of time, rather than trying to explain all phylogenetic mechanics by uniformitarian
extrapolation from microevolutionary processes, on the third branch.
At the second level of validation for proposed revisions in the structure of
evolutionary theory, I have tried to develop broad arguments and empirical
justifications for major changes and expansions on each of the three branches of
Darwinian central logic. On the first branch of agency, the theory of punctuated
equilibrium itself, initially formulated by Niles Eldedge and me, establishes the
species as a true and potent Darwinian individual, and grants a minimal guarantee
of descriptive independence to macroevolution by requiring a treatment of trends
as the differential success of stable species rather than the adaptive anagenesis of
lineages by accumulated and extrapolated organismal selection alone. Beyond
punctuated equilibrium, the general rationale for a hierarchical theory of selection,
as presented here through the interactor approach based on emergent fitnesses at
higher levels, may establish a complete (and tolerably novel) framework not only
for grasping the consistent logic of hierarchical selection, but also for viewing each
level as potent in its own distinctive way, and for recognizing the totality of
evolutionary outcomes as a realized balance among these potencies, and not as the
achieved optimality of a single causal locus—a substantial difference from
Darwinian traditions for conceiving the dynamics of evolutionary change. In
working through the differences among levels—see Chapter 8, pp. 714- 744 —I was
particularly struck by the surprising, but accurate and challenging, analogies
(Lamarckian inheritance at the organismal level with adaptive anagenesis at the
species level, for example); and by the different modes of equally effective change
implied by disparate structural reasons for the establishment of individuality at
various levels (particularly, the domination of selection over drift and drive at the
organismal level vs. the potent balance among all three mechanisms at the species
level).
On the second branch of efficacy, I have tried to make the most
comprehensive case yet advanced for internal constraint as a positive director and
channeler of evolutionary change, and not only as a negative brake upon pure
Darwinian functionalism. I proceed by explicating two conceptually different
forms of constraint—structural constraints as consequences of physical principles,
and historical constraints as channels from particular pasts. I argue that each
category challenges a different central tenet of Darwinism—structural constraint
by establishing a substantial space for non-selectionist origin of important
evolutionary features, and historical constraint for explaining the markedly
inhomogeneous filling of morphospace as flow down ancient internal channels of
deep homology, and not primarily as a mapping of adaptive design upon current
ecological landscapes. Beyond any novelty in this general formulation, I have
attempted to develop a conceptual space, and to establish practical criteria, for the
identification of non-adaptive sequelae (spandrels), the evolutionary importance of
their later cooptation for utility (exaptation), and the importance of such reservoirs
of potential (exaptive pools) in explicating the important concept of "evolvability"
in structural rather than purely adaptational terms.
50 THE STRUCTURE OF EVOLUTIONARY THEORY