Non-monotonic epistemic aspects of scientific explanations

Yao-Hua Tan


In this paper we show that explanations based on incomplete information do not always comply with Hempel's Covering Law Model of scientific
explanation. We show that the relevant covering law is usually not known beforehand in this type of explanations. This results in a breakdown of the symmetry between prediction and explanation in Hempel's DN-model in the case of explanations based on incomplete information. We argue that in this type of explanations it is possible to derive from the observed facts a weaker type of law, which we call unspecific laws, which are strong enough for explanation, but too weak for prediction. Furthermore, we argue that this derivation of unspecific laws from observed facts presupposes a new type of arguments, which we called Law-Finding-From-Facts or L3F Arguments, which are supplementary to Hempel's covering law model. These L3F arguments cannot be mod- elled in classical logic, and should not be considered as inductive arguments either. We show that Shoham's non-monotonic epistemic logic is the best logic to model these L3F arguments.

Full Text:



  • There are currently no refbacks.