I have been discussing causation over at Crooked Timber and despite a couple of attempts to explain myself I am not being understood. Having chucked a simple textbook example at me folks seem to have just ignored my point which I find interesting. To recap the point at issue is when does a set of correlations become a cause? I have proposed that it becomes a cause when some of those correlations lie in the future, when there are predictions involved and the correlation is surprising—i.e., is the correlation is true it adds to our knowledge of the world (see here; my thinking here has been entirely shaped by the late, great Richard Feynman). So if I claim that when you jump up and down on one leg while picking your noes, your tooth ache will always disappear then you can try it out and see it is it works—see if you observe this correlation the next time you get a tooth ache. If you do (and repeatedly so) then you have some new tentative causal knowledge that will become strengthened as you reliably see the correlation in a variety of circumstances. The textbook example that people have been throwing at me is that if I take causation is correlation too seriously then I will be forced to conclude that cock’s crows cause sun to rise, but this isn’t a problem here. Suns rising after cocks crowing isn’t surprising to me—I am not looking to explain that correlation having a perfectly satisfactory set of causal relationships to explain it (but thanks anyway).
I’m not sure what you’re getting at Chris. Discovering a correlation is an invitation to further study. One shouldn’t leap to conclusions. The problem is that no matter how fine grained our mechanism is there will always be a leap involved. So we are left with observing that B follows A and concluding that A causes B. We call that deduction but there is a gap in our understanding. There always will be.