Textual problems with mathematical assessments of relationship networks in medieval literature (or whythe numbers don’t add up in Táin Bó Cúailnge)
The purpose of this paper is to highlight some methodological problems that arise during attempts to apply statistical mechanical tools to the analysis of character networks within medieval literature, with particular reference to one attempt undertaken in a recent paper by Pádraig Mac Carron and Ralph Kenna. In that study Mac Carron and Kenna applied statistical mechanical tools to the study of relationship networks between characters in the Iliad, Beowulf and Táin Bó Cúailnge (treating those of each text separately). Character relationships were divided into hostile and friendly and it was suggested that the perceived artificiality of the Irish narrative can be traced back to anomalous features associated with six characters. Speculating that these are amalgams of several entities or proxies, renders the plausibility of the Irish text comparable to the others from a network-theoretic point of view. I do not claim to be a mathematician nor competent enough to challenge most aspects of their statistical analysis, but rather wish to suggest that a fuller appreciation of the nuances of the literary material upon which they based their analysis exposes underlying flaws within the definition of part of their subject area and in the enumeration of part of their data set. This paper will focus on the example of the medieval Irish narrative about a war between the kingdoms of Connacht and Ulster, known as Táin Bó Cúailnge (‘The Cattle Raid of Cooley’— henceforth Táin), studied by Mac Carron and Kenna, but some of the arguments are applicable to the other narratives in their study (particularly the Iliad) and more broadly, to statistical analysis of any literary text or corpus of texts.