||Certain events cause, or bring about others. But what is it for one event to cause another?
Some have attempted to explain the notion of causation in terms of constant conjunction. An event of one kind is said to cause an event of another just if all events of the first kind are constantly conjoined with events of the second. So, for example, lighting the match in the room full of gas caused the explosion because every event of the first kindâ€”every lighting of a match in a room full of gasâ€”is accompanied by an event of the second kindâ€”by an explosion.
Others have attempted to explain the notion of causation in terms of counterfactuals. Given that two events actually occurred, then the first caused the second just if, counter to fact, the first event had not occurred, then the second would not have occurred either. For example, the lighting of a match in a room full of gas and an explosion actually occurred; the first caused the second because if, counter to fact, the match had not been lit, the explosion would not have occurred.
The counterfactual account of causation has the advantage that, unlike the account in terms of constant conjunction, it can allow that not all causes determine their effects. Smoking 20 cigarettes a day caused Tina to get lung cancer. But smoking 20 a day merely raises the probability, and does not determine, that one will get lung cancer. The constant conjunction account wrongly implies that smoking 20 a day did not cause Tina to get lung cancer, because smoking that many is not constantly conjoined with lung cancer. Not everyone who smokes that much gets lung cancer. In contrast, the counterfactual account correctly implies that smoking 20 a day did cause Tina to get lung cancer. For if, contrary to fact, Tina had not smoked, then she would not have got lung cancer.
However, while the counterfactual account can allow that some causes merely raise the probability of their effects, it faces other problems. One problem concerns overdetermination. Suppose that two assassins simultaneously score direct hits on the president\'s brain. Then his death is causally overdetermined. There were two independent causesâ€”the first assassin\'s firing his gun and the second assassin\'s firing hersâ€”each of which was causally sufficient for the death. So the first assassin\'s firing of his gun did cause the death. But if, counter to fact, the first assassin had not fired his gun, the president would still have died, because the second assassin would nevertheless have fired hers. The counterfactual account of causation thus seems to fail to accommodate cases of causal overdetermination. AJ
See also event.Further reading D. Owens, Causes and Coincidences; , R. Sosa (ed.), Causation and Conditionals.