ABSTRACT
The distinction between intrinsic and extrinsic conditions is fundamental to the notion of a 'causal' or 'dispositional' property. Relational properties would seem to be necessarily 'extrinsic', at least with respect to physical things and substances. It seems, therefore, that the question which of the properties of a thing constitute its 'nature', and which are extrinsic, depends on tests just as much as does any question about its potentialities. For example, a refusal to accept Hume's doctrine that power is nothing in the agent beyond its exercise can draw some strength from the present argument, according to which unexercised power is a matter of what its possessor would do, having the nature it now has, in conditions not affecting its nature. 'Actualism', the doctrine that nothing ever has the power to do what it does not actually do, is generally grounded on the principle that, if an action is caused, it follows that the agent cannot do otherwise.
