Doing Internal Evidence First in Textual Criticism (Using Accordance)

For much of the history of the discipline of New Testament textual criticism, practitioners have overwhelmingly favored beginning with the external evidence before looking at the internal evidence. This has been largely a necessity because one could not determine by simply looking at the text the type of textual variant that would be found in the apparatus. Tischendorf’s magisterial Editio octava critica maior, with its extensive list of textual variants, nevertheless did not indicate in the text what kind of reading one would meet in the apparatus. Von Soden’s magnum opus also lacks any such pointers. The UBS text fares better in that it at least gives a footnote number after a word. But it still does not hint at what sort the variation is.

Perhaps this is why external evidence has been the first step in solving a textual problem: there was simply no other way to do it. Once someone glanced at the apparatus and saw their favored witnesses—whether they be א B, D F G, 𝔐, or any number after 𝔓—all too often the textual problem was considered solved. Second-year Greek students, regardless of instructions otherwise, tend to use internal evidence only as confirmation on the decision already arrived at on an external basis. Internal considerations are merely an afterthought, certainly not given equal weight with the external.

The Nestle tradition, however, gives sigla in the text to indicate what kind of variant one might expect to see in the apparatus, as follows:

⸆       insertion
⸀       substitution of one word
⸂ ⸃      substitution of more than one word between these two symbols
o       omission of one word
⸋ ⸌    omission of more than one word between symbols
⸉ ⸊   transposition between symbols
[ ]     word(s) in brackets omitted in witnesses listed in apparatus

At least the Nestle-Aland text gives some clues to the user as to the kind of variation one can expect to find. These symbols were not in the early editions of the Nestle but have been included for many decades.

This is fortuitous for the approach I take to solving textual problems. First, I ask the student to start with the Nestle-Aland text and refrain from looking at the apparatus. Second, I ask the student to come up with some guesses as to what the variant(s) might be. This is of course not necessary for omissions, simple transpositions, and bracketed words; the variants can be deduced from the sigla. But substitutions and insertions require some guesswork. And if the student can guess what the variant is, this reveals a predictable reading. If the modern student can come up with it, then scribes whose work has no genetic connection to each other’s could have come up with it on their own. But even when there is a genetic connection, working this way helps the student to create a more level playing field between external and internal evidence.

Take, for example, Phil 1.14. The Nestle text reads: καὶ τοὺς πλείονας τῶν ἀδελφῶν ἐν κυρίῳ πεποιθότας τοῖς δεσμοῖς μου περισσοτέρως τολμᾶν ἀφόβως τὸν λόγον ⸆ λαλεῖν. There is an addition after λόγον. Obviously, some adjunct, probably a genitive modifier. What is the ‘word’ that these Christ-followers might dare to speak? Typical student answers are the word ‘of God’ or the word ‘of the Lord.’ And this is exactly what we find in the apparatus: του θεου or κυριου.

This is where things get a bit muddled, however. Students notice immediately the pedigree of the longer readings: του θεου is found in major Alexandrian witnesses, along with several significant witnesses of other types (ℵ A B [⸉ D*] P Ψ 048vid. 075. 0278. 33. 81. 104. 326. 365. 629. 1175. 1241s. 2464 al lat syp.h** co; Cl); κυριου is found in F G, two leading Western MSS. The Nestle text reading is found in 𝔓46 D1739. 1881 𝔐 r vgms; McionT. Even with the papyrus and 1739 the evidence is not nearly as impressive as the ‘word of God’ reading. If students begin with the external evidence, as has been the customary practice, they may well be prejudiced against the shorter reading from the get-go because of its lack of credentials. This, in fact, seems to be the case with the third edition of the UBS text: the shorter reading garnered only a ‘D’ rating; the fourth and fifth editions elevated it to ‘B’ status.

What if students could look at the internal evidence without bias? What if they could ignore the witnesses in the apparatus and work out the problem before listening to the external voices? As we have noted, students can do this with certain kinds of variants with the Nestle-Aland text. But I did not know of any way to assist students in not letting their left hand know what their right hand was doing. Until now.


Feature in Accordance

During the spring semester of 2019, while teaching an elective on NT textual criticism at Dallas Seminary, I wrote to Helen and Roy Brown of Accordance to see if they could create a module that would enable students to do internal evidence first and without prejudice. As is typical with Accordance, I received a quick reply. They worked on this problem and soon realized that the software already could do just what I was asking for!

Here is what Helen wrote:

The illustration below shows a tab with the apparatus where the Witness field is apparently hidden, while the tab behind it has the regular display. Both are tied to the text so the user can consult whichever version he wants.

Screen Shot 2019-05-23 at 8.19.54 PM

You can do this in a separate tab (not a parallel pane), searching the Witness field for *? to highlight all the contents of that field. Then go to Set Tool Display.

[You go to Tool Display by clicking “command,”; this window will pop up:]

Screen Shot 2019-05-23 at 8.20.40 PM

You click the Customize button, and choose White as the Search Highlighting color (or whatever color your background is set to). This effectively hides all your hits. DO NOT click Use as Default as this will apply to all new views of the tool. You can however, save the workspace and the tab will retain its characteristics when it is reopened.

Screen Shot 2019-05-23 at 8.21.40 PM


See what we have done For BibleWorks Users.

Dr. Helen A. Brown
Chief Administrative Officer
Accordance/OakTree Software, Inc.


I would also recommend saving the session so that you can return to it any time you’re working with the apparatus. I called mine “NO MSS.accord.” And here is what that shows for Phil 1.14:

Screen Shot 2019-05-23 at 8.22.11 PM


I wonder if textual critics have for a long time made a virtue out of a necessity. Of course, since the days of Lachmann, Tischendorf, and Hort, the emphasis in most schools of thought has been on the external evidence. But more and more, textual scholars are recognizing that internal evidence must have its say, and it needs to do so with blinders on (as much as is possible) about what the external evidence reveals. Perhaps now that day has come.


I wish to thank Helen Brown of Accordance for help in seeing yet another potential use of this outstanding Bible software program.