Deductive-nomological model

The deductive-nomological model (DN model) of scientific explanation, also known as Hempel's model, the Hempel–Oppenheim model, the Popper–Hempel model, or the covering law model, is a formal view of scientifically answering questions asking, "Why...?". The DN model poses scientific explanation as a deductive structure, one where truth of its premises entails truth of its conclusion, hinged on accurate prediction or postdiction of the phenomenon to be explained. Because of problems concerning humans' ability to define, discover, and know causality, this was omitted in initial formulations of the DN model. Causality was thought to be incidentally approximated by realistic selection of premises that derive the phenomenon of interest from observed starting conditions plus general laws. Still, the DN model formally permitted causally irrelevant factors. Also, derivability from observations and laws sometimes yielded absurd answers. When logical empiricism fell out of favor in the 1960s, the DN model was widely seen as a flawed or greatly incomplete model of scientific explanation. Nonetheless, it remained an idealized version of scientific explanation, and one that was rather accurate when applied to modern physics. In the early 1980s, a revision to the DN model emphasized maximal specificity for relevance of the conditions and axioms stated. Together with Hempel's inductive-statistical model, the DN model forms scientific explanation's covering law model, which is also termed, from critical angle, subsumption theory. The term deductive distinguishes the DN model's intended determinism from the probabilism of inductive inferences. The term nomological is derived from the Greek word νόμος or nomos, meaning "law". The DN model holds to a view of scientific explanation whose conditions of adequacy (CA)—semiformal but stated classically—are derivability (CA1), lawlikeness (CA2), empirical content (CA3), and truth (CA4). In the DN model, a law axiomatizes an unrestricted generalization from antecedent A to consequent B by conditional proposition—If A, then B—and has empirical content testable.