Cause and Effect II

 Yesterday I wrote about cause and effect through the prism of the work of Wesley Salmon, late of the University of Pittsburgh. 

Many philosophers have disagreed with Salmon, for example with his emphasis on spatio-temporal continuity. Why should we presume to rule out of court the very possibility of action at a distance? 

One philosopher who seems willing to rule action at a distance back in, for example, is Philip Kitcher. Kitcher places "causation" within the broader term "explanation," and he has a unificationist theory of what explanation means. We generally explain X by placing it in a larger category, unifying it with other phenomena so that X is just an illustration of X(1). Then we explain X(1) in the same fashion, perhaps a generation later with reference to X(2). So, for example, Franklin explained lightning by calling it "electricity," an example of the same shock we get when we walk on a shag rub then touch a metallic doorknob. Some decades later, Faraday explained electricity in turn by putting it within  a broader thing called electromagnetism. 

As to causation specifically, Kitcher's view is that we can begin with a unificationist view of explanation and in due course pull causation out of that silk hat. 

Salmon, in putting limits on what causation can consist of, seems to Kitcher excessively inhibited. 

Another critique of Salmon arises from the notion, dear to some, that ideas about causation are ideas about counterfactuals. "Since I salted the clouds on Tuesday, there was rain on Wednesday." The corresponding counterfactual statement is, "if I had not salted the clouds, there would have been no rain." On some accounts, then, every possible causal connection has four terms: Cause (1) which led to Effect (1) and C(2) (perhaps a whole alternative-state-of-the-world on Tuesday) would have led to E (2), a longer drought.

There are difficulties stating the precise role of the counterfactuals in this sort of schematic.  But in general it helps explain the pressing distinction between causation and correlation. Let us suppose that we observe that private school students get higher grades on a standardized test than their public school peers. Do we attribute this to the superior instruction students in the former group enjoy? Maybe. But there are other considerations we would want to consider. Maybe it is simply the case that the kids who enter private school have parents who are in a position to pay the tuition.  That fact may provide those kids with a lot of other advantages that could lead (before standardized tests push into the picture) to a greater knowledge base, greater fluency in writing single-paragraph answers, and so forth. 

The counterfactualist theory of causation looks at a quarrel over such questions as this and says, "Aha! See what I mean? We cannot say of any student (called, say, Joe) that IF Joe had gone to public school, he would not have thrived, which is precisely the same thing as being unable to prove that the private education was the cause of the thriving." 

Counterfactuals, then, at a minimum help illustrate the difference between cause and mere correlation. If that makes them the essence of causation, then so much the worse for Salmon's theory, in which they play no part.