Causality is one of those topics that is going to get me in trouble if I pretend to know too much about it. It's an issue in physics because of quantum mechanics, which ascribes an irreducible randomness to nature on the smallest, subatomic scales. (See my first answer in this series.) If you don't know when an atom is going to decay or what direction a photon (the discrete carriers of light in quantum theory) is going to go when an atom loses energy, then you can't say "why" it went one way or another or one time or another. Some part of causality has been lost.
There is another issue with cause and effect, it seems naively to me, when you fold in time. The basic laws of physics are supposed to work equally well whether time is going forward or backward. You can't tell, watching a movie of a pair of billiard balls or a pair of electrons colliding, whether the film is running forward or backward. If the universe really works this way, then you can't have cause and effect uniquely defined. One time the red ball causes the blue one to go flying; the other (reversed) time the blue one causes the red one to fly off.