Deductive-Nomological Model
The deductive-nomological (DN) model holds that to explain a phenomenon is to deduce a statement of it from premises that include at least one law of nature.
Definition
On the DN model an explanation is a sound deductive argument whose conclusion is the explanandum and whose premises (the explanans) contain essentially at least one general law together with statements of antecedent conditions, all of which must be true.
Scope
This topic covers Hempel and Oppenheim's covering-law account of explanation, its adequacy conditions, the symmetry it posits between explanation and prediction, and the classic counterexamples (irrelevance, asymmetry, explanation by effects) that motivated later causal and pragmatic accounts.
Core questions
- What conditions must an argument meet to count as a DN explanation?
- Does the DN model wrongly treat explanation and prediction as structurally identical?
- Why do counterexamples like the flagpole and the barometer undermine the model?
- Can the model accommodate explanations that do not cite exceptionless laws?
Key concepts
- explanans and explanandum
- law of nature
- nomic expectability
- explanatory asymmetry
- explanatory irrelevance
Key theories
- Covering-law thesis
- Every adequate scientific explanation subsumes its explanandum under general laws, so explanation is fundamentally a matter of nomic expectability.
- Symmetry thesis
- Hempel holds that explanation and prediction have the same logical structure, differing only in whether the explanandum is known before or after the inference.
History
Introduced in Hempel and Oppenheim's 1948 paper and elaborated in Hempel's 1965 essays, the DN model became the orthodox account of explanation. Wesley Salmon's 1989 retrospective documents how counterexamples concerning relevance and asymmetry led to its decline in favour of causal accounts.
Debates
- The asymmetry problem
- From a flagpole's height one can deduce its shadow's length and conversely, yet only the former explains; the DN model, being purely logical, cannot capture this directional asymmetry.
Key figures
- Carl Hempel
- Paul Oppenheim
- Wesley Salmon
Related topics
Seminal works
- hempeloppenheim1948
- hempel1965
Frequently asked questions
- What is the barometer counterexample?
- A falling barometer reading and an approaching storm are both effects of a drop in atmospheric pressure. The DN model would let one deduce and seemingly 'explain' the storm from the barometer reading, which is clearly not a genuine explanation, showing the model needs a notion of causal relevance.