• In his introduction to The New Testament in the Original Greek, F. J. A. Hort laid out a taxonomy of evidence for text-critical judgments that is still followed today. In addition to the external evidence pertaining to the textual affinities, dates, and provenances of manuscripts, versions, and patristic quotations, he divided the internal evidence of readings into two categories: intrinsic probabilities, concerning what the author was most likely to have written, and transcriptional probabilities, concerning how emendators were most likely to change this. Hort’s description of internal evidence in terms of probabilities, while mostly figurative, turned out to be prophetic: after a century and half of development (largely outside of New Testament textual criticism), cutting-edge genealogical methods work in the realm of probabilities and approach the task of reconstructing a textual history as one of testing different hypotheses in the presence of unknown variables.

    In this talk, I show how Bayesian phylogenetics, a technique for inferring family trees in evolutionary biology, aligns nicely with Hort’s taxonomy of text-critical evidence and offers a clear separation of concerns between the parts of this taxonomy that rely on human judgments and those that can be automated by the computer. Using worked examples from my dissertation on Ephesians, I will demonstrate how assessments of internal evidence can be input into this process. Then, using toy examples from the UBS collation data for different New Testament corpora (Gospels, Praxapostolos, Pauline Epistles, and Revelation), I will demonstrate how stemmata relating the external evidence are processed and output by this process.