Abstract

I suggest that we may settle the question of their relatedness by way of two arguments. The first argument holds that two worlds might be identical in structure but different in their dispositions and subsequent behaviors. This argument loosens the relation of dispositional to structural properties; but, though plausible in itself, the argument has disastrous implications for the uniformity of processes within each world. The second argument supports our intuitive belief that the dependency of a thing’s dispositions upon its structure must be complete, e.g., as the knife’s powers for cutting devolve only upon the fine edge of its rigid blade. This second argument is also a defence against the chaotic implications of the previous one. But it has these two effects only because of affirming that structural properties are, more accurately, geometrical-structural properties. Together, these arguments justify the conclusion that no theory of dispositions is comprehensive, unless it provides for these two factors: Geometrical-structural properties are necessary and sufficient to determine what a thing’s dispositions shall be; but these structural properties are distinguishable from dispositions as properties constitutive of a thing are different from its qualifications for relatedness, and especially causal relatedness, to other things.