Theology Thursday - Reasoned Eclecticism & the New Testament Text

Daniel Wallace is a scholar who advocates a reasoned eclectic approach to New Testament textual criticism. Here, in this short video, he briefly explains this approach:1

In this short excerpt from his discussion of New Testament textual criticism from the Lexham Bible Dictionary, Dan Wallace explains the nuts and bolts of the reasoned eclectic approach to textual criticism:2

Different schools of thought exist as to how best to duplicate the wording of the autographs. These schools can be organized by the weight each places on the two broad databases of textual criticism—external evidence and internal evidence …

Priorities given to either of these two groups of evidence reveal a continuum of text-critical practices. At one end of the continuum are those who believe that the internal evidence is the only sure method for recovering the autographic text; on the other end are those who view the external evidence as the only objective approach. In the middle are the majority of textual critics today: they believe that both internal and external evidence are subjective in some degree and that both are needed to reach the wording of the autographs. Textual criticism is, after all, both a science and an art.

The methods used in New Testament textual criticism can be placed on the following continuum:

Reasoned eclecticism [is] the view that is followed by a majority of New Testament scholars is reasoned eclecticism. It gives both external and internal evidence an equal hearing, without absolute preference for either type of evidence or for any text-type or group of manuscripts—though the Alexandrian witnesses are usually favored. Reasoned eclectics recognize that even though all internal evidence is subjective, it is not all equally subjective, and although all manuscripts are corrupt, they are not all equally corrupt. Also, reasoned eclectics try to address the cultural and theological backgrounds in which variants arose. Although there are problems with this view, no theory has yet replaced it. As we examine the practice of New Testament textual criticism, we will do so from the perspective of reasoned eclecticism.

Application of New Testament Textual Criticism

The basic guideline of textual criticism is: Choose the reading that best explains the origin of the others. The more that external and internal evidence point to the same reading as authentic, the more certain it is that it is the wording of the autograph. However, in many cases the data do not line up so neatly—internal and external evidence often point to different readings as closest to the wording of the original. These cases require further examination.

​Internal Evidence

Internal evidence is an examination of the wording of the variants in order to determine which reading gave rise to the other(s). The reading that is evidently the source for the other(s) is most likely to be authentic.

Canons of Internal Evidence. Almost everyone practices one aspect of textual criticism every day whenever they use internal evidence to get at the author’s meaning. When we read a newspaper article—or an email from a friend—we instinctively correct misspellings, typos, and statements that are known to contradict the facts. We need no other documents to make comparisons—what the author intended to write can be determined by examining the wording to detect known errors. This is internal evidence—the reading that is most likely on such bases is considered internally probable. (Of course, the analogy breaks down since emails are original creations, not edited and copied documents.)

There are several guidelines for interpreting internal evidence, but two are the most important:

  1. The shorter reading is to be preferred
  2. The harder reading is to be preferred

A harder reading is one that is ambiguous, grammatically or stylistically awkward, lexically less common, or theologically in tension with other portions of the New Testament. A shorter reading is one that lacks at least one word, and sometimes whole phrases or even verses. These guidelines, or “canons,” should be applied with consideration for other possible causes of corruption. If a reading could have been created unintentionally, the shorter and harder reading rules are generally ignored. The great majority of accidental readings will be harder (many are even nonsensical), and many shorter readings are the result of the scribe’s eye skipping a letter, word, or line of text. Fatigue and lapses in hearing, eyesight, and memory are major causes of unintentional variants. Thus, the possibility of unintentional changes needs to be ruled out before these two guidelines can offer any level of certainty.

Since scribes tended to add to the text—especially in the later centuries—and smooth over grammatical and theological difficulties, these two rules are quite useful for determining the wording of the autographic text. But there are other considerations, as well. The Western manuscripts, for example, are known to omit whole verses. Although this text form is quite early, it also was produced in a somewhat carefree manner. Thus, when there are internally strong arguments for a reading found in manuscripts that have a poor pedigree, the external evidence is also weighed heavily.

Divisions of Internal Evidence: Transcriptional versus Intrinsic Probability.

Transcriptional probability looks for what a scribe was likely to have written that is different from faithful copying; it then eliminates that reading from consideration. Scribes produced two kinds of alterations to the text—intentional and unintentional changes (see Metzger-Ehrman, Text, 250–71).

Copyists often consciously changed the text—either for grammatical, theological, or explanatory reasons. Because of this known tendency of the scribes, the two canons of shorter and harder reading are important guides to the wording of the original text.

For example, in John 1:34, the manuscripts are divided between two readings: “I have seen and I have testified that this is the ‘Son of God’ ” and “I have seen and I have testified that this is the elect one of God.” Looking just at transcriptional probability, the second reading appears to be more primitive. First, it is not as rich theologically as the former reading; second, “Son of God” is a favorite expression of John’s, and scribes would know this. It is easy to see, then, how a copyist would change the text of his exemplar from “elect one of God” to “Son of God.” A change in the opposite direction is not nearly as plausible.

Scribes, however, did not just change the wording on purpose; due to problems of sight, hearing, memory, judgment, or fatigue, they often altered the text unconsciously. A common mistake was to write once what should have been written twice (haplography)—this occurred when a scribe’s eye skipped a second word or line that ended the same way as the word (or line) before it. Another was to write twice was should have been written once (dittography)—this occurred when a scribe accidentally reproduced the word or line that he or she just wrote.

A classic example that is still debated today is in 1 Thess 2:7. The manuscripts have either “we became gentle” or “we became little children” there. The difference in Greek is a single letter, the letter ν. But the preceding word ends in a ν. Did Paul say ἐγενήθημεν νήπιοι or ἐγενήθημεν ἤπιοι? That is, did the scribes write one ν when they should have written two, or two when they should have written one? In this instance, transcriptional probability only points to the mistake, but it does not help to determine which reading is correct. Although most translators have regarded the νήπιοι reading as too hard, two modern translations have adopted the reading “little children” here—the NET and the NIV 2011, supported by the NA27 Greek text.

Intrinsic probability tries to determine what the New Testament author was likely to have written. This is based on passages that are undisturbed by significant textual variation, which is the vast majority. As in transcriptional probability, two key issues are involved (though there are others as well): context and style.

Which variant best fits the context? In Romans 5:1, there is a textual problem involving a single letter in Greek. The text reads either “Therefore, having been justified by faith we have peace with God through our Lord Jesus Christ,” or “Therefore, having been justified by faith, let us have peace with God through our Lord Jesus Christ.” The indicative (“we have peace”) fits well with the overall argument of the book to this point. From chapter 1 through chapter 5, Paul is grounding the believers in what Christ has done for them. There is only one imperative up till this point, while from chapter 6 on there are 61 imperatives. The apostle also seems to presuppose that the audience has peace with God (via reconciliation) in 5:10. This seems to assume the indicative in verse 1.

Which reading better fits the author’s style? How does he normally express himself, what are his theological emphases, what motifs does he employ, what are his spelling habits? Although intrinsic probability is usually the most subjective aspect to textual criticism, for longer passages it becomes one of the most objective. Thus, a primary reason why most scholars regard the last 12 verses of Mark (16:9–20) as a later addition is that the vocabulary, grammar, motifs, style, and even theological outlook are significantly different from Mark 1:1–16:8. Further, when intrinsic probability is combined with transcriptional probability, the evidence becomes overwhelming in favor of excising this passage: scribes did not wish to let Mark’s Gospel end with “for they were afraid” (ἐφοβοῦντο γάρ). They seem to have written in the several “endings” found in the manuscripts. Combined with the external evidence—the earliest and best manuscripts and versions, and early patristic comments that support concluding the gospel at 16:8—the evidence is overwhelmingly in favor of viewing Mark 16:9–20 as a later addition (see Wallace, “Mark 16:8 as the Conclusion to the Second Gospel,”1–39).

Once all the internal evidence has been examined, textual critics usually have a good idea as to which reading was the origin of the other(s). To the degree that the intrinsic and transcriptional probabilities confirm each other, this preferred reading likely reflects the original wording.

To be continued …


1 The video is an excerpt from an edition of the John Ankerberg Show, entitled, “Which English Translation of the Bible is Best for Christians to Use Today?”The entire show is available for purchase here. This excerpt is publicly available on YouTube.

2 Daniel B. Wallace, “Textual Criticism of the New Testament,” ed. John D. Barry et al., The Lexham Bible Dictionary (Bellingham, WA: Lexham Press, 2016).

2847 reads

There is 1 Comment

TylerR's picture


One thing people rarely talk about is whether there is any profit in going with a single-source manuscript for the NT. After all, this is what we do with the modern OT text (BHS, for example, is based on the Leningrad Codex). Some scholars have suggested we can do the same with Sinaiticus or Vaticanus.

Tyler is a pastor in Olympia, WA and works in State government. He's the author of the book What's It Mean to Be a Baptist?

Help keep SI’s server humming. A few bucks makes a difference.