Question Let's say I have a machine with a button and a light bulb where the bulb lights up if and only if I press the button. I don't know anything about it's inner workings (gears, computers, God), I only know the "if and only if" connection between button and light. Can I say that by pressing the button I cause the bulb to light up? (I would say yes). It seems to me that for the causal connection it doesn't matter that I don't know the exact inner workings, or that I don't desire the effect (maybe I press the button just because I enjoy pressing it, or because there is strong social pressure to press it, ...), and that I consider it very unfortunate that the bulb lights up wasting electric energy. Let's now change the terms: instead of "pressing the button" we insert "having a kid" and instead of "the bulb lights up" we have "the kid dies" (maybe when adult). I think the "if and only if" relationship still holds, and so does the causal connection. It would seem to me that parents are causally connected to the death of their kids (e.g. creating a person also causes the death of such a person), and that it doesn't matter that they don't want their kids to die, or that they don't understand exactly how a human being is created or dies, etc... it also doesn't matter if the kid will live till his 90s, or commit suicide as a teenager, or be poisoned. Those are the irrelevant "inner workings", the only certain thing is that he will surely die, one way or another. Any particular holes in this line of reasoning?

Accepted:
November 23, 2011

Comments

Great set of thoughts, here. But maybe one quick mode of response is to remark that much depends on just what you take the word "cause" to mean. You could take it to mean something like this: "x causes y" = "y if and only if x", as you've suggested. Then, granting that both cases above are cases fulfilling the "if and only if", sure, giving birth would count as a cause of the later death. But now two things. (1) Why should "cause" mean precisely that? Wouldn't it be enough if the x reliably yielded the y, even if things other than x could yield the y too? (i.e. couldn't you drop the 'only if' part, so 'x causes y' would mean 'if x, then, y', even if it might also be true (say) that 'if z, then y'?) Going this route would preserve your intuition that both cases above are cases of causation, but focus on whether your particular definition is the best one. (2) Perhaps more importantly, though, one might examine the 'pragmatics' of causation -- how people actually use the word, different from how very precise philosophers or scientists might define it. So, for example, we often restrict the word 'cause' not just to every factor which may be necessary or sufficient or both for an effect, but to the most salient factors, the most explanatorily relevant ones, the most proximate ones, etc. So, you strike a match and it lights; strictly speaking many things are at least necessary for that (the presence of oxygen, the existence of the match, the laws of physics, etc.) but we often say 'the striking caused the lighting', even though all those other factors were necessary. Indeed the striking was neither necessary nor sufficient for the lighting -- the match could have been lit other ways, and striking on its own (without oxygen etc) wouldn't light. So our ordinary use of 'cause' is far looser than some technical philosophical definition. So the question for you is: how, ultimately, are you going to use the word 'cause', and are you justified in choosing that use in light of competing uses?

hope that's useful ...