How can know that (sensory event) A is the cause of some (sensory event) B? Since A and B are distinguishable, we do not think of one being the cause of the other until, through experience, we find constant conjunction between A and B (coupled with "contiguity" (closeness) of A and B, and the priority of A to B). This constant conjunction gives rise to a superstition that there is a necessary connexion between A and B but this notion is just superstition, in that we might have had a long run of coincidences. Since A and B are separable, and we can conceive them existing apart, there is no purely rational basis for deriving B from A; and appeal to some general principle derived from experience (i.e., the future will be like the past) is not helpful because any such principle suffers from the same problem as "A causes B" -- because this too can be coincidental.