Entries with tag cbgm .

Online Tools for the ECM

The ECM of Mark is currently being printed and will be available soon. Once it appears in print, we will make our online tools for it accessible. These will include the digital edition, the CBGM, a Docker container, and a list of textual changes compared to the NA28.


Even though the ECM of Acts was published back in 2017, we realized we have not posted the list of textual changes online for Acts like was done for the ECM of the Catholic Letters. So, in this post, I thought I would take the opportunity to review what is already available for the ECM of the Catholic Letters, update online resources for ECM Acts, and explain what will be available for the ECM Mark.


The INTF’s homepage lists a number of “Online Utilities”. These cover different topics but I’ve singled out the pertinent ones for the ECM and CBGM and have listed them here for convenience.

Image: ECM Volumes and available Digital Tools and Downloads


ECM IV: The Catholic Letters

Textual Changes

The ECM of the Catholic Letters (2nd edition, 2013) contained 33 textual changes compared to the NA27. These changes were adopted in the NA28 and are listed in the NA28 on pages 50*-51* and posted on the INTF’s website under “NA28” and Textual Changes. The first printing of the NA28 listed 34 textual changes, but the entry concerning the elision in ἀλλά in 1 Peter 2:25 was removed in later printings (since it is only orthographical); this resulted in 33 textual changes. Spellings were changed in a number of locations in the NA28. For a complete list see Orthographical standardization under “NA28” in “Online Utilities”.


Split Primary Lines

A split primary line occurs when the editors leave the decision open where two or more variants of about equal weight should be adopted in their reconstruction of the Ausgangstext. There are 43 split-line (diamond) readings in the ECM of the Catholic Letters, which were incorporated into the NA28. A list of diamond readings was posted under the “NA28” link under Split Primary Line in ECM2. The NA28 itself does not list these.



The Coherence-Based Genealogical Method (CBGM) offers tools for reconstructing the Ausgangstext in the ECM which is based on full transcriptions of witnesses. Decisions are based on textual criticism and philological study of all variants. The CBGM and its data for the Catholic Letters is available online here.



Textual Changes

Compared to the NA28, the ECM of Acts has 52 textual changes. For a list, see Textual Changes under “ECM Acts” in “Online Utilities”.


Split Guiding Lines

There are 155 split lines in ECM Acts. A list of these is found under Split Primary Line.



The CBGM for Acts is also available online here. Phase 4 of the CBGM for Acts uses the new interface designed by the Cologne Center for eHumanities.


Textual Commentary and Digital Edition

It’s important to note that all of the textual changes and split lines are discussed in the online textual commentary on the NTVMR, explained here. The “Text-Critical Commentary” gives concise reasons why one variant is favored over another (in the case of textual changes) or explains why the decision has been left open (in the case of split lines).

            This commentary has been integrated into the digital ECM (dECM). The dECM displays the text of ECM Acts (different from the NA28) and offers interactivity that is not possible in a printed edition. For example, the apparatus links to transcriptions and images of manuscripts, there is more versional data included than what was in the printed ECM, and every variant unit has a link to the specific passage in the CBGM.


Patristic Citations

There is also the online database of Patristic citations available here. What is innovative about this database is that the reader is not only given the specific work of the author cited but also the full context of the quote. Nikolai Kiel has described how the ECM treats Patristic citations.


Docker Container

The newest addition to ECM Acts is the Docker container, which is a downloadable package that enables you to run the CBGM for Acts on your own computer. Different from the online CBGM, the program enables you to make different textual decisions and reestablish the local stemmata to your own theories. Video instructions for the CBGM Acts Docker are found here, which also includes a short introduction to the CBGM.



After the ECM of the Gospel of Mark appears in print (26 July 2021), we’ll upload a list of textual changes and split guiding lines online. Like Acts, there will be an online textual commentary, a digital version on the NTVMR, the CBGM (with downloadable docker container), and the Patristic citations database.


Image: Advertisement of ECM Mark from German Bible Society


We hope these resources will guide readers to better understand the data behind the editions and can provide a solid starting place for further research to take place. Now that a Docker container is available for Acts, anyone can now experiment with the CBGM, which may be the best way to learn how the method works firsthand.

Download the CBGM Docker Container

There is now a docker container available for the Coherence-Based Genealogical Method (CBGM) for Acts. It can be downloaded here: https://hub.docker.com/r/scdh/intf-cbgm-db-server

Here is a tutorial on how to install it and a brief introduction to how the CBGM works: https://www.youtube.com/watch?v=k0_tlbz_YVQ



Mink, Gerd. “The Coherence-Based Genealogical Method (CBGM) – Introductory Presentation.” Release 1.0, 2009, egora.uni-muenster.de/intf/service/downloads_en.shtml.


---. “Problems of a Highly Contaminated Tradition: the New Testament. Stemmata of Variants as a Source of a Genealogy for Witnesses.” Studies in Stemmatology II, edited by Pieter van Reenen, August den Hollander, and Margot van Mulken, John Benjamins Publishing Company, 2004, pp. 13-85. Limited Google Books preview


Novum Testamentum Graecum: Editio Critica Maior, ed. the Institute for New Testament Textual Research. Volume III: The Acts of the Apostles, ed. Holger Strutwolf, Georg Gäbel, Annette Hüffmeier, Gerd Mink, and Klaus Wachtel. Stuttgart: Deutsche Bibelgesellschaft, 2017.


Wachtel, Klaus. “An Interactive Textual Commentary on Acts.” INTF Blog, https://ntvmr.uni-muenster.de/intfblog/-/blogs/an-interactive-textual-commentary-on-acts.

Remarks on Carlson, “A Bias at the Heart of the CBGM” (Guest post by Gerd Mink)

Just recently, Stephen Carlson’s article, “A Bias at the Heart of the Coherence-Based Genealogical Method (CBGM),” appeared in Journal of Biblical Literature. While we at the INTF read this with great interest, we were soon disappointed by the number of misunderstandings contained therein. Criticisms of the CBGM are always welcomed, and we are eager to incorporate suggestions for improving it. Unfortunately we were not able to use Carlson’s publication as fodder for making improvements because his article evinces a general lack of understanding of the method.


The suggestion to integrate “a common-error criterion within the mechanisms of the CBGM” shows that Carlson has not understood how the CBGM approach is fundamentally different from Lachmannian stemmatology, particularly with regard to errors. In a recent publication, Klaus Wachtel targets this exact point:


For the CBGM, coherence is the pre-eminent feature of the New Testament manuscript tradition for which it was developed in the first place. In Lachmannian methodology, common errors are used to trace genealogical structures. [...]

          In the context of the CBGM, all grammatically sound, or at least tolerable textual differences, which are not merely orthographical, are considered variants. An indicative error would have to be a variant in this sense to be genealogically useful because, as a rule, clerical errors were corrected, not copied, by the scribes. The CBGM abstains from identifying variants as errors, a principle that offers two advantages over against the common-error method: (1) we do not have to know at the outset, relying only on our philological acumen, which variants are errors and which are true renderings of the text in a pristine exemplar; and (2) we are not immediately confronted with the problem of contamination which admittedly [...] is the biggest problem for Lachmannian genealogy.

          Instead, we can make use of quantitative data regarding similarities and differences between witnesses, i.e. pre-genealogical coherence, to get an impression of the consistency of attestations. [...]

          The most important innovation brought about by the CBGM compared to Lachmannian methodology is the perception and description of genealogical relations. According to [the neo-Lachmannian scholar] Trovato, the relationship between any two manuscripts A and B can be assigned to one of three types, A>B, B>A, or A<x>B. For the textual tradition of the Greek New Testament, it would not be a reasonable goal to describe the relationship between any two manuscripts following this pattern. For any two New Testament witnesses A and B, i.e. states of text preserved in manuscripts, there is textual flow of the type A>B, which stands side by side with flow in the opposite direction (B>A), as well as A<x>B, due to contamination

(Wachtel, “The Development of the Coherence-Based Genealogical Method (CBGM),” 438-439, referring to Trovato, Lachmann’s Method, 57).


Gerd Mink, who first devised the CBGM, has been retired for some years now but has taken the opportunity to engage with the main points of Carlson’s discussion. Because he wanted to respond in a timely manner, he decided a blog post would be best. Therefore, below I present Mink’s remarks.


1. Invented Textual Genealogy: Carlson’s Scenario 2


In his article “A Bias at the Heart of the CBGM,” Carlson complains that the CBGM cannot figure out a simple scenario that he has invented himself. His scenario 2 has 18 variant passages with two variants each, which are assigned to five witnesses (including the initial text A); a simple case indeed. Carlson speaks of errors; I will use the term variants and assume that their agreements in the same place are not coincidental and rest upon high connectivity variants. Thus, both errors and variants are equally able to connect witnesses stemmatically.

Image: Figure 4. Carlson, 330


The stemma in fig. 4 shows Carlson's results, and it is not surprising that it corresponds exactly to the invented case. Carlson uses phylogenetic software to display such figures. The analysis behind them evaluates the agreement of the witnesses and their distance from each other, measured by the number of disagreements that separate them. To put it simply: agreements argue for belonging to the same branch; disagreements cause the branch to split.

Image: Chart of 18 Passages. Carlson, 330


The case seems well constructed to fit the resulting stemma. According to Carlson, the CBGM is not able to reconstruct this scenario and therefore must be rejected.


It is an unfortunate disadvantage for Carlson that his own scenario has such a small number of witnesses and variants because it gives the reader other options to come up with different stemmata than he did. A higher number would have been better since it would restrict the range of possible combinations in the global stemma. In the present simple case the 18 variant places correspond to 18 local stemmata of variants (‘a’ is the prior variant and ‘b’ is the posterior variant: a>b). These also allow for global stemmata of witnesses other than what are presented by Carlson in fig. 4 (see above). In addition to the four witnesses and the initial text A, Carlson invented two lost hypothetical witnesses, X and Y, that have left no traces in the local stemmata and are only visible to the inventor of the scenario.


Therefore, the text critic can proceed only from the variants at the 18 passages. It is not difficult to find stemmata that are compatible with all 18 local stemmata, for instance:

The stemmata have 5 nodes (= 4 witnesses and the reconstructed initial text A) and 6 edges (= arrows connecting stemmatic ancestors and descendants) each. These stemmata are more parsimonious than Carlson's fig. 4 because they do not require hypothetical witnesses.

It may be helpful to give an explanation about the left stemma. Here are what the arrows represent:

  • Arrow A > B: B agrees with A at the variant places 2-18. At place 1, the variant of B is derived from A. All 18 places are agreements or are prior variants in A.
  • Arrow A > E: E agrees with A at the variant places 1, 6-11, 13-16. E is derived from A at the variant places 2-5, 12, 17-18. All 18 places are agreements or are prior variants in A.
  • Arrows pointing to C: at variant place 1, C agrees with A and E. At variant places 2-5, C agrees with E. At variant places 6-11, C is derivable from A or E. At variant places 12, 17-18, C agrees with A. At variant places 13-16, C agrees with A and E. All 18 places are agreements or are prior variants in A and/or E.
  • Arrows pointing to D: at variant place 1, D agrees with A and E. At variant places 2-5, D agrees with E. At variant places 6-11, D agrees with A and E. At variant place 12, D agrees with E. At variant places 13-16, D is derivable from A or E. At variant places 17-18, D agrees with A. All 18 places are agreements or are prior variants in A and /or E.

It is a little irritating that Carlson always speaks of “witnesses”, but also calls his invented witnesses “manuscript copies” and uses the formulation “stemma codicum” (p. 326; p. 336 referring to real witnesses), which, however, he contradicts on the same page (cf. note 49). In fact, Carlson has developed two scenarios of manuscripts that are copied from each other and whose copyists make errors (on implications of the distinction between witness and manuscript cf. below and Mink, “Manuscripts, Texts, History, and the Coherence-Based Genealogical Method,” 281-283).


The CBGM deals with witnesses, not manuscripts. In Carlson's constructed case, manuscripts and witnesses are the same thing, a fact that cannot be deduced from the data. The data does not reveal anything about the completeness of the tradition, i.e. whether witnesses are directly or indirectly related. The latter is the normal case in the NT tradition and renders a stemma codicum impossible. In the CBGM, a global stemma displays a structure of the data according to specific rules; it does not immediately display the actual history of transmission—a phylogenetic stemma does nothing else, by the way.


The alternative stemmata above contain contamination. In a global CBGM stemma, which can show only the preserved tradition and does not include lost links, several arrows pointing to a witness do not mean that contamination has necessarily taken place in that witness. Still, it may appear there as a result of contamination in lost predecessor witnesses. (On contamination as a process and as a result of a loss of witnesses cf. Mink “Introductory Presentation,” 58-63; cf. also the section “How to Understand a Global Stemma” in Mink, “Manuscripts, Texts, History, and the Coherence-Based Genealogical Method,” 284-287.)


In other stemmatic methods, it may be that hyparchetypes represent lost predecessor witnesses. Let us take the left alternative stemma again. Arrows point from E and A to D. E and D show many agreements. At some places, however, we read in E an older variant, in other places vice versa. That is what the local stemmata indicate. The reason for this situation can be that the transmission is split or is contaminated. The data do not tell us anything about it. Maybe a lost hyparchetype caused the textual state in D. Therefore, the left alternative stemma could be compatible with Carlson's stemma. We cannot know how many hyparchetypes and where in the stemma they should be assumed, especially in a more complex situation. (On the needlessness of hyparchetypes, see Mink, “Manuscripts, Texts, History, and the Coherence-Based Genealogical Method,” 289; Mink, “Problems of a Highly Contaminated Tradition,” 48, 59-67; Paulson, “Improving the CBGM,” 301f.)


In his essay, Carlson takes scenario 1 (10 variant passages) as his starting point. The only difference in scenario 2 (18 variant passages) discussed here is that C and D have additional variants compared to scenario 1. Yet, according to Carlson's invented scenario, the stemma should remain the same—except for the greater distances of C and D from A. In order to test how the CBGM processes his hypothetical scenario, Carlson has chosen the completely wrong approach; he uses potential ancestors and textual flow diagrams instead of stemmatic coherence. The above mentioned alternative stemmatic possibilities, based on scenario 2, would also be compatible with scenario 1 since its data is only a subset of the data of scenario 2. Carlson's example does not demonstrate what he wanted it to demonstrate.


The essential point is: the CBGM does not claim to reconstruct the historical sequence of copying activities. Phylogenetic methods do not achieve this either (cf. Bordalejo, “Genealogy of Texts”). Also Carlson has not reconstructed but rather invented a copying scenario following what his method is able to represent. He must introduce hypothetical witnesses X and Y because his graphic (bifurcating and allowing terminal nodes only for non-hypothetical witnesses) requires this. Yet, X and Y left no definite traces in the data, so in a parsimonious graph there is no need to assume them. And what has left behind no traces, cannot be found. The data is not unambiguous.


As we see, we can invent still other copy scenarios for the same local stemmata. We can declare that one of them (including contamination) is the actual one, and—as it appears—Carlson's method used for fig. 4 would not find it, despite its simplicity. Would Carlson then make the same judgment about the method he applies as he did about the CBGM?


2. Epistemological Premises


Here is Carlson's verdict on the CBGM:


“If a method cannot handle this simple case correctly, it should be rejected or fixed so that it can” (p. 325).


“If a method is misled in the simplest of cases due to some bias, how confident can one be that it will work in the more complicated cases?” (p. 335).


We should bear in mind that no method, not even computer-aided, can reconstruct historical events in detail (see again Bordalejo). This is even more obvious when elements (here the hypothetical witnesses) are hidden, as is the case in Carlson’s chart of 18 passages.


In principle, where we have positive knowledge, we do not need hypotheses. If we know some details of the copying history, we must apply this knowledge in the CBGM or any other method. Normally, we know only the variants. As for the stemma, many possibilities may arise. The witnesses in Carlson's fig. 4 have their place due to the introduction of hypothetical witnesses and because Carlson knows the copying history, as he invented it himself. Again, it is easy to invent another copying history with the same texts, but it would not be represented by the method used for fig. 4.


3. Key Terms and Concepts of the CBGM Approach Misunderstood


The core problem of Carlson’s article is that he does not seem to understand the overall concept of the CBGM. Although he refers to CBGM terms, which are explained in almost all publications on the CBGM, he does not understand them according to their rigid and precise definitions; these definitions correspond to rules on how the associated values are obtained and the defined terms do not allow for connotative interpretations. Moreover, he does not observe the intentions and claims of modules contained in the Genealogical Queries.


3.1 Textual Flow Diagrams


One of Carlson’s key misunderstandings is that he reads textual flow diagrams as if they were genealogical representations of actual textual history. He does not follow the definitions of potential ancestor and textual flow diagram (even though he cites them!). The potential ancestors are, of course, hardly ever the actual ancestors of any manuscript. They are also not proxies for which there would be some text to reconstruct. They do not represent anything but themselves. The textual flow diagrams are not stemmata (Gurry wrote explicitly on this point in “The Harklean Syriac” p. 198). Nor does the coherence of an attestation determine whether a reading is the initial text.


It is crucial in the CBGM not to confuse genealogical coherence with stemmatical coherence nor potential ancestors with stemmatic ancestors (or even actual historical ancestors of manuscripts). Only a few of the potential ancestors have a chance to become stemmatic ancestors in a substemma of the descendant in question, even if they have the highest ranking numbers. On the other hand, witnesses which are not potential ancestors can become stemmatic ancestors in an optimal substemma.


3.2. Relatives Tables


To interpret textual flow diagrams correctly, it is important to know that they are only simplifying graphic representations of the data; they must be viewed with the Relatives table in the new interface for Acts, phase 4 (or in “Show Tables” option for the Catholic Letters) for a better understanding of the genealogical scenario.


For an example in a new interface which has not yet been transferred to the Catholic Letters, go to Coherence and Textual Flow in Genealogical Queries Acts (phase 4).

Image: Genealogical Queries for Acts


Next, click on any manuscript and see the result.



Example of Relatives Table for 03 in Acts


The column with the percentages shows pre-genealogical coherence, that is, the textual agreement between two witnesses which does not change no matter how many prior variants a witness has. The columns W1<W2 and W1>W2 show the results of the construction of the local stemmata between two witnesses: in the example above, 01 has 192 prior variants to 03, and 03 has 328 prior variants to 01. Using textual flow diagrams without consulting the data behind them (i.e. the Relatives tables) can be misleading. (See also Wachtel on this: https://ntvmr.uni-muenster.de/intfblog/-/blogs/an-interactive-textual-commentary-on-acts; and on how to interpret the listings of potential ancestors, see especially Mink, “Introductory Presentation,” 255-297.)


3.3 Connectivity


Another fundamental problem is that Carlson does not seem to comprehend the purpose of the connectivity option in Genealogical Queries (cf. Carlson, 325, 334, and passim).

Image: Option to select connectivity in Genealogical Queries


This option enables users to test the stability of the resulting diagrams by setting different values. In many cases, the values of 10 (in the Catholic Letters) or 5 (in Acts) are only reasonable starting points. Instability will raise doubts if high connectivity has been assumed on internal grounds. No definitive statements about connectivity are made regarding textual flow diagrams. There is no right or wrong value. The inserted value is based on the user’s (preliminary) assessment. Different areas in a textual flow diagram may even require different connectivity assumptions (depending on closer or more distant relationship of witnesses in an area). Definitive statements are required during the construction of substemmata to decide on whether possible stemmatic connections are necessary. (Cf. Mink, “Introductory Presentation,” 529-537.)


4. Essential Methodological Procedures Neglected


In short: Carlson has used the tool “Coherence in Attestations” for something for which it was not designed. Instead, he should have dealt with stemmatic coherence, the formation of optimal substemmata, and the global stemma; only there do hypotheses about the stemmatic structure of the tradition develop. Carlson, however, neglects this and other major parts of the methodology he criticizes.


4.1 The Text Is the Witness, the Manuscript its Carrier


It is also unclear whether Carlson understands that in a contaminated tradition almost every witness, even a potential ancestor, has a proportion of both older and younger variants compared to any close relative. In his section that deals with 1 Jn 1:7 (scenario 3), Carlson writes (p. 336),

This situation not only resembles that of scenario 2, but it inspired it. In both cases, the potential ancestor bias manifests itself against witnesses that branched off early from the predominant textual flow but acquired a large number of secondary and singular readings of their own. The texts that correspond to scenario 2's C and D in the textual transmission of 1 John are 01 and 02.

Carlson correctly identifies the reasons that lead to greater distances from A than we see in 1739. His phylogenetic software displays these distances, too (for the place of 01 in light of CBGM data, see Mink, “Introductory Presentation,” 270-295, especially 290.). As for scenario 2, we do not know the kind of variants which produce the distances there.


To take 01, 02, and 1739, which are cited by Carlson, as examples: in the Catholic Letters. The text of 1739 does not have a genealogically older text than 01 or 02 in every place but in the majority of places, i.e. it has ancestor variants more often than vice versa. Only these places could offer a connection pointing from a stemmatic ancestor to a descendant in the global stemma (i.e. not the places where the ancestor reads text posterior to the descendant). And of course, there are places where 1739 has text posterior to 01 or 02. Carlson claims there is “a bias against texts on old lineages like 01 and 02, and a bias in favor of stemmatically later texts like 1739 whose copying is more strictly controlled. The net effect of this bias is to overvalue the witness of 1739 at the expense of 01 and 02” (p. 337).


On which basis are texts considered “stemmatically later”? Is an early textual error more valuable than textual accuracy documented in a later manuscript? It is neither bias nor contradiction to identify older variants in younger manuscripts. At the many variant places where the witness 01 reads a posterior variant, it cannot be an ancestor of the witness 1739. Carlson jettisons the important distinction between the age of a manuscript and the age of its text as represented in the majority of its variants. In this context again, it is of great significance not to confuse potential ancestors with stemmatic ancestors, nor textual flow diagrams with global stemmata, and a stemma of witnesses (=texts) must not be read in the same way as a stemma codicum.


4.2 The CBGM Is a Tool, not a Decision-Maker


At the end of his essay, we find the following curious statement: “The potential-ancestor formula favors certain witnesses at the expense of others in certain genealogical configurations that it cannot detect a priori” (p. 339). The role of potential ancestors is not to detect genealogical configurations. A potential ancestor is a witness with more prior variants than the witness being compared. Genealogical configurations are offered in optimal substemmata. The role a given potential ancestor will play there cannot be read from lists of potential ancestors or textual flow diagrams.


Carlson continues: “For 1 John, it appears that the CBGM favors 1739 at the expense of Codex Sinaiticus and Codex Alexandrinus, and the full extent of the bias is not apparent. It has probably made the omission of δέ in 1 John 1:7 more viable than it really is.” Apart from the fact that there is no bias, the textual non-decision (i.e. split primary line) in 1 Jn 1:7 is not required by the CBGM as such. The CBGM provides tools and does not make textual decisions. Text-critical decisions like the one in 1Jn 1:7/3 are never enforced by some algorithm or automatism of the CBGM. These decisions are made by the editors. Anyone who has familiarized themselves with a basic introduction like Wasserman and Gurry’s A New Approach to Textual Criticism would know this.


5. Conclusion


In closing, Carlson's assumption that there is bias in the CBGM is the result of misunderstandings and wrong application. Any method will have advantages and disadvantages as well. Proper understanding of its possibilities and limitations is crucial. Different methodological approaches to textual criticism are highly welcome as is informed and substantiated criticism. Overall, however, it appears that Carlson’s latest article has only very selectively grappled with literature which explains the CBGM and has not grasped the overall concept of the method and its key components. Wachtel's essay, The Development of the Coherence-Based Genealogical Method (CBGM), its Place in Textual Scholarship, and Digital Editing, would serve here as a useful corrective to understanding that the CBGM is deliberately non-Lachmannian.

The problem is not that Carlson clearly favors a different method than the CBGM. In fact, Edmondson's 2018 Ph.D. thesis demonstrates that an analysis of the CBGM is also possible from a phylogenetic perspective (Edmondson, An Analysis of the Coherence-Based Genealogical Method Using Phylogenetics). In only selectively and superficially engaging with the relevant literature, Carlson’s article has unfortunately cultivated a new series of misunderstandings about the CBGM and its functionalities, most of which could have been avoided had he just made use of my entry level “Introductory Presentation.”

It is hoped that my blogpost encourages interested researchers to form their own opinion based on the relevant literature on the CBGM. The aforementioned “Introductory Presentation” may be a good start.


Works Cited


Bordalejo, Barbara. “The Genealogy of Texts: Manuscript Traditions and Textual Traditions.” Digital Scholarship in the Humanities, vol. 31, no. 3, 2016, pp. 563-577. Links to publisher and Academia


Carlson, Stephen C. “A Bias at the Heart of the Coherence-Based Genealogical Method (CBGM).” Journal of Biblical Literature, vol. 139, no. 2, 2020, pp. 319-340.


Edmondson, Andrew Charles. An Analysis of the Coherence-Based Genealogical Method Using Phylogenetics. 2019. University of Birmingham, PhD dissertation. https://etheses.bham.ac.uk/id/eprint/9150/


Gurry, Peter J. “The Harklean Syriac and the Development of the Byzantine Text: A Historical Test for the Coherence-Based Genealogical Method (CBGM).” Novum Testamentum, vol. 60, 2020, pp. 183-200. Links to publisher and Academia


Mink, Gerd. “The Coherence-Based Genealogical Method (CBGM) – Introductory Presentation.” Release 1.0, 2009, egora.uni-muenster.de/intf/service/downloads_en.shtml.


---. “Manuscripts, Texts, History, and the Coherence-Based Genealogical Method (CBGM): Some Thoughts and Clarifications.” The New Testament in Antiquity and Byzantium: Traditional and Digital Approached to its Texts and Editing. A Festschrift for Klaus Wachtel, edited by H.A.G. Houghton, David C. Parker, and Holger Strutwolf, De Gruyter, 2019, pp. 281-293. Link to publisher


---. “Problems of a Highly Contaminated Tradition: the New Testament. Stemmata of Variants as a Source of a Genealogy for Witnesses.” Studies in Stemmatology II, edited by Pieter van Reenen, August den Hollander, and Margot van Mulken, John Benjamins Publishing Company, 2004, pp. 13-85. Limited Google Books preview


Paulson, Gregory S. “Improving the CBGM: Recent Interactions.” The New Testament in Antiquity and Byzantium: Traditional and Digital Approached to its Texts and Editing. A Festschrift for Klaus Wachtel, edited by H.A.G. Houghton, David C. Parker, and Holger Strutwolf, De Gruyter, 2019, pp. 295-307. Links to publisher and Academia


Trovato, Paolo. Everything you always Wanted to Know about Lachmann’s Method. 2nd ed., Libreriauniversitaria.it, 2017.


Wachtel, Klaus. “The Development of the Coherence-Based Genealogical Method (CBGM), its Place in Textual Scholarship, and Digital Editing.” The Future of New Testament Textual Scholarship, edited by Garrick Allen, Mohr-Siebeck, 2019, pp. 435-446. Links to publisher and Academia


---. “An Interactive Textual Commentary on Acts.” INTF Blog, https://ntvmr.uni-muenster.de/intfblog/-/blogs/an-interactive-textual-commentary-on-acts.


Wasserman, Tommy and Peter Gurry. A New Approach to Textual Criticism: An Introduction to the Coherence-Based Genealogical Method. SBL, 2017.


Recent Bloggers Recent Bloggers

Greg Paulson
Posts: 17
Stars: 42
Date: 10/06/21
marie-luise lakmann
Posts: 2
Stars: 8
Date: 23/04/20
Darius Mueller
Posts: 1
Stars: 0
Date: 19/03/20
Theodora Panella
Posts: 2
Stars: 4
Date: 09/03/20
Jan Graefe
Posts: 2
Stars: 0
Date: 18/09/19
Siegfried G. Richter
Posts: 1
Stars: 3
Date: 01/07/19
Nikolai Kiel
Posts: 1
Stars: 4
Date: 06/06/19
Klaus Wachtel
Posts: 1
Stars: 1
Date: 14/09/18