The Ship of Theseus: Using mathematical and computational models for predicting identity judgments


Reasoning processes have been one of the central targets for cognitive modeling. Modeling of reasoning processes appears as an even harder challenge during paradoxical conditions such as the Ship of Theseus paradox. This work attempts to model empirical data from a behavioral study on paradox resolution with different modeling techniques: discriminant analysis (DA), decision tree and neural networks. While each method has its own advantages and disadvantages, this paper attempts to compare and to contrast these methods trying to select the best model for future work. Identity judgments have long been at the center of philosophical debates, e.g., is a car still the same after being fixed after a serious accident? Beyond the philosophical debates on the nature of objects and the concept of identity, it has also been a matter of interest how laymen respond to the identity question under different circumstances. This empirical study focuses on a famous paradox from ancient Greece, the Ship of Theseus. Answers to this paradox are tried to be predicted by a Conceptual Tendency Test (CTT) tapping the concept of “sameness”. The main aim of this paper is to present and compare the results of three predictive models in terms of their accuracy (predictive success) and to discuss the theoretical basis of the findings.