In the recent issue of New Testament Studies, I have an article that presents the results of a long project studying the textual variants in sixteen key manuscripts for Acts and the Catholic Epistles:
“Quantifying New Testament Textual Variants: Key Witnesses in Acts and the Catholic Letters,” NTS 64/4 (2018): 551–572.
The abstract reads:
This article interacts with Peter J. Gurry’s recent estimate of the total textual variants in the Greek New Testament (NTS 62 (2016)) by (i) employing a different (and complementary) method using data from the Editio Critica Maior and (ii) producing an estimate that is narrowly confined to the ‘key’ manuscript witnesses for Acts and the Catholic Letters (a mix of majuscules and minuscules, both Byzantine and non-Byzantine). The results prove more useful for framing the development and distribution of textual variants in this group of key witnesses.
It was subsequently discussed over at the Evangelical Textual Criticism blog (thanks, Peter).
What is this article about?
About two years ago, I got involved with Peter Gurry and Elijah Hixson on their book project dealing with New Testament textual criticism (Myths and Mistakes in New Testament Textual Criticism, with IVP Academic). My main focus for that project was on minuscule manuscripts (see also my CBR article and ETS-2017 presentation). However, as part of the broader work I was doing, I became interested in using the robust set of data made available by the ECM project (which is charged with updating our Greek New Testaments) in order to understand in a more robust/quantitative way the nature of textual variation in the NT.
Gurry had, while we were in Cambridge, done some excellent work in quantifying the total number of known textual variants in the Greek NT textual tradition. But with the new data available and a lot of Excel crunching, I realized that his work could be taken further.
My particular goal in this sub-project was to understand the profile of textual variation not for all the ~5,600+ Greek NT manuscripts, but for (what turned out to be) the sixteen or so key ones that represent the most important parts of the textual tradition. To me, those manuscripts were most relevant because they represent the waterfront in terms of what various parts of the church were using as the text across the empire and for about a 1,000 year period. They also happen to be the ones that textual critics use the most in making decisions.
How did I do this analysis?
The article provides a high level summary of the method I used to wrangle the ECM data into usable form, drawing on the two websites that make it available:
- General epistles: http://intf.uni-muenster.de/cbgm2/Comp4.html
- Acts: http://ntg.cceh.uni-koeln.de/acts/ph4/
It took lots of man-hours, but ultimately I was able to produce a massive database that included all textual readings for each of these 16 manuscripts at each point of variation, so that I could do something with it. It looked something like this:
Yes…it goes out beyond column BH, which means there were over 60 columns of data (and over 10k rows). (And here I give a shoutout to the Boston Consulting Group, where I learned nearly everything I know about MS Excel).
From that data set, I was able to slice, dice, filter, compute, etc. to my heart’s content. Most of the analyses I did were not all that illuminating, even if they produced cool charts like this:
But after swimming in the data for a while, I developed a few testable hypotheses got to work on them. This quantitative analysis ran in parallel with a ton of reading research I was doing for the other projects, particularly on the minuscules and Byzantine tradition.
Ultimately I married the two strands of research together, and one of the results was the NTS article.
What are the key findings (or: why should I care)?
As I state in the article, some of the key learnings—which are limited to Acts and the Catholic Epistles (CL), for we only have full data for those at this point—are as follows:
- 15% of the text in Acts/CL is entirely non-variant for the hundreds of manuscripts collated in full (which are, basically, the only ones that matter according to the Münster team). That is to say, 1 out of every 7 words has no known variation.
- These 16 key manuscripts perfectly agree with each other nearly 70% of the time. This includes both early ‘Alexandrian’ manuscripts as well as the chief ‘Byzantine’ manuscripts. That’s a tremendous amount of textual stability over time.
- These 16 key manuscripts attest a relatively small portion of the total known variants. That is to say, the ‘trunk’ of the tree is consistent over time; we don’t see wild variability (contrary to what many critics have held in the past).
- (Note: in §4.4 of the article, I regretfully used the word “account for,” which suggests that these variants are only found in these 16 manuscripts. That is not what I meant, though my word choice is confusing. I meant “attest,” as used later in the article. My apologies for the poor word choice in that section. Thanks to P. Gurry for pointing that out.)
- For 90% of the text for which there is any known variation, we have at most 2 competing readings. While it is true that some parts of the text of Acts/CL features over a dozen competing options, this is by far the exception. In the vast majority of cases, we’re dealing with only two options (and in nearly all of these the decision between the two is not rocket science). This further corroborates the controlled and stable nature of the textual transmission.
In short, for Acts/CL at least, we have empirical evidence that demonstrates that the Greek NT text across the main parts of the textual tradition (so-called “Alexandrian,” “Byzantine,” and mixed types reflected in non-Byzantine minuscules) is remarkably stable. Yes, there are ~500,000 total textual variants (as Gurry originally estimated), but that is due to the staggering number of total copying events. When you narrow in on the manuscripts that matter most, the situation looks quite different.
If you are interested in seeing the full article, send me an email.