294 Speciesintroduction of a biology-alone species concept that the biological species was not
identical to the Aristotelian species, as I have shown above. Although the received
view is historically false—biologists before Darwin and logicians knew that living
species could not be defined and were decomposable, unlike logical species which
were both definable and indivisible—there is sufficient truth in it to set out a philo-
sophical problem addressing the nature of kind terms like “species” in biology and
their relation to natural kinds overall. I shall argue that biology does not have natural
kinds (NK) of the sort that physics and the other ahistorical sciences do, and that the
appropriate sortal in biology is a natural group (NG).Classes in Biology
A natural kind is defined by Wilkerson^33 as a real essence; an intrinsic property or
set of properties that make a thing the kind of thing it is, irrespective of any system
of classification. It is not important here whether this commits a Natural Kindist to
scientific essentialism, modal realism, or the logical or causal necessity of laws of
nature. What matters is that NKs imply that class terms involve a de re commitment
to a fixed set of properties intrinsic to any particular that is a member of, or falls
under, the natural kind.^34 I use the term “class” in this section for any sortal con-
cept covering more than one particular, without prejudice with respect to essentialist
definitions or not, as Aristotle did with universals. Here, clouds, clades, and grades
are all class concepts, and data organized using them are classifications. Effectively,
a classification is the organization of any set of data or objects under general terms.
It is clearly the case that if biological taxa are NKs, they cannot evolve (despite
LaPorte’s contrary assertion). What has to occur in evolution, if taxa are NKs, is that
the actual ancestor-descendent, or genealogical, lineages move out of one species/
NK and into another. Species would become, on the NK account, timeless abstract
classes, leaving the dynamic and historical entities of evolution to be the lineages
themselves.^35 Moreover, it would then also be the case that “being a species instance”
would be sharply defined. If there were a microstructural essence, variation else-
where in the genome or phenotype other than in what is essential would not affect
the essence (would be accidental in the old terminology), but as soon as the essence
varied beyond its limits the lineage would no longer be of that species; allowing,
with LaPorte and Aristotle, that the essence is within a range, so that a member of a
species can vary right up to the critical threshold of being a species member, and that
any further divergence from the typical mode will push it over into a novel species
(^33) Wilkerson 1995.
(^34) LaPorte’s claim that his kindism, or essentialism, is historical seems to me to be another example of
the metaphysical chimerism that is found in Mayr.
(^35) LaPorte discusses this objection [2004, 9–10] but fudges this point. He shifts from discussing change
of species to change of members of species. If species are NKs, they do not change, which he admits,
but if they are, then their members are actually only instances of the species. There then needs to be
a connection or commensuration between instances of the first and second species in an evolutionary
sequence, and hence the need for lineages. That said, Ghiselin’s objections to which he is responding
that species have no NK essences must be contingently false. If a species consists entirely of clones,
or the species is reduced to a single closely inbred population, it can easily have properties that are
unique to all and only its members. Most species will vary (be polytypic) but not all need to.