The Development of the Philosophy of Species 291
A clade is, as the Greek root word klados, meaning “branch,” suggests, a histori-
cal class.^27 This notion involves a causal history, which may or may not result in a
set of JSSNPs, and in the leading school of biological classification named after it
(cladism) a clade is understood to be uniquely isolated from the rest of the tree by a
single stem or “cut.”^28 Such clades have the property of being monophyletic, which is
usually defined as the stem (taxon, or species) of the branch and all of its dependent
branches. Clades nest within more inclusive clades. The diagram of relationships
between taxa forms a cladogram; there is a dispute over whether or not a cladogram
is a direct or indirect representation of the evolutionary history of these groups.
Those who think it is a direct tree usually refer to the internal nodes of the clado-
gram as earlier taxa; those who do not refer to internal nodes of the tree as formally
equivalent to inclusive sets; that is, each branch represents a set of smaller sets, or of
individual organisms.
A cloud is just a cluster in some abstract morphometric or phase space of char-
acters of organisms. On the well-founded assumption that organisms vary over dis-
tribution curves for any measurable character, numerical taxonomy, known later as
phenetics (in contrast to cladistics), attempted to find “natural kinds” of organisms—
that is, of taxa—in a purely empirical and atheoretical way. Unfortunately, organ-
isms vary in ways that are often uncorrelated with other traits they bear (due to
mosaic evolution), and so it transpired that what appeared as a phenomenal cluster
on one set of principal components was divisible in different ways on another set.
Many of the analytic techniques of phenetic analysis have been assimilated by both
the cladistic and the gradistic schools of classification, but the underlying philosophy
of pure operationalism is pretty well a dead issue in biological systematics.
The difference between these schools was not over the task of classification, for
each thought that it was to find the “joints between natural things.” The dispute
largely turned on what was acceptable in the way of theory-dependence in clas-
sification. Pheneticists were extreme empiricists; a classification should reflect no
prior theoretical commitments. Gradists are comfortable with the notion that we can
use theoretical kinds to distinguish objects, in ways I shall explore below. Cladists
assumed that we are recovering some signal of the history of the organisms through
the employment of a formal and logically coherent methodology, but that some gen-
eral knowledge of the organisms and the taxa in which they exist is required to
separate the informative from the uninformative characters.(^27) The term was coined by Julian Huxley, and defined as “delimitable monophyletic groups” [Huxley
1957b] and “groups of common ancestry” [Huxley 1957a, 90]. He also used the term grade to mean
an advance “sometimes independently achieved, sometimes in common.” I am deliberately not adopt-
ing the rather insulting terms employed by Mayr and Bock [2002, 180] to distinguish between “clas-
sification,” which they define as, or rather assert is, based on similarities, and “cladification,” or
genealogical “ordering.” It appears that Mayr and Bock wish to restrict the nature of classification to
grouping by similarities (in conjunction with phylogenies). This is contrary to historical antecedent,
and restricts, deliberately, the role of systematics or taxonomy to the collation of grades as defined
here. Alternatively, should one wish to assert that cladification is one kind of classification, which I
do, the other kind might be better termed “gradification,” or the making of grades. I first used this
term in Wilkins 2002. See also Hennig 1975.
(^28) Sober 2000, 164.