Principle saying that, though the same cause must have the same effect each time, the same effect need not have the same cause each time. (Of course the cause on one occasion may be complex and involve many contributory factors; but could these be replaced by different factors when the effect next occurs?)
The principle seems plausible, but is this because the effect is only vaguely described? Death can have many different causes, but could the precise death undergone by Smith at midnight last Thursday?
Is it true that, given a complete description of the universe at one moment, we could in principle know what its future will be but not what its past was (ignoring for convenience problems about self-prediction and so on)?