In this paper today, I would like to offer a new analysis of causation and of causal claims. It is an unorthodox one, as you will see, but I suspect that in the not too distant future it will be seen as intuitively, perhaps even trivially, true.

​

               I hardly need defend the urgency of my project. Ever since Hume, philosophers have wondered whether there are causes. This is a desperate situation. With no causes, it's hard to see how brushing my teeth is likely to prevent tooth decay. Indeed, it would not be unreasonable to read Hume as an advocate of rotten teeth, which might explain the sad state that many British mouths find themselves in today. The attentive listener will have noted that I said Hume's advocacy of rotten teeth might explain the abysmal state of British oral hygiene. Of course, if Hume is right about causation then nothing explains anything, and that explains why I have been tentative in my claim.

​

               The account I would like to propose is this. The claim ‘x causes y’ is to be understood in the following way: ‘x makes y happen’. That is, to say that x is the cause of y is just to say that x makes y happen. Or, to put it more succinctly, if x is the cause of y, then x makes y happen.

​

               This is no doubt a startling claim, and one in need of further clarification and defense. To begin, I should like to contrast my analysis with another that might, on its surface, appear similar. Suppose one were to claim that 'x is the cause of y' means that x brings y about. But ‘bringing about’ is hardly an informative verbal clause, and does little ampliative work. This way of putting it lacks the opaque transparency that we’ve come to expect of philosophical analyses of causation.

​

               Now this new account is not necessarily inconsistent with other, more traditional analyses, such as Lewis and Hausman's analyses of causation in terms of counterfactuals or Eells' probabilistic theory of causation. Consider first counterfactual analyses of causation. These are efforts to account for the meaning of causal dependencies. 'X causally depends on Y' means that were X not present then neither would Y be present. But, with just a little ingenuity, it is easy to see that this counterfactual analysis can be made to fit with my suggestion. For surely, if X makes Y happen, then whatever is counterfactually true in the causal case is true in the makes happen case, which is just the result we want. Of course, those not intelligent enough to grasp this point may harbor doubts, but to appease these doubts I would have to say things that are even more difficult to understand and this would be a waste of time. To be sure, questions regarding the nature of time are a worthy pursuit in themselves, and were it not for lack of it I would take them up now.

​

               Related to counterfactual analyses of causation is the British philosopher J.L. Mackie's suggestion that causation be analyzed in terms of INUS conditions, where 'INUS' is an acronym. Of course, the greatest challenge a proponent of INUS conditions faces is to remember what this acronym stands for. Philosophers of science often disagree about this. Nancy Cartwright takes 'INUS' to mean "I never understood Spanish," and cites Mackie's Britishness as evidence. Van Fraassen (personal communication), insists that 'INUS' stands for "I need unprotected sex," also mentioning Mackie's nationality in defense, but adding to this the highly confirmed fact that Belgium is just across the channel from England. One final theory is that, in fact, INUS is a typographical mistake, and should have read ‘ANUS’, which stands for “I’m just talking out of my ass.”

​

          Whatever the case, INUS conditions purportedly capture an important feature of causal relationships and, without worrying about the details, I can assure you that if this is true it is so only to the extent that they matter to whether X makes Y happen. Q.E.D.

​

          Let's now look at a probabilistic analysis of causation, according to which one thing is a cause of another if it raises the probability of the other. The notion that x is the probabilistic cause of y will be abbreviated by the phrase ‘x is the p-cause of y’. Thus, on my account, to say that x is the p-cause of y is to say that x makes it more probable that y will happen. The fit could not be more perfect. As Russell might have observed, the "makes it happen" locution has all of the advantages of honest labor without having to steal anything.

​

               Notice, too, that the account is neutral with respect to agent or event causation. That ‘x causes y’ means ‘x makes y happen’ whether x and y are agents or events. On the agent reading, ‘x causes y’ means ‘an agent x makes y happen’; on the event reading, ‘x causes y’ means ‘an event x makes y happen’. As you can see, the account holds either way.

​

          Another pleasing feature of my "makes it happen" analysis of causation is that it clears up a puzzle about ceteris paribus clauses. The mystery is this. It is not true that every time I strike a match a flame ignites. Perhaps the match is wet. In such a case, it is false that striking a match makes happen a flame. It is for this reason that some have seen the need to introduce ceteris paribus clauses. But, there is simply no consensus whether these clauses are to be pronounced "seteris paribus" or "keteris paribus." Probabilistic and counterfactual analyses of causation have little to say about this (but see Sober forthcoming). However, in my "makes it happen" analysis of causation, we need no answer to this question. In fact, we can sidestep the debate altogether in this way. We translate "X causes Y, seteris paribus" as "sometimes X makes Y happen and sometimes it doesn’t." Although I will not attempt a proof here, I assure you that the same result applies to the claim "X causes Y, keteris paribus."

​

Objections:

          Now one might object that there are many instances where x is the cause of y but where x does not make y happen. But that’s just stupid.

​

               Another possible objection is that the inverse of my analysis does not hold. That is, it could be argued that just because x does not make y happen, it does not follow that x is not the cause of y. But I suspect you already see the fatuousness of this kind of objection, and so I won’t belabor it any longer.

​

               Finally, there is the matter of truth conditions. If ‘x causes y’ means ‘x makes y happen’, then when will it be the case that the proposition ‘x causes y’ is true? This aspect of the theory needs further work, and I can here offer only tentative steps toward a solution, but here is a first step. On this account, the proposition ‘x causes y’ will be true if and only if x makes y happen. Whether this holds equally well for token and type causal claims is a question I hope to pursue in the future.