do Different Readings Affect Theology?
Now, let me attempt a candid answer to whether variations in ancient manuscripts could affect our theology. We can be quite confident that variant readings do not and will not affect major doctrines because these doctrines do not depend on one or two verses. Probably the most famous example of this is the spurious version of 1 John 5:7-8 found in the TR/KJV. Opponents of the doctrine of the trinity (such as Jehovah’s Witnesses) have wreaked a great deal of havoc among the orthodox by pointing out that the passage was not originally in the Bible. However, conservative scholars have for a very long time acknowledged this fact and pointed out that the trinity is nevertheless biblical, due to other proof texts.
The same came probably said about various minor doctrines as well; we just don’t worry about them as much. One that has attracted my interest is subject to change due to variant readings, in my opinion: the doctrine of forgiveness. It seems clear that Jesus was generous with forgiveness and that He taught generosity in forgiving others who do us wrong. Society, in general, seems to find that appealing, and we tend to consider those especially noble who forgive their offenders even when the latter are incorrigible and do not ask for forgiveness. In the case of those who have been wronged and are reluctant to forgive, we encourage them to do so, and this is frequently taken a step farther: the victims are encouraged to forgive for their own benefit, whether their kindness is acknowledged by the offenders or not.
I’m sure we have all seen a conversation played out in which a reluctant victim is gently disciplined by being told that God expects him or her to forgive regardless of what the offender does. If a proof text is called for, the most likely to be used probably is Luke 23:34: “Father, forgive them; for they do not know what they are doing.” We certainly want to emulate Jesus, and it is crystal clear that when He said this, the people crucifying Him were not acknowledging their sin or asking for His forgiveness.
But did Jesus actually say this? If you check your study Bible (a note may be in Bibles without commentary as well), you will find that early manuscripts do not have the verse. Actually, the information may not be as specific as that so you may need to consult a commentary or other source. Textual critics are agreed that the line is not scripture because it is missing from the best manuscripts. That is only external evidence, but I expect that the reading without the line will prevail eventually in the CBGM as well. In terms of internal criteria, the shorter version better serves as the origin of the longer than the reverse. We can easily imagine a scribe adding the line to his copy from another source when his main exemplar did not have it. It is far less likely that the scribe would delete the line when his exemplar had it.
It can be argued that Jesus might actually have said this even though it was not recorded as scripture, but when I considered that the line was spurious, it occurred to me that Jesus’ teaching seems rather to be generosity in forgiveness whenever the offender comes to the victim and acknowledges the wrong. Probably the best example of His teaching on the subject is Matt. 18:21-35, in the parable of the hopelessly indebted slave and the king. It begins with Peter asking Jesus how many times he should forgive an offending brother, the answer essentially being an unlimited number. In the brief parallel in Luke 17:3-4, Jesus actually points out that the offending brother repents.
Following all this out, I came to the conclusion that Jesus quite possibly did not teach a blind forgiveness, i.e., forgiveness whether the offender cares or even knows about it, but unlimited forgiveness that is conditioned on the offender repenting or seeking it. This would not make Jesus less generous in my mind, but rather more sensible. When we seek forgiveness, we want to restore a broken relationship and/or avoid the consequences of a wrong we have committed. It seems clear from NT teaching on forgiveness that it is for the benefit of the offender. If that person is unrepentant, however, he or she has no reason to seek forgiveness, other than to avoid any painful consequences. But avoiding the latter will again lead to seeking out the victim. So the only situation that makes sense for dispensing forgiveness is that of the offender coming to the victim in repentance. It seems to be necessary in the relationship between God and man (cf. 1 John 1:9).
I could say much more about the topic, but the point here is that the authenticity or lack of it for a variant reading can conceivably affect doctrine. In the particular case of forgiveness, it could be that we have overlooked an important detail and, as a result, some poor advice has been given to victims who are suffering unnecessarily.
Let’s consider one more example, in this case, another traditional one: Luke 2:14. Many people are aware now that the KJV’s beloved “good will toward men” is a variant reading supported by a large number of weak manuscripts. The reading supported by the best manuscripts, which is also the harder reading, is “men [or people] of good will/pleasure.” The difference in the Greek is purely grammatical. Good will can refer to people who are kind or have good intentions toward others. The Greek can also be understood as “of good pleasure,” in which case it can be interpreted as referring to God’s good pleasure. From that, we get the idea, “with whom He is pleased.” In a good study Bible, you will see some note about this.
If you haven’t already noticed it, the difference is significant for the doctrine of salvation. “Good will toward men [mankind]” is easily understood as God’s best wishes for all mankind, pretty much the kind of Christmas that the average person believes in. On the other hand, “men of good will” would indicate that the blessings of Christmas are only for good people, not everyone, and there are rewards for good deeds. If we narrow it down to “with whom He [God] is pleased,” it can be understood as those of the right faith, or even as the elect, Calvinisticly speaking. So there are several interesting interpretations of this verse–a verse which much, if not most, of the English-speaking world, is familiar. But determining the correct interpretation begins with choosing a variant reading through the practice of textual criticism.
I hope this answers the “so what” questions. There remains a “Now what?” question: what changes can we expect to see in the New Testament (and by extension in the OT, though we are discussing only the NT) going forward as the result of textual criticism? In simpler times, when I was also much younger and naïve, I would say that new manuscript discoveries are taking place all the time. That was true then, and it is still true, but I am going to continue with a candid appraisal here. I think that, at least for conservatives, the impact of new discoveries will be minimal compared to the impact of new systems for evaluating and choosing variant readings. For example, we are not likely to discover any manuscripts from the early centuries, with the possible exception of fragments. Late manuscripts will continue to surface, and their value depends in large part on one’s principles of textual criticism.
Fortunately, I have a real-world way to provide examples of future changes. I have indicated that I take mainly a positive view of the CBGM and that in any case, it will play a major role in TC because of its use in editing the current and future editions of the NA/GNT text. The current (NA28) edition incorporates 34 changes in the General (Catholic) Epistles relative to the previous edition. I propose to go through some of them, explaining not only what they are in the Greek, but also how the English text might be translated differently as a result. Whether these changes will prove to be representative of all changes that will eventually be made to the NT text as a result of the CBGM, we cannot know, but it seems more likely than not.
We have already looked at three of the 34 changes in the chapter on the CBGM. The first was Jude 5, where the grammatical subject “the Lord” has been changed in NA28 to “Jesus.” This decision, ironically, need not have had anything to do with the CBGM, though it did. Traditional TC supported “Jesus” as having better external support and being the harder reading. It was rejected on the basis of the provision to the latter criterion that it was an impossible reading. “Lord” was deemed possible because there was the potential for confusion between the nomina sacra abbreviations for “Lord” and “Jesus.”
The next example was James 1:20. In this case, the difference between the readings in NA28 and the previous edition (NA27/26) is much more subtle: the verb for “accomplish” or “achieve” is intensified in the Greek of NA28. However, the difference in meaning is so slight that the translation of the previous NA could be retained.
The third example was 2 Pet. 3:10, and here again, the decision ironically had nothing really to do with the CBGM. It was another case of rejecting the harder reading “will be found” as too difficult to be possible. In looking at additional examples, therefore, I will select only examples that were not clearly outcomes of a rejected harder reading.
James 2:3 is a good example, where the issue is the order of wording, specifically the position of the Greek for “there.” NA27 (and previous editions) have it with “stand,” supported mainly by manuscript A (02) if we take a traditional perspective. If we consult the CBGM online tools, the calculations tell us that A inherited the reading from 01 and other related manuscripts.
In NA28, the position of “there” has been changed to follow the command “sit,” supported mainly by B (03). The difference in meaning is interesting. The rich man has just been directed to a good seat “here.” If we follow NA27 (as in the NASB), the poor man is told to “Stand there, or sit by my footstool.” We can clearly see the preferential treatment. On the other hand, if we follow ms B, the translation would be “Stand, or sit there by my footstool.” It seems clear that this is harsher treatment, arguably a better fit for the context. From a traditional TC viewpoint, it also has better external support than the previous reading supported by A.
Given these appraisals, you might wonder why the reading supported by A (02) was previously chosen by Metzger et al. for earlier editions of NA. In his textual commentary, he maintains that the scribes of B (03) and other witnesses transposed the Greek “there” to produce a parallelism of two, rather than three, references to places–i.e., the “here” for the rich man’s seat and the “there” where the poor man is told to either stand or sit. Metzger surmised that the scribes missed the point that the footstool was the nearest place to the speaker. He does not say anything about a harder reading; rather, this decision seems to be solely about reconstructing the origins of the variants. As such, it illustrates the first fundamental principle, as I noted in the chapter on criteria above.
I think we will all agree that the choice of reading here will not affect theology in any way. Nevertheless, we may well prefer the reading chosen for NA28 just because of its contribution to the context. There are a few other changes in James, but their effects on translation are not significant for our purposes here. To illustrate, I cite the last one, which is found in 4:10. It is a question of whether “Lord” has the Greek article (“the”) or not. Since we know that James is referring to deity, we would translate with the article in any case as opposed to “a lord.”
More of the 34 changes occur in the epistles of Peter than in the other books combined: a total of 19. Again, however, not many of these are significant for translation. We have already discussed one that is significant in 1 Peter, i.e. 4:16, and I refer the reader to that discussion. There are five variant readings before that where NA28 differs from NA27, but none affects translation or is really significant.
The next reading that does affect translation is the presence or absence of “Therefore” (Greek OUN) in 1 Peter 5:1. The Greek can be translated in different ways, but in other passages, Peter consistently uses it to infer something from what precedes, so “Therefore” fits. In this particular case, the text-critical issues are interesting, so I will address them. We begin with the fact that the editors of NA28 chose the reading supported by the Byzantine text, which omits “Therefore” (OUN), as opposed to the Alexandrian manuscripts.
At first glance, this looks as though it could be a choice of a Byzantine reading over the Alexandrians for its own sake. It is more complicated than that, however. While traditional criticism would tend to ignore late manuscripts and consider mainly (or only) the early Alexandrians, we must bear in mind that the CBGM ignores the dates of manuscripts. In this case, 1739 and 81, which both reflect the Alexandrian “text-type,” part company with the older manuscripts and have the Byzantine reading. So as far as the CBGM is concerned, we have here a split in the Alexandrians, with important representatives lending support to the Byzantine variant. If we consult the CBGM diagrams online, we also find better coherence on the Byzantine side.
What we have just considered amounts to external evidence in the CBGM. Internally, we should ask which reading better accounts for its counterpart. For an inferential particle (often called a “marker”), there must be something in the preceding text that serves as a logical basis or precedent for the statement with the particle. The particle becomes difficult to explain when there is little or no logical connection between what follows and what precedes. I think it is a matter of opinion here, but I see little connection with the preceding (the end of chapter four). If I am correct, it follows that a scribe would not add “Therefore” if it were not before his eyes in his exemplar. On the other hand, it seems plausible to me that the Greek word could be omitted by accident due to its resemblance to the endings of the words before and after it, i.e. both -OUS. I must add that my preference is a qualified rejection of the criterion favoring the shorter reading. So on balance, it is not difficult to see why the ECM editors chose the reading found in Byzantine mss. We can also conclude that while their choice affects translation here, like other variants so far, it has no implications for theology.
There are two remaining textual variants that have been changed for NA28 in 1 Peter, one in 5:9 that does not affect translation, and one in 5:10 that actually does not change the choice presented to translators in NA27. This brings us to 2 Peter, where most of the changes occur.
I should acknowledge the first variant, which occurs in 2:6 because the textual choice significantly affects the translation. We will have something like either “an example to those who were going to behave in an ungodly way,” or “an example to ungodly people of things [i.e. punishments] to come.” In NA27 the choice was clearly marked in the text by brackets, indicating uncertainty on the part of the editors. In NA28 the brackets are gone, leaving one choice (the Greek infinitive) preferred, probably because it exhibits perfect coherence in the CBGM analysis. However, any translator paying attention to the apparatus will find the same options with their support, so I doubt that the change in NA28 will be of consequence.
The next variant reading of significance occurs in 2:18, where the ECM editors have chosen “really” or “actually” in place of “barely” in the phrase “barely escape….” The two words are almost antonyms in both the Greek and the English, and “barely” seems a much better fit for the context. Orthographically they are so similar that scribes could easily miscopy either or assume that one was mistaken for the other and change what is found in the exemplar. It happens that the choice of the ECM editors is also the reading of the Byzantine witnesses, who join 01 against 03. Such an opposition of readings could be very significant if the subject matter involved doctrine. It is also interesting that the original choice of the other reading had the complete confidence of the editors (an “A” decision); this was not a case of a slight tipping of the scales.
A change of what was a rather doubtful reading occurred for 2 Pet 2:20, with the deletion of “our” following “Lord.” It is of little or no significance to the meaning of the passage, so I think we can move on to the next variant unit, the form of the relative pronoun following a preposition in 3:6. NA28 now has the pronoun in the accusative case, changing the meaning of the preposition from “through/by” to “because of” (“because of which [water] the world…was destroyed”). There is a significant difference in meaning, and we would need to perform a good deal of exegesis to do the matter justice.
I think it is fair to say that after 2 Pet 3:10 (discussed earlier), the only other variation unit in 2 Peter of significance for our purposes here occurs in 3:16, where the uncertainty is the tense of the verb “distort.” In NA27 the present tense was the accepted reading; in NA28 the reading chosen is the future. It is a significant difference in meaning (“the untaught distort/will distort, as they do…”) but the point seems to be the same either way, one that does not involve doctrine.
There are four changed readings in 1 John, but only one, found in 5:18, is significant for our purposes. The question is whether the third-person personal pronoun follows the verb “keeps,” or the reflexive pronoun (“himself”). This, in turn, leads to (or follows) a decision whether the subject of the verb is the divine Son or the believer. If the former, the implication is that the Son protects the believer from a sinful way of life, which would be a doctrinal matter. In NA27 the personal pronoun was chosen supporting this doctrine; in NA28 it is the believer keeping himself from sin.
There are two changes in 2 John, both changes of word order, neither significant for us. In 3 John there is one change, the presence or absence of the article before “truth” in verse 4. Some would question whether such a small change is significant. I would say that “walking in truth” (without the article) could refer to honest behavior, while “walking in the truth” more likely would refer to following Christian doctrine faithfully.
We come finally to Jude, where there are three changes. The first is quite interesting, mainly substituting “Jesus” for “Lord” in verse 5. This would be a clear reference to the activity of the preexistent Christ in the OT. As such it could probably serve as a Trinitarian proof text if the reading were strongly favored by text-critical criteria, but that is not the case. Of the two remaining changes, the first is very uncertain and of minimal importance, while second is virtually insignificant for translation.
What are we to conclude about these changes? First, there are very few of them compared to the total number of variant units in the epistles, and if anyone expects the outcome to match the expectations that we might have had given the use of a major new tool (the CBGM and nearly exhaustive database) in textual criticism, he or she will be underwhelmed. Even then, many if not most of the changes will seem inconsequential, depending on one’s perspective. I suppose we can say positively that this is a tribute both to the stability of the biblical text and the objectivity of the new technology. We may, of course, be in for some surprises as work continues on the rest of the New Testament.
So much for the quantity of changes; what about the quality of the changes that are significant? Conservative theologians undoubtedly will continue to say with great confidence that no major doctrine is affected, and for my own part, I find it difficult to imagine that to be untrue. But for a few changes discussed above, we did see the possibility of doctrinal implications. We also saw at least one case where a previously confident decision on a reading was reversed. I have, moreover, seen such reversals on readings between the publications of ECM (first edition) and ECM2. So there are no guarantees that decisions about variant readings will remain the same, at least not for a number of future ECM/NA/GNT editions. For translators, even one significant change is important; and for theologians, any change in a verse affecting doctrine needs to be evaluated for its validity and any possible effects it may have on the doctrine.
What We Can Expect to See
In closing, I think we are entering a new era of textual criticism and Bible translation in which not just two or three mainstream competing Greek New Testaments will continue to be available, but eventually, a handful will be or even more. For translators, in particular, it will be necessary either to take a “frozen text” position or put on the additional hat of a textual critic, unless they simply entrust themselves to a favorite textual scholar or organization to make text-critical decisions for them. I think a professional translator has an obligation to wear both hats, not only for intellectual honesty but also for the practical purpose of responding properly to challenges about his or her work. “Why did you translate X this way?” is a frequent question, and I believe it always deserves an intelligent answer. To say that the Greek reads this way in the Nth NA edition, or that Dr. So-and-so recommends it, is an honest answer, but not really intelligent. For those who are content if the Greek reads such a way in the TR, I say, God, bless them, and I’m sure He does.
 See above, pp. 154 f.
 The list of changes can be found in several sources, including NA28, 6* and 50*-51*.
 Metzger, Textual Commentary, 609-610.
 See p. 239.
 This is an English problem: “presence of Lord” (if the anarthrous Greek is preferred) would be too awkward, forcing the translator to add either “a” or “the,” and “a” (indefinite) typically implies a plurality of members in a given class.
 See above p. 275.
 Some might disagree about the participle in 1 Pet. 1:6, but I think the accusative simply assumes an implied infinitive of being.
 For example, in P72 it is very easy to misread the Greek for “barely” as “really,” the other Greek word.
Please Help Us Keep These Thousands of Blog Posts Free for All