How flexible is the principle of causality?

[Alex K]

Question: If a specific effect is the consequence of a specific cause, will that same cause or causes always produce the same effect(s)? It seems like the presumption that science is largely conducted on. But if A causes B, can it also cause C or D or E? Or if AB = CD, is it due to necessity? Or can AB also = EF or GH? What degree of flexibility lies in the apparent regularities of causes and effects?

As I like to do so often, I'll ask a question back: when do you say that B is caused by A, i.e. what is your definition, notion of causation?

But concerning your (and my) question:
Famously, in quantum systems there are many processes without deterministic outcome. Likewise, in classical chaotic systems, tiny perturbations get blown up macroscopically such that asking for e.g. the cause of a storm does not seem to be a well-defined question.

Finally, the notion of an arrow of time we have is largely based on entropy: processes without entropy increase are called "reversible" for a reason: if you watch them played back backwards, you will not be able to tell. Your notion what is earlier and what is later therefore relies on statistics, the second law etc, and the second law is not strictly true, but only on average. If you wait long enough for a given number of particles in a system, you will see a violation.

http://en.wikipedia.org/wiki/Poincar%C3%...ce_theorem

The time you have to wait however goes exponentially with the number of particles, and so for any macroscopic system never comes for all practical purposes. Still, I think it's an important point to remember.