Development: Three Grades of Ontogenetic Involvement 185modern synthesis theory that is up for grabs here. If the most plausible account
of the explanatory role of development is inconsistent with the fragmentational,
sub-organismal interpretation of the modern synthesis, then so much the worse for
that interpretation.
4 THREE GRADES OF ONTOGENETIC INVOLVEMENTI distinguish three possible positions regarding the role of development in evolu-
tion. While these positions are probably neither exhaustive or mutually wholly
exclusive, they do at least serve to demarcate increasing grades of commitment
toward the significance of development to the study of evolution. As we move
from one grade to the next, ontogeny takes on a more significant role in explaining
the mechanisms of adaptive evolution. Only the first I believe is wholly consistent
with replicator biology.
4.1 Grade I: Development as Constraint
Lewontin [1974] illustrates the relation between the transmission of replicators,
ontogeny and the selection of organisms as the sequential interplay between two
distinct property spaces: genotype space and phenotype space. The processes of
transmission of inherited material from one generation to the next, the recombina-
tion of replicators, and the introduction of new mutations, induce changes in what
Lewontin calls ‘genotype space’. Their immediate effects are registered as changes
in the relative frequencies of, or the relations between, genes. Selection, however,
distinguishes between organisms on the basis of their phenotypes, consequently,
its immediate effects are changes in ‘phenotype space’. If evolution consists of
changes in relative frequencies of genotypes as a function of selection, then it must
involve an interplay between processes operating over genotype space and those
operating over phenotype space. We need an account of the relation between
these two spaces. It’s easy to map changes in phenotype space onto changes in
genotype space. When organisms differentially survive, reproduce or die, so do
their genes. So changes in phenotype space are directly transposed into changes in
genotype space. However, the mapping of changes in genotype space onto changes
in phenotype space is less straightforward, because it is mediated by development.
Development provides the so-called ‘genotype/phenotype’ map.
Organisms face the tribunal of natural selection as corporate entities, not as
loose aggregates of traits. This fact imposes certain demands on development. At
each stage development is conditioned by twin constraints: (i) it must maintain the
organism’s viability and (ii) it can only build phenotypes out of the materials and
resources at its disposal. These demands introduce a bias into the range of available
phenotypes that development can attain [Amundson, 1994]. Some potentially
viable phenotypes may be closed off to development, because of the particular
path it would have to take to attain them. Others may be particularly easy to
attain, irrespective of the starting point in genotype space. Development biases the