190 D. M. Walsh
are not simply measured in changes in genotypes, but changes in developmental
processes. We should measure evolution, then, not as change of relative gene fre-
quencies, but change in relative frequencies of inheritable developmental systems.
Grade II ontogenetic involvement, then, stands in stark contrast to first Grade
involvement.
Yet Grade II ontogenetic commitment does preserve at least some of the dis-
tinctions among evolutionary processes that mark out Grade I. It preserves the
independence of the supra-organismal process of adaptation from the organism-
level processes of inheritance and development. Adaptation, on this view, just
as in Grade I, is the consequence of selection operating over heritable variants
[Griffiths and Gray, 2001; 2004]. Indeed advocates of DST take great pains to
emphasize that their approach to development and inheritance leaves the integrity
of the Darwinian process untouched.^11
We hope it is now clear how DST can explain adaptation, in the mod-
ern sense of that term. Change over time in the developmental system
of a lineage is driven by the differing capacity of variant developmental
systems to reconstruct themselves, or, in a word, differential fitness.
What is fitness? In contemporary evolutionary theory fitness is a mea-
sure of the capacity of a developmental system to reproduce itself....
[2001, 209]In their conclusion to the same article Griffiths and Gray offer some definitions
of standard evolutionary terms from the perspective of DST:
Natural Selection — the differential reproduction of heritable variants
of developmental systems due to relative improvements in their func-
tioning.
Adaptation — the product of natural selection. [Griffiths and Gray,
2001, 214]Grade II commitment takes a conservative approach to the adaptiveness of
adaptive evolution. Explaining the adaptedness of organisms to the conditions
of their existence is the exclusive domain of the population-level process of selec-
tion. The individual-level processes of inheritance and development make no great
contribution to adaptedness, except insofar as they serve up the variants among
which selection winnows. Grade II involvement substitutes developmental systems
for genes as units of selection [Griffiths and Gray, 2004], but its conception of the
process of adaptation is still resolutely sub-organismal.
The charge that DST is ‘sub-organismal’ is bound to rankle its adherents. After
all, it as a DST shibboleth that the organism/environment boundary is arbitrary
or irrelevant ([Oyama, 1985]; [Oyamaet al., 2001]). DST is to be credited with the
recognition that developmental systems extend beyond the skin, but that alone
(^11) Depew and Weber [2001] have their doubts.