Species

The Development of the Philosophy of Species 303

properties. However, he holds that Linnaean taxa are defined by their intrinsic

properties:

Intrinsic Biological Essentialism: Linnaean taxa have essences that are, at least partly,

underlying, probably largely genetic, intrinsic properties.^62

The partial nature of his definition is because the nature of the species as taxa is due

to relational, as well as intrinsic, properties, while IBE is a claim about the explana-

tions of the properties of organisms and their membership in those taxa, not the cat-

egory of species. The “largely genetic” component includes developmental programs.^63

Devitt and others have appealed to a causal essence. This is to say, the essential

traits (which is the biological instance of “properties”) must be causally efficacious

in maintaining the species. The causal account that he and many others indicate

underpins a realistic biological essentialism is our third option: Richard Boyd’s

Homeostatic Property Cluster account, to which we now turn.

Homeostatic Property Cluster Kinds

One of the early (1960s) arguments against scientific essentialism (outside of phys-
ics) was that natural kinds do not have uniquely shared properties among all their
members. In short, organisms vary. As we have seen, variation within species arose
as a concern in the early nineteenth century, and by the mid-twentieth century, spe-
cies are known widely to be polytypic, with variants and multimodal distributions of
traits common. One essentialism proposal by Richard Boyd took this variation to be
real and normal, and sought to account for kinds in the face of their statistical nature.
Boyd argued that natural kinds do not, generally, in science require JNSSPs.^64 His
“accommodation thesis” is that natural kinds had better account for scientific prac-
tice, and that the archetypical natural kinds—species—did not require the “traditio-
nal” essentialism, but were cluster concepts. However, Boyd said, the clusters were
not merely social constructions, but causal clusters that kept the kinds in homeosta-
sis with which the social constructive practices of science engage. This is called the
Homeostatic Property Cluster (HPC) concept of natural kinds. Like Devitt’s essen-
tialism, who in part follows him here, this is a causal essentialism.^65 Boyd’s exem-
plar is species, and this has been taken up by numerous others.^66 He has extended the
HPC account to genealogically based higher taxa (i.e., monophyletic groups) in the
process redefining monophyly as a conserved developmental modularity.

(^62) Devitt 2008, 378.
(^63) Devitt does not argue in favor of a genetic definition here, but merely appeals to such things as DNA
barcoding and genome comparisons between species. As I have argued above, barcoding does not
define or even with certainty identify species. See also DeSalle et al. 2005, Wheeler 2005, Will et al.
2005 for critiques. As to cross-genome comparisons, Devitt is conflating identification of a “typical”
organism within a species with what makes the species bear typical traits, in this case.
(^64) Boyd 1991. See also Boyd 1999b, Boyd 1999a.
(^65) Although Boyd’s HPC kinds view has gained wide currency among philosophers and biologists, it
is not without its critics. Ereshefsky and Reydon 2015 argue that HPCs include some non-scientific
kinds, and excludes some scientific kinds, for example.
(^66) Brigandt 2009, Rieppel 2009, Barker and Wilson 2010, Martínez 2015, Neto 2016.