Is being dependent an intrinsic property of an entity?
Suppose we say that it is intrinsic. Then we have the following interesting consequence. Assuming there are dependent entities, it is possible to have an intrinsic property, D, whose possession entails the obtaining of a genuine relation (a dependence relation) to another entity, but where D is, nonetheless, not relational. This would force us to deny strong recombination principles in accounts of modality. And that would be a good thing from my point of view. For one, it would force a humility in the move from apparent conceivability to possibility. (The modal problem of evil is one place where this matters.)
Could we say that being dependent is not an intrinsic property? That, I think, would be odd. If being dependent is not an intrinsic property, or at least is not entailed by the intrinsic properties of the entity (all I need for the arguments of the previous paragraph is that being dependent is entailed by the intrinsic properties of a thing), then being dependent is not a matter of some kind of inner need or lack in the entity. If George could survive without water, and without any substitute (natural or supernatural) for water, and without without any intrinsic difference in him, then he is not really dependent on water for his existence. My intuition is that the notion of a dependence that does not supervene on the intrinsic properties of a thing and that (therefore) is merely accidental is a sham dependence. I don't yet have a very good argument here, beyond just restatements of the intuition.
If I am right, then Hume has no conceptual resources to affirm that any entity is genuinely causally dependent. For on his view, "causal dependence" would have to be an extrinsic property of an entity, and hence, if I am right, would at best be just a sham dependence.