||Scientists have discovered an impressive number of laws of Nature. One example is Boyle\'s law, which states that the pressure of a gas varies inversely with its volume at a constant temperature: so the smaller the volume of a gas the greater its pressure, given that its temperature remains the same.
Philosophers of science have attempted to explain just what laws of Nature are. Some have held that laws of Nature are merely true generalizations, stating that whenever an event of one sort occurs, an event of another sort also occursâ€”that whenever an F event occurs, a G event also occurs. One problem with this is that not all laws of Nature state that all F events are accompanied by G events. Some laws, such as some laws of physics, are merely probabilistic and so can only be represented as stating that, for instance, 75% of F events are accompanied by G events.
Further, laws of Nature are not just true generalizations stating that a certain fixed percentage (possibly but not necessarily 100%) of F events are accompanied by G events. They do not just state that, as a matter of fact, x% of F events are accompanied by G events. For it can be accidentally true that x% of F events are accompanied by G events. Suppose that, as a matter of complete coincidence, every time Gail sneezes in London, Pierre sneezes in Paris. Then it is true, but not a law of Nature, that whenever Gail sneezes in London, Pierre also sneezes in Paris. This generalization is not a law of Nature because it is merely accidentally true.
If it is a law of Nature that x% of F events are accompanied by G events, then it is not merely accidentally true that x% of F events are accompanied by G events. Laws of Nature (non-logically) necessitate their instances. Since it is a law that the pressure of a gas varies inversely with its volume at constant temperature, it is (non-logically) necessarily and not just accidentally true that whenever the temperature of a gas remains the same and its volume decreases, its pressure increases.
So laws state not merely that x% of F events are accompanied by G events, but that necessarily, x% of F events are accompanied by G events. Generalizations which are merely accidentally true do not support counterfactualsâ€”they do not tell us what would have been the case if the world had been different from the way it actually was. The accidentally true generalization that whenever Gail sneezes in London, Pierre also sneezes in Paris, does not allow us to infer that if, counter to fact, Gail had sneezed more often in London, then Pierre would have sneezed more often in Paris. In contrast, we can infer from Boyle\'s law that if, counter to fact, the temperature of a certain gas had remained the same as its volume decreased, then its pressure would have increased.
So laws of Nature state that certain generalizations are (non-logically) necessarily true, and they therefore support counterfactuals. AJ
See also modality; philosophy of science.Further reading D. Armstrong, What is a Law of Nature?; , C. Hempel, Aspects of Scientific Explanation.