1314 THE STRUCTURE OF EVOLUTIONARY THEORY
changes under two rubrics: the "random" and the "different rules" models for
alternating regimes of faunal turnover in mass extinction vs. ordinary, sequential
(and, in Darwin's own preferences, overtly competitive) replacement in "normal"
times between these infrequent but intense episodes.
In the purely random model, which I do not consider of great importance in life's
history, a group might die in a mass extinction for no reason of particular sensitivity
towards the catastrophic agents of extermination, but for reasons little beyond the
luck of the draw—the "bad luck" vs. the traditional "bad genes" in Raup's amusing
characterization (Raup, 1991a). (I consider this random model as less important than
deterministic reactions to the "different rules" of these parlous times because, even in
these extreme episodes, the species number of major taxa generally remains large
enough to preclude full removal by random inclusion of all members of one kind
within a percentage of totality. All ten red beans in a bag of 100 may disappear often
enough in a random destruction of 75 beans. But we would not expect all 10,000-
dinosaur species in a fauna of 100,000 land taxa to die in a truly random reduction to
25,000 taxa.)
Nonetheless, a fact of the fossil record, not widely known or appreciated outside
the community of professional paleontologists, may grant the random model an
important role in a few crucial circumstances during life's history (and such moments
make all the difference in contingent sequences, as Jimmy Stewart discovered when
his guardian angel showed him the alternate history of his town, absent his existence,
in It's A Wonderful Life): some groups, formerly dominant in their habitats and
therefore viewed as persistently "major" in our conceptions, fluctuated enormously in
diversity throughout geological time, and just happened to be surviving at very low N
(a situation from which they had always rebounded before, and in normal times)
when a mass extinction intervened. And when an event of latest Permian magnitude
occurs—the largest of all mass dyings, with estimates of species loss ranging up to 96
percent (see Raup, 1992)—and your group contributes only a lineage or two to the
global fauna, you can easily disappear, entirely and forever, for little reason beyond
bad luck (those two red beans, both included among the 96 percent of dead benthic
taxa, but entirely equivalent in ordinary Darwinian prowess to the 4 percent that
survived).
Among several plausible cases in this mode, the two specifically cited by
inherent problems, see pp. 236-248), unless its conclusion played a central role in his full
system, and he could devise no other path toward a desired outcome. Hence, and in any
case, in thus de-emphasizing the themes of geological surrogacy, I am only paying homage
to Darwin's own preferences in the structuring of his theory. Fourth, disconnectedly, finally,
and entirely personally, I have done no technical paleontological research on these
geological themes (beyond some early, and largely philosophical, analysis of the concept of
uniformitarianism, including my first published paper of Gould, 1965), whereas I have
devoted my career to paleontological studies of the first two biological legs of hierarchy and
constraint. Thus, I lack the competence and the "feel" of personal expertise (or the folly of
personal engulfment) to treat this theme at comparable length, for which all readers should
rejoice (as even this footnote has hypertrophied quite enough already)!