The Ordinary Concepts of Cause and Counterfactual Dependence
How intuitively plausible is it that we ordinarily understand the meaning of causal claims in terms of the truth-conditions of counterfactual claims? For my part, I doubt that anyone will accept any counterfactual as true unless they believe they know the underlying causality. The claim ‘I wouldn’t have gotten a jackpot on the one-armed bandit if I hadn’t tapped it three times on the side and stomped hard once’ will be considered mere superstition, unless we know there is a mechanism inside the machine that has to be triggered by tapping and stomping toproduce three aces in a row. The same goes for claims involving the healing potency of homeopathic concoctions, say: ‘I wouldn’t have survived cancer if I hadn’t drunk extremely diluted amounts of ginger’. Only those who accept the causal powers of homeopathic concoctions will accept the claim. Someone who feels ignorant about the connection will remain agnostic.
Consider a plausible counterfactual: ‘if there hadn’t been lightning there wouldn’t have been thunder’. Do we first believe that there would be no thunder without lightning and therefore infer that lightning is the cause of thunder, or do we believe that lightning is the cause of thunder and therefore there will be no thunder without lightning? Note that the counterfactual is only true if we first stipulate that the thunder of which we speak is the sound produced by lightning, or, in other words, if we define thunder as the sound caused by lightning. The counterfactual is false if we allow ‘thunder’ to also include sounds from explosions, landslides, and earthquakes. Conceptually, the connection depends on the stipulation of a causal link.
Bear in mind that it was not always believed that there would be no thunder without lightning. It was a scientific discovery that lightning is an atmospheric electrical discharge that produces a shockwave in the air, which we in turn perceive as thunder. The pagan Scandinavians believed that lightning and thunder were two correlated effects of a common cause: the spark and sound, respectively, produced when Thor went berserk with his hammer. On that understanding there may well have been thunder without lightning, since Thor’s hammer may not always have ignited a spark even though it always produced a sound. Here it is our understanding of the causal connection that determines whether we accept or reject the counterfactual, and, as far as I can tell, the same is true for every causal counterfactual. This argument is completely open to refutation. Just come up with a counterfactual we accept while being ignorant of a causal connection, and which establishes that there is indeed a causal connection. But in the absence of a plausible example, consider the idea refuted that CTCs can clarify the content of the ordinary concept of causation.
I have demonstrated that you can make sense of counterfactuals if you presuppose the causal realist understanding of causation paraphrased by Bunge. On the other hand, I can see no discernible prima facie reason for thinking that counterfactuals can explain causation. Of course, if one is already persuaded by philosophical arguments that the causal realist conception is false—that causation is contingent and thus cannot ground counterfactual dependence no matter what our ordinary concept says— one might be tempted to find another basis for counterfactual claims, such as similarity between possible worlds. But will any other ontology do the trick, in particular an ontology of possible worlds? If it is not at all clear which ontological model of possible worlds will ground the truth of counterfactual claims, how optimistic can we be about the promise of
A Critique of Counterfactual Theories of Causation 157 explaining an unclear notion of causation by appeal to a clearer notion of counterfactual dependence?
