A philosophical critique of the probability calculus

pp. 312-327

in: Hans Reichenbach, Selected writings 1909–1953, Berlin, Springer, 1978

Abstrakt

The uncertainty of opinions regarding the validity of the laws of probability stems from the apparent contradiction between these laws and the accepted methods of acquiring physical knowledge. The basic method of physics consists in tracing observable events to relations of dependency, in depicting the present events as the effect of a past event and the cause of a future event. The resultant causal chains are considered to be unambiguously determined functional connections, and even where they have not succeeded in discovering causal chains, physicists steadfastly maintain that they exist in principle and will ultimately be discovered. In contrast, the probability calculus makes assertions about non-causal connections, demands, indeed, the causal independence of its objects as a condition of the validity of its propositions. In games of dice, for instance, it is presupposed that the individual throws are independent of one another; the causal chain leading to the result for each single throw is totally disregarded, and the resulting distribution of the throws is designated expressly as a chance distribution, as a "game", in contrast to the causally connected course of other natural events. It looks, then, as though probability and causality are mutually exclusive, as though they introduce into physics the question, "Chance or law?", forcing the physicist to declare himself for the one or for the other.