An open problem in computational stemmatology - a model for contamination

Abstract

In this contribution, two open problems in computational stemmatology are being considered. The first one is contamination, an umbrella term referring to all phenomena of admixture of text variants resulting from scribes considering more than one manuscript or even memory when copying a text. This problem is one of the biggest to date in stemmatology since it implies an entirely different formal approach to the reconstruction of the copy history of a tradition and in turn to the reconstruction of an urtext. (Maas 1937) famously stated that there is no remedy against contamination and (Pasquali and Pieraccioni 1952) coined the terms 'open' vs. 'closed' recensions to distinguish contaminated from uncontaminated. We present a graph theoretical model which formally accommodates traditions with any degree of contamination while maintaining a temporal ordering and give combinatorial numbers and formula on the implication for numbers of possible scenarios.