The 'theoretical virtues' theory : towards a joint solution to the meta-descriptive and meta-normative problems

  • Panagiotis Saranteas

Student thesis: Doctoral Thesis (MPhil)


There are at least four main problems that the 'theoretical virtues' theory needs to solve. Those are the descriptive, meta-descriptive, normative, and meta-normative problems. The first two problems are about virtue discovery and its methods, and the latter two are about virtue justification and its methods. The principles constituting a solution to the descriptive problem need to adequately describe the order of epistemic priority between the theoretical-explanatory virtues, and their combinations in various degrees, based on the considered judgments of expert natural and social scientists, and philosophers of science, on actual cases of explanatory-theory choice. The principles constituting a solution to the normative problem need to adequately describe the order of epistemic priority that should guide the judgments of both experts and non-experts in rational explanatory-theory choice in science, philosophy, and everyday reasoning. In this study it is argued that the two meta-level problems are prior to the two lower-level problems, and also, that they are interdependent and can be solved jointly. Such a joint solution is explored with the aim of eventually solving the two lower-level problems. After evaluating the most recent solutions that have been proposed to the descriptive and normative problems, the historical philosophy of science method and experimental philosophy of science method are also evaluated for their meta-descriptive and meta-normative potential. After their limitations are examined, it is argued that as part of any adequate solution to the two meta-level problems, the concept of a ‘theoretical-explanatory virtue’ would need to be conceptually re-engineered into, one, combinations of features of explanatory theories, and, two, a set of principles that describe each feature’s and each combination’s weight in an order of epistemic priority. Finally, the proposed solution to the meta-level problems is further developed by utilizing techniques from the methods of philosophical artificial intelligence and reflective equilibrium.
Date of Award2 Dec 2021
Original languageEnglish
Awarding Institution
  • University of St Andrews
SupervisorPeter Milne (Supervisor)

Access Status

  • Full text open

Cite this