Excerpt from: S. Carlson, Hypotyposeis,
TC Blog, (2009)
Last Updated: Feb 15, 2010
Recently a variety of methods have been applied in an attempt to group manuscripts and text-types, and to organize them genealogically. One such method is the Coherence-Based Genealogical Method (CBGM) proposed by Gerd Mink, and discussed by Klaus Wachtel. Here Mr. Carlson reviews the technique and attempts to rate it as a tool in an ever-expanding arsenal of novel methods and techniques.
[A Review of]
Coherence-Based Genealogical Method
Stephen C. Carlson
Tuesday, December 01, 2009
For more reviews and updates by S. Carlson, see his web-blog here:
Headings have been added for clarity and navigation purposes.
Stephen C. Carlson (a review)
Tuesday, December 01, 2009
Given the recent interest in the Coherence-Based Genealogical Method (CBGM) at SBL and other places, I have decided to post an article review I had prepared about the CBGM for a seminar. Owing to the complexity of the Method, it is very possible that I have not completely understood various aspects of it, and I welcome comments and corrections. - Review follows.
Gerd Mink’s Coherence-Based Genealogical Method (CBGM) has been described in a number of publications beginning in 1993. These articles are listed in the appended bibliography. The initial articles, from 1993 and 2000, are the result of Mink’s preliminary investigations and represent an earlier state of his thinking. The most comprehensive description of his method is found in the 2004 chapter, “Problems of a Highly Contaminated Tradition,” and Klaus Wachtel has just recently published an accessible summary of the method in 2008.
The purpose of the CBGM is to deal with the circular reasoning that Mink identifies as inherent in textual criticism: witnesses are good because of their good readings, but readings are good because of their good witnesses. Mink does not think that this circular reasoning can entirely be avoided but he proposes that it can be controlled through his CBGM. In particular, he sees the CBGM as being repeatedly applied to the state of the critical text (now called the “initial text”), each time successively refining the text to be better supported on Mink’s genealogical principles, especially that of genealogical coherence. It is important to understand what the CBGM is and what it is not. As Klaus Wachtel (2008: 127) put it:
The CBGM does not provide a means of automating the reconstruction of the initial text, nor the ‘royal way’ to it. Furthermore, the CBGM often does not make textual decisions more secure. It may cast new doubt on cases that seemed to be settled. It is a way to analyse the structure of a manuscript tradition and to integrate methodically our growing knowledge about it.
Accordingly, the CBGM is not an automated method for generating a stemma of the New Testament or otherwise to derive the textual history of the New Testament. Rather, it is a technique for lending more precision for weighing the external evidence in support of various readings in the course of deciding the critical / initial text. As rational eclecticism has commonly been practiced, the weight of external evidence has generally been put upon a few, subjectively pre-eminent witnesses, such as Codex Vaticanus and Codex Sinaiticus. Given that Muenster has been looking at a much larger set of manuscripts, the CBGM has been developed to evaluate these witnesses in a more rigorous and precise manner.
A Description of CBGM
The CBGM works as follows. First, it assumes an “initial text” that is relatively close to the form of the text from which the textual tradition of a New Testament book has originated. The first iteration of this “initial text” seems to have been the Nestle-Aland 27th ed. Then, at a large number variation units, a local genealogy of the variant readings is drawn up, showing the evolution of the variants without regard to the witnesses that transmit those readings. At this point, it is not necessary to perform this analysis for all variation units, just the ones whose local genealogy is fairly clear and uncontroversial. A local genealogy can be represented by a local stemma, and it indicates which variants are to be considered prior or posterior to other variants.
Next, the local genealogies are considered for each witness to identify which other witness is a potential ancestor. The best potential ancestors are those which have a high degree of agreement in readings and, of the variant units where they do not agree, the potential ancestor has a high proportion of the variants that are prior to the readings found in the witness under consideration. Thus, a “substemma” can be established for each witness that indicates which other witnesses are the best potential ancestors for it.
These various substemmata can be aggregated to produce a global stemma, which shows the textual flow of the tradition. Now, Mink is careful to warn the reader that his global stemmata is not the same as the classical stemmata in the Lachmannian or Maasian sense. In particular, Mink eschews the use of hyparchetypes (hypothetical ancestors) and allows for most witnesses to have multiple ancestors (to account for contamination). Another major difference is that, in classical stemmatics the Archetype usually has two, sometimes three immediate descendants, while in the CBGM there can be a dozen or more witnesses whose best potential ancestor is the initial text.
Once the global stemma has been produced, the variation units of the initial text can be revisited, either to revise the local stemma, or even to modify the reading of the initial based on the quality of the “coherence” of the respective variant readings. As best as I can tell, Mink does not provide a rigorous or mathematical definition for coherence, but its essence seems clear enough. Basically, a reading exhibits good coherence if it is found only among witnesses that are closely related genealogically, and poor coherence if it is found among witnesses that are distantly related. In other words, a reading with poor coherence is likely to have originated multiple times in the transmission of the text, and for that reason may be considered transcriptionally a later reading. On the other hand, a reading should belong to the initial text if it has good coherence among those witness whose best potential ancestor is the initial text. After the initial text has been revised based on genealogical coherence, the CBGM process is repeated over and over until the repeatedly revised initial text stabilizes.
Applied at Jude 5
Wachtel (2008: 121-126) has mentioned that the CBGM was used to change the critical text for the variation unit at Jude 5, where the subject of the ὅτι clause can be either “Jesus” (02 036 33C 1739 etc. with ὁ, 33* 88 915), the “Lord” (01 044; with ὁ, 61 69 etc. [Byz]), the “Lord Jesus” (1735), “God” (04C2 5 etc.) or “God Christ” (P72). Previously, the NA27 read “the Lord,” but the NA28 will read “Jesus” instead. According to the CBGM, the “Jesus” reading has good coherence because that reading is found among closely related witnesses, most of those have a best potential ancestor as the initial text. On the other hand, the “Lord” reading has poor coherence, because, according to the text flow of the global stemma, it must have originated at multiple times in multiple places. As a result, the CBGM indicates that the “Jesus” reading has strong external support, while “Lord” is likely to be a transcriptionally later reading.
To its credit, the CBGM appears to be more rigorous and precise about evaluating external evidence than the heuristic rules of thumb currently employed in reasoned eclecticism. Its main drawback is that this promised precision may be more apparent than real. The theoretical basis for the CBGM is the work of one man, and the technique has been applied only to the Catholic Epistles. It has not been tested for other texts and its operation is still poorly understood outside of Muenster. Indeed, it has not borrowed its concepts from classical stemmatics or even from phylogenetics in biology, which have been tested and found to be robust over many different applications. It is still very much an unproven method; time will have to tell as more people gain experience with it.
* Mink, Gerd, “Eine umfassende Genealogie der neutestamentliche Überlieferung” [A Comprehensive Genealogy of the New Testament Tradition], NTS 39 (1993): 481-499.
* ________, “Editing and Genealogical Studies: The New Testament,” Literary and Linguistic Computing 15 (2000): 51-56.
* ________, “Was verändert sich in der Textkritik durch die Beachtung genealogischer Kohärenz?” [What is Changed in Textual Criticism by the Observation of Genealogical Coherence?] in Recent Developments in Textual Criticism, New Testament, Other Early Christian and Jewish Literature: Papers Read at a Noster Conference In Münster, January 4-6, 2001 (STAR 8; W. Weren and D.-A. Koch, eds.; Assen: van Gorcum, 2003), 39-68.
* ________, “Problems of a Highly Contaminated Tradition: The New Testament; Stemmata of Variants as a Source of a Genealogy for Witnesses” in Studies in Stemmatology II (P. van Reenen et al., eds.; Amsterdam: Benjamins, 2004), 13-85.
* ________, “The Coherence-Based Genealogical Method—What is it All About?”, online at http://www.uni-muenster.de/INTF/Genealogical_method.html (Accessed Feb. 13, 2009).
* Wachtel, Klaus, “Reconstructing the Initial Text of the Editio Critica Maior of the New Testament Using the Coherence-Based Genealogical Method,” oral presentation reported by Paul Foster, “Recent Developments and Future Directions in New Testament Textual Criticism: Report on a Conference at the University of Edinburgh,” JSNT 29 (2006), 229-235.
* ________, “Towards a Redefinition of External Criteria: The Role of Coherence in Assessing the Origin of Variants” in Textual Variation: Theological and Social Tendencies (TS 6; H. A. G. Houghton and D. C. Parker, eds.; Piscataway, N.J.: Gorgias, 2008), 109-127.
Response by Wachtel
"Thank you for the excellent summary of the CBGM. Just one remark: when we assess local genealogy of variants we take all available evidence into account, including intrinsic and transcriptional probability and, of course, pre-genealogical coherence and other external criteria applied with care. While first constructing local stemmata, however, we beware of preferring a variant just because it is attested by "good old manuscripts", thus confirming our own prejudices.
There is now online a comprehensive introduction by Gerd Mink at:
A useful documentation of some CBGM procedures is here:
Klaus Wachtel, INTF
Institut für Neutestamentliche Textforschung
Tuesday, December 08, 2009 5:45:00 AM