Friday, June 08, 2012

Carlson's "The Text of Galatians and Its History"

Stephen Carlson’s dissertation, “The Text of Galatians and Its History,” is now available at Duke Space here.


This dissertation investigates the text of Paul’s Epistle to the Galatians and its history, how it changed over time. This dissertation performs a stemmatic analysis of 92 witnesses to the text of Galatians, using cladistic methods developed by computational biologists, to construct an unoriented stemma of the textual tradition. The stemma is then oriented based on the internal evidence of textual variants. After the stemma is oriented, the textual variants near the base of the stemma are examined and the text of Galatians is established based on stemmatic and eclectic principles. In addition, two branches of the textual tradition, the Western and the Eastern-Byzantine, are studied to assess the nature of textual variation in their history. This study reaches the conclusion that a modified stemmatic approach is an effective way to study both the text of a New Testament book and its history.


  1. Thanks, SC and Duke people, for making this interesting work available. I look forward to sifting through it carefully.

    Selectively turning to page 320, this statement seems interesting:

    "There is little evidence that the Byzantine text is the result of a recension aside from a small influx of Western readings into the common ancestor of the Syrian group and the Byzantine text."

    I wonder what Daniel Wallace -- recently seen online still on board H.M.S. Hort, apparently still confident that the holes in the hull are but a scratch -- thinks of that.

    Yours in Christ,

    James Snapp, Jr.

  2. Thanks for letting me know, Tommy!

  3. Great stuff Stephen,

    I haven't gotten through the whole thing yet (I'm in Chapter 4), but I did have some comments generated from the previous chapters.

    1) The intro is probably one of the best summations of the goals and techniques of TC I've yet read. Certainly it's a great introduction to orient someone in how TC works and what the new techniques are.

    2) Tractability: As a programmer (working in the cryptographic world), I translated your 10**166 step number into binary. That's 2**547. In cryptographic terms, if you had a computer that could check all possibilities, then you could break pretty much any cryptographic key ever made. The strength is twice the size of the maximum security level defined by NIST.

    This brings up a interesting question, though. Clearly smaller stemmas are and easier problem to deal with than larger Stemmas from the exhaustive search point of view. Do you know if there is was any research on breaking the problem into smaller stemmas, completely solving the smaller stemmas, then merging the resulting solved stemmas into a larger one. I suspect it may be and NP or NP-complete problem. It clearly falls into graph theory, so someone versed in graph theory could probably answer this question right away.

    2) As a next step you talked about doing 1 Corithians. It seems to me that Romans, 1 Corithians, 2 Corithians, and Galatian should pretty much have the same stemma. Once you run your stemma finding program on these texts, if the stemma is different from Galatians, it seems you would want to calculate the changes in Galatians with the new stemma to see if it's an improvement on your existing stemma, and vice versa (see if your existing stemma as less changes in 1 Corithians than the new one you calculated). The good thing about your method is that if anyone finds a stemma with fewer changes, it doesn't take weeks or months to verify that it has fewer changes.

    3) You mentioned that the Majority text editions don't help in either finding the original text or the history of the text, but I think you meant, a priori, they don't help. With your stemma for Galatians, we can't we presume that the majority text editions are actually fine representatives Byzantine base.

    4) I was surprised to find some discussion of the external evidence in Chapter 3. I was under the impression the you weren't going to use external evidence to orient the stemma. Doesn't that bias the orientation, or are you only using the unbiased stemma in your external evidence?

    Any thanks for this work, I still have more to read, but I've read so far sounds really cool (particularly to the programmer in me;).