The Development of the Philosophy of Species 297
else, these laws are known only to apply to terrestrial organisms. In short, biological
NKs are universals in a very small and domain-specific universe.^42
Biological generalizations, usually ecological ones such as the Lotka-Volterra
predator−prey relationship, define the universals that are active under that regime.
“Predator” is defined as a grade of organism that consumes other organisms, and
there is an implied relationship between population size and structure of the one
and the other. Predators must be fewer in number than their prey, simply due to
the energy-relationships implied by organisms subsisting on other organisms. When
they over-predate, there is a population crash of prey followed by a population crash
of predators shortly after. What, from this model, do we know about the instances of
predators and instances of prey? We know they will evolve in an arms race if they
are tightly linked trophically. We know that there is a general ratio of energy budgets
where the energy budget of the predator is greater than the energy budget of the prey,
and so on. All these things are implied by the model. But should we discover that
predator A has some feature—say a claw structure—can we therefore conclude that
another predator, B, will? Without further information, no. One predator might be
an eagle, while another might be an ant-lion. In fact, without phylogenetic informa-
tion, nothing else apart from the definition of the grade “predator” can be inferred.
Model-based grades are not inductively projectable, phylogenies are.^43 Ecological
laws apply in any case where there are “transactions” of energetic capture, lim-
ited space, and resources. They are laws of economics irrespective of whether the
objects over which they range are biological or not. The model (that is, in this case,
the Lotka-Volterra equation plus some rules for interpretation, and a possibly tacit
delimitation of the explanatory domain) offers only a description of the relations
between two populations. It might equally describe an inorganic system; say, an
economic market, a computer model, or an engineering process.
However, for all that, a clade, or a monophyletic group, may instantiate a grade
so long as a number of conditions are met. It must of course be monophyletic. Any
taxonomic group that is only partial, or which is formed by the independent attain-
ment of the grade, is considered non-natural in cladistic classification. The so-called
“evolutionary systematists,” such as Mayr and Simpson, believed that this was per-
fectly acceptable so long as the grade was salient, as illustrated above. However, cla-
dists rejected this on the grounds that it returned classification to an arbitrary state.
So long as the individual systematist considered the grade significant, any grade-
based classification can be proposed and defended equally as well. But in a cladistic
system, the grade, which is an intensionally defined class, and the clade, which is
an extensionally defined class, must coincide for the grade to be (contingently or
historically) natural. As soon as any other lineage attains the grade, or a lineage of
the clade leaves it, the grade is no longer representative of a natural group. In short,
NKs may be NGs and vice versa, but they need not. A grade may be a polyphyletic
(^42) Richard Dawkins has claimed [1989, 191f] that natural selection must apply to any living thing irre-
spective of it having evolved from terrestrial antecedents. But I am unsure that natural selection is
a law as such. It is a formal model, but it need not apply in every case or even most cases of living
organisms even on earth, as stochastic processes like genetic drift indicate.
(^43) Nelson 1989, Griffiths 1994. I discuss this further in Wilkins and Ebach 2013.