Sunday, July 27, 2025

H. A. G. Houghton on the Coherence-Based Genealogical Method (CBGM)

  

Genealogical Coherence

 

One of the problems facing editors of the New Testament is the phenomenon of mixture (sometimes called ‘contamination’) from multiple sources within the text of a single manuscript. This makes it very hard, if not impossible, to construct a stemma, or family tree, of how manuscripts relate to each other. In other textual traditions, this is often done based on ‘shared errors’ which demonstrate the dependence of certain witnesses on others and thereby indicate that their text is likely to be secondary. The adoption of digital editing processes, however, has led to new methods of examining relationships within the textual tradition. Chief among these is the Coherence-Based Genealogical Method (CBGM), used in the production of the ECM. A fundamental difference between this method and traditional stemmatics is the distinction between the manuscript and the text it carries. In the framework of the CBGM, the text is the witness, not the manuscript with its codicological and paleographical properties. While traditional stemmatics focuses on common errors, the CBGM addresses the issue of textual contamination by considering different types of coherence. The first measure is the percentage agreement of the text of any two witnesses included in the edition (their ‘pre-genealogical coherence’). By creating a ‘local stemma’ of how readings relate to each other in each variation unit, it is then possible to compare the relationship of witnesses wherever they differ, expressing it in terms of the proportion of ‘prior’ (earlier) and ‘posterior’ (later) readings. Each local stemma is constructed by the editors using traditional philological principles to establish how the readings are related in each unit, largely based on internal criteria. The ability of the computer to process the relationships of witnesses based on each place of variation provides editors with an indication of the closest potential ancestors for each witness and, beyond that, the textual shape of the tradition as a whole. The measure of ‘genealogical coherence’ is the extent to which each variant reading is attested in witnesses which are related overall (‘coherent attestation’) or whether it emerged on multiple, unrelated occasions (‘incoherent attestation’). Beyond this is the concept of ‘stemmatic coherence,’ which expresses the extent to which the text of each witness can be explained by the smallest possible combination of its potential ancestors in an optimal substemma. Combining all substemmata creates a ‘global stemma.’

 

This genealogical approach is used alongside the traditional criteria by the editors of the ECM when constructing the editorial text (described as the Ausgangstext or ‘Initial Text’). It thus constitutes a novel means of enhancing reasoned eclecticism by adding information which can only be provided by computer processing. Like the understanding of the textual quality of individual documents, the CBGM data is built up in an iterative ways, beginning with straightforward variation units and then using the genealogical relationships built up from these to consider more complex cases, it is worth underlining that the CBGM does not itself determine the earliest reading: rather, it provides analyses of the consistency of the textual tradition which may assist editors in identifying the earliest reading, especially in cases where other criteria are evenly balanced. In addition, the CBGM sheds new light on the development of the biblical text by indicating the extent to which particular readings are likely to be related or may have arisen independently. (H. A. G. Houghton, A Textual Commentary on the Greek New Testament: A Companion to the Sixth Edition of the United Bible Societies’ Greek New Testament [Stuttgart: Deutsche Bibelgesellschaft, 2025], 21*-22*)