The Grammar of Science - Chapter V

Chapter V

"Contingency and Correlation - The Insufficiency of Causation"

1. Routine in perceptions is a relative term; the idea of causation is extracted by conceptual processes from phenomena, it is neither a logical necessity, nor an actual experience. We can merely classify things as like; we cannot reproduce sameness, but we can only measure how relatively like follows relatively like. The wider view of the universe sees all phenomena as correlated, but not causally related.
2. Whether phenomena are qualitative or quantitative, a classification leads to a contingency table. From such a table we can measure the degree of dependence between any two phenomena. Causation is the limit to such a table, when it contains an indefinitely large number of "cells," but in each array only one such cell is occupied. Mathematical function arises when the belt of dots which are the actual result of all experience shrivels up into a curve. It is a purely conceptual limit which is just as much a conceptual limit to actual experience when we use a multiplicity of "causes."
3. The intellectual gain of this contingency category lies in the fact that it sees variation as the fundamental factor in phenomena. Determinatism is the result of supposing "sameness" instead of a mere classificatory "likeness" in phenomena. Variation and correlation include causation and determinatism as special cases, if indeed they have any actual existence in regard to phenomena. No experience we have at present justifies us, however, in assuming them to be anything but conceptual limits created by human need for economy of thought, and as little inherent in phenomena themselves as geometrical surfaces or centres of force.