A theoretical virtue in science is a property of a scientific theory that is considered desirable. Standard theoretical virtues include testability, empirical accuracy, simplicity, unification, consistency, coherence, and fertility. First highlighted by Thomas S. Kuhn in a seminal paper in 1977, theoretical virtues have come to play an important role in a number of philosophical debates. A central bone of contention in many of these debates is whether theoretical virtues are epistemic, i.e., whether they are indicative of a theory’s correctness, or whether they are just pragmatic, concerning only the convenient use of a theory. Particularly contested virtues are simplicity and unifying power. In the scientific realism debate, in which philosophers argue about whether or not scientific theories allow us to uncover the reality behind the phenomena, scientific realists have argued that virtuous theories are more likely to be correct than a less virtuous ones, even when they accommodate the same data. In the closely related debate about the so-called Inference to the Best Explanation, realists have argued that not only can we determine the best explanation on the basis of its virtues, but we can also determine which explanation is the true one. In discussions about “theory choice” or “theory appraisal,” philosophers discuss which virtues might be most decisive in scientists’ deliberations about which theory they should adopt. Here a theory’s successful novel predictions, or novel successes for short, have been a particular focus. Philosophers have also discussed possible trade-offs between various virtues and the difficulties which these may pose for theory choice. Samir Okasha has argued recently that there cannot be any rational algorithm for theory choice. Theoretical virtues also play a role in philosophical accounts of the laws of nature. One extremely prominent account, namely David Lewis’ Best System Analysis, appeals to simplicity and unifying power to determine what generalizations qualify as genuine laws of nature (rather than just accidentally true generalizations). Even in philosophical theorizing about science, theoretical virtues have been appealed to: Rudolf Carnap believed that simplicity and fruitfulness were important desiderata guiding the explication of scientific concepts. Finally, psychologists have started to investigate the role of theoretical virtues in picking explanations. There is work that appears to show that children and adults have preferences for simple and broad theories.
Knowledge-based Algorithms for BDI-agents
