The Development of the Philosophy of Species 299
have a Nominal Essence—that is, they need not be definable—but they have, in
Locke’s metaphysics, a Real Essence, in the sense that they have something in com-
mon that causes them to be a group. The essences are, of course, the pedigrees of
these things so related. A pedigree is a real causal relationship. People, and horses,
have “a community of blood,” or as we would say today, a genetic or genealogical
relationship.^45 Darwin himself later wrote that all true classification was genealogi-
cal, in the Origin. This was taken up, but seemingly not further developed, by Kant.
His “scholastic,” or “academic” classification, based on similarity, is a class-based or
gradistic classification. The “natural system” is based on the actual causal relations
of the organisms, as with Locke.
Such groups are causally generated. And they are as natural as anything is capa-
ble of being. But they are not NKs, as they have no definitions, unless it happens that
accidentally with respect to the propinquity of descent (Darwin’s term) that they all
share some character or trait, in which case it is better thought of as a diagnostic trait.
If all members of the A family have red hair, they are not a family by virtue of their
auburnity but by virtue of being related by (real) parent−child lineages that are more
closely connected in a network graph representation than they are with the outgroups
used to diagnose it.
Edward Zalta, in his interpretation of metaphysics,^46 introduces a useful distinc-
tion between abstract objects, which are spatiotemporally unrestricted, and concrete
objects, which are indexed as to time and space, which is to say they are bounded in
time and space (though they need not be contiguous or even connected). A species
exists at a time and in a location, even if it happens to vary over ranges (and even
if it re-evolves). A natural kind does not. Once an NK-object, always an NK-object.
If a species is an NK, then it can come and go into and out of concrete existence
many times. It can be caused by different physical processes. For example, an NK T.
rex can evolve by ordinary evolution once, and once extinct, might be reformed by
Jurassic Park scientists. But an NG T. rex can evolve only once. The latter “T. rex” is
only a copy and deserves its own name, and it will have quite distinct causal relations
to the earlier “example.”^47 In the case of a respeciation event such as described for
some species of fish, there is a less clear-cut “birth” in evolutionary terms for the spe-
cies, but the “speciation event” is still singular, and the species is still a metaphysical
particular, and thus a concrete object. Vagueness is expected in evolution, and pres-
ents no direct challenge to the reality of the outcomes of its processes.
How does this generalize beyond biology? Are there NGs in other disciplines
and domains? In physics, a particle has the same properties no matter where it
appears. An electron always has the same charge, components, and propensities as
(^45) Interestingly, Stamos 2003, 40−47 reads Locke’s views on classification quite differently, seeing him
as a gradist pure and simple. I think, however, that this is due to taking Locke’s views on classifica-
tion purely from his discussion in the Essay on logical species. It is true, though, that Locke does not
treat causal relations as the basis for [logical] species. Stamos’ own relational account of [biological]
species could be traced to this passage in Locke. That Locke was a gradist is also argued, not in those
terms, by Ayers 1981.
(^46) Zalta 1988.
(^47) Gradists such as Ruse and possibly even Darwin of course, challenge this, although I believe Darwin
was merely unclear about the sole necessity for genealogical classification, and not confused, as
Padian claims [1999]. However, see the discussion in Stamos 2003.