'Laws of Nature: Necessity vs. Regularity'

Core Idea

A law of nature is an exceptionally stable, universal regularity governing how things behave. But are laws merely regularities describing what happens, or do they express metaphysical necessity determining what must happen? This question profoundly affects theories of causation, explanation, and the unity of science through reduction.

Explainer

Your study of laws of nature established what we want from them: universality, support for counterfactuals ("if this were copper, it would conduct"), explanatory power, and a distinction from mere accidental regularities. "All copper conducts electricity" feels different from "all the coins in my pocket are silver" — even if both happen to be true. Your work on necessity and contingency gave you the modal vocabulary: necessary truths hold in all possible worlds; contingent truths hold in some but not others. The debate about laws asks which of these categories laws belong to.

The Humean Regularity view denies that laws are anything over and above stable patterns. Laws are the generalizations that describe what actually, universally happens — nothing more is needed or available. David Hume argued that when we observe regularities, we add the idea of necessity from our own psychological habits, not from the world itself. The most sophisticated Humean account is the Best Systems Analysis (Mill-Ramsey-Lewis): laws are the axioms of the simplest, strongest, most unified systematization of all particular facts. "All copper conducts" is a law because it appears in the best theory; "all coins in my pocket are silver" does not. There is no hidden necessity in the world — just patterns and the best descriptions of them.

Non-Humean accounts insist that this picture leaves something out. The necessitation view (Armstrong, Dretske, Tooley) holds that laws express genuine metaphysical necessities: relations of nomic necessitation between universals. When "all F's are G" is a law, it is because the universal F stands in a necessitation relation to the universal G — not merely because all the F-instances happen to be G-instances. This extra metaphysical ingredient explains why laws support counterfactuals: it is not just that every actual piece of copper conducts, but that the property of being copper necessitates the property of being conductive.

What is at stake goes beyond academic metaphysics. If laws are mere regularities, scientific explanation becomes a puzzle: regularities describe patterns, but do they explain them? Saying "the copper conducts because all copper conducts" seems circular. Non-Humeans argue that genuine explanation requires necessitation — the law tells you why the particular event had to happen. The debate also bears on induction: if laws are merely the best systematization of past facts, what justifies projecting them onto unobserved cases? And it connects to intertheoretic reduction: if chemistry reduces to physics, is that reduction merely the discovery of correlating regularities, or the identification of genuine necessities that make the upper-level laws hold with metaphysical force?