The Development of the Philosophy of Species 293
a candidate taxon, Homothermia^31 —which clades can attain through evolution, but
that is a purely theoretical name. To define it you require some reason to think it
is an important kind and not simply a subjectively interesting one. You must have
some prior commitment to the significance of warm-bloodedness (and of course it
was important in the minds of those who “discovered” it because they were warm-
blooded). In short, once has to have some sort of model from which to draw signifi-
cant and salient criteria to define the grade out of an infinite range of possible grades.Natural Boundaries......................................................................
In the cladistic approach to classification, the classes are causal, not descriptive. One
often cannot define a clade, only identify (or diagnose) it and point to it. Theory, and
theoretical terms such as “warm-blooded,” “intelligent,” “predator,” “parasite,” can
only be applied or not to a clade according to the standards of a theoretical structure
and interpretation, and they are no more informative than the model on which the
grades are based. If you only know that something is a predator, you only know
whatever “predator” implies (that it is a heterotroph, that it is ambulatory in most
cases, that its population size will co-vary with its prey’s in Lotka-Volterra cycles
under some ceteris paribus conditions^32 ); you know little else about it. On the other
hand, if you know that something is a member of a particular clade, you know a
lot about it—you know most of: its dental structures, skeletal structures, cell types,
developmental sequences, genetic features and functions, overall lifecycle, and so
on. On the cladistic account, these things are affinities, as they might have said in the
nineteenth century, while ecological and other universalizing theories only provide
information about analogies (in the biological sense of the similarity relation that is
not due to common descent, opposed to homology).
Natural kinds have always been an attempt to draw natural boundaries among
universals. The thing about universals, though, is that they are timeless and abstract.
A universal is something that any historical thing may instantiate, but which contin-
ues to exist if nothing instantiates it. A historical entity, though, is a particular, an
individual. When nothing of that particular exists any more, the particular is gone,
extinct. At best, particulars instantiate, or subsist as examples of, universals. So we
must ask ourselves a number of questions—some historical and some formal. One
question is, how is classification in biology to proceed, via universals or via particu-
lars, on the basis of laws or actual processes? Another is, why do we insist that clas-
sification must be based on universals? There are others, which we shall encounter
as we proceed, but these two offer a starting point.
Because Aristotle’s logic, in the Categories and elsewhere, begins with predicates,
it is perforce an essentialistic (definitional) logic, although pretty well everybody
from him through to Locke understood that the definitional account of essence did
not apply in a simple manner to living things. Moreover, it was understood after the(^31) Hu x ley, loc cit.
(^32) Mark Colyvan noted in commenting on an earlier draft that the Lotka-Volterra equations also apply
to the relationship between grazers and the plants they graze, so perhaps it turns out that the naïve
category of “predator” in fact is artificial, or an arbitrary partition of a natural kind (see Colyvan and
Ginzburg 2003).