On the Weighing and Counting of Variants: The Coherence Based Genealogical Method, Potential Ancestors, and Statistical Significance

Though the theory behind the Coherence Based Genealogical Method has been published a number of years ago, as yet it has received little discussion and evaluation within the wider scholarly forum. This paper will discuss some of the strengths and limitations of the method and suggest some improvements as to its flexibility. It will do this by concentrating on one of the fundamental elements in the CBGM, that of the Potential Ancestor. Part 1: When is a 'Manuscript' a Potential Ancestor? At the moment in the CBGM every 'manuscript' that contains more earlier than later readings than a second 'manuscript' is considered a potential ancestor. The ratio between earlier and later readings is established by counting all readings for which a decision is made. However, it can be shown that in many cases this leads to statistically insignificant outcomes. Part 2: Which Variants Make the Difference? As an example, the differences between minuscules 33 and 2344 in the Catholic Epistles are discussed and it is shown that the relative positioning of these two manuscripts might be different than the CBGM currently suggests. Part 3: Weighing and Counting. It will be suggested that full-scale comparisons between manuscripts overestimate the role of contamination and underestimate that of independent emergence of variants. The CBGM could be used to identify variants (and therefore classes of variants) which are the result of independent emergence, and has the potential to construct approximations of the textual flow that are not, as presently is the case, the result of a single, fixed hypothesis on the nature of variants, but are the result of a variety of ways of counting the relevant and weighed data.