I'm dealing with certain aspects of Lowe's metaphysics and certain ways of reading Aristotle and it occurred to me to ask:

What possible argument could justify natural kinds as a primitive or a kind of whole not reducible to some kind of constituents?

It's puzzling to me both from a phenomenological position- since it's not clear to me that natural kinds appear as something over and above a collection of their properties -and from an historical perspective, since Aristotle is ambiguous about reduction and both Plato (see the Timaeus) and some Platonic followers (Xenocrates explicitly) tend to want to reduce Forms to numbers or mathematicals in the World Soul or something of that sort.

Once I gave up the idea that "reduction" of natural kinds entails some kind of materialism (and thus gave up a kind of naive Platonism of especially mundane things) I've come to see less and less justification for natural kinds on the level of metaphysics. I can see their merit from the level of a kind of scientific, psychological, social, humanistic, or ethical realism, but once I proceed to cash out the objects of these domains into their fundamental constituents I have no argument.