That something like the following thesis is true seems very plausible to me, though some refinement is called for:
- We can only know of, and perhaps even have a belief in, the existence of a basic class of properties or entities if some member of that class enters somehow into an explanation of some belief of ours.
But what about evaluative properties such as normative or axiological ones? Let me focus on the axiological. I take it to be almost definitive of modern naturalism that axiological properties do not enter into genuine explanations. While the existence of a soup kitchen might be explained in terms of a belief in the goodness of a soup kitchen, the goodness of the soup kitchen is entirely explanatorily impotent if naturalism is right. I think the naturalist's best bet would be to identify the property of goodness with some property such as contributing to organic flourishing. But the problem reappears in regard to the property of "flourishing".
If we reject modern naturalism, we can say that flourishing is a property of accord to a thing's nature, and can account for that nature in an Aristotelian way that makes the nature be explanatorily efficacious. Or we can give a theistic story.