Tiers of Time and Trials of Extrapolationism 1315
Darwin (see p. 1302) to express his confidence in gradualistic and uniformitarian
models—trilobites and ammonites as the "signatures" of the Permian and Cretaceous
extinctions respectively—may, ironically, fit this potential scenario best. Trilobites
abounded in most Paleozoic faunas (at least in the affections of fossil collectors), but
they had been reduced to only two lineages of low diversity by latest Permian times.
The death of these two, and the consequent termination of one among only four great
arthropod classes, may reveal no insufficiency of trilobite anatomy, ecology or
development, and may only record the failure of any among a very few coins to come
up heads at a weird moment that diverted 96 percent of all flips into a tailspin
towards oblivion.
The case of ammonites is both more complex and more instructive. They
suffered greatly during the last three of five great Phanerozoic dyings—latest
Permian, latest Triassic, and latest Cretaceous. They barely survived the first two,
with only one or two lineages persisting in each case, and then died entirely in the K-
T event (although their less speciose relatives, the chambered nautiloids, survived to
this day to become favored items in aquariums and shell collections, and to become,
as the "ship of pearl," Oliver Wendell Holmes's celebrated metaphor of eternity:
"Build thee more stately mansions, O my soul...").
After the first two restrictions, the ammonites reradiated mightily and became
major components of Triassic, and then of Jurassic-Cretaceous, faunas. Perhaps we
do need an "ammonite-specific" reason for the final Cretaceous death, as Ward has
shown (1992) that their latest Maastrichtian diversity remained respectable. But I
raise a different point here: Can we specify any "real" mathematical or biological
difference among reduction to 1, 2 or 0 lineages? That is, can we say that the
Cretaceous extinction was causally worse for ammonites than the preceding Permian
or Triassic events? Surely not from the numbers themselves (and we have no other
basis for such an assertion, at least at this moment of research and understanding).
We can state no causal or statistical difference among 0,1 or 2 survivors. These
results are effectively equivalent. We might as well put slips with the three numbers
in a bag and let nature choose one at random for each of the last three mass
extinctions. And yet, in a real world of genealogical history, zero vs. anything else
makes all the difference in the universe: absolute termination absolutely forever, vs.
the possibility of redemption and reassertion. In this sense as well, and again in a
sequence so influenced by contingency, effectively random removal based on small
numbers in the face of catastrophe can impact the history of life in truly fateful and
permanent ways. Principles regulate ranges of events, but particular and
unpredictable events make history.
Nonetheless, I am confident, and I think nearly all paleontologists would agree,
that the "different rules" model—a more conventionally causal view of history—
plays a far more important role in expressing the power of catastrophic mass
extinction to fracture the crucial extrapolationist premise of Darwinian central logic. I
devoted the first part of this section (pp. 1296-1303)