her father's wrath; or she plays an immoral
game to dissemble and win his affection. She tumbles into this plight because
Lear cannot conceptualize the proposition that Cordelia's silence might
signify her greater love--that nothing can be the biggest something.
Cordelia's
dilemma is deeper and more interesting than publication bias, is we glimpse
the constraining role of neurological, social, and psychological conditioning
in our struggle to grasp this complex universe into which we have been
so recently thrust. Publication bias is only a guard at the party door
giving passage to those with the right stamp on their hands. At least
the guard can see all the people and make his unfair decisions. Those
rejected can gripe, foment revolution, or start a different party. The
victims of Cordelia's dilemma are "unpersoned" in the most Orwellian
sense. They are residents in the last gulag in inaccessible Siberia, the
last outpost of Ultima Thule. They are not conceptualized and therefore
do not exist as available explanations.
These two forms of nonreporting have different solutions. Publication
bias demands, for its correction, an explicit commitment to report negative
results that appear less interesting or more inconclusive than the "good
story" of positive outcomes. The solution to Cordelia's dilemma--the
promotion of her nothing to a meaningful something--cannot be resolved
from within, for the existing theory has defined her action as a denial
or non-phenomenon. A different theory must be imported from another context
to change conceptual categories and make her response meaningful. In this
sense, Cordelia's dilemma best illustrates the dynamic interaction of
theory and fact in science. Correction of error cannot always arise from
new discovery within an accepted conceptual system. Sometimes the theory
has to give first, and a new framework be adopted, before the crucial
facts can be seen at all. We needed to suspect that evolution might be
true in order to see variation among individuals in a population as the
dynamic stuff of historical change and not as trivial or accidental deviation
from a created archetype.
I am especially interested in Cordelia's dilemma, and its resolution
by using new theories to promote previously ignored phenomena to conceivability
and interest, because the "main event" of my early career included
an example that taught me a great deal about the operations of science.
Before Niles Eldredge and I proposed the theory of punctuated equilibrium
in 1972, the stasis, or nonchange, of most fossil species during their
lengthy geological lifespans was tacitly acknowledged by all paleontologists,
but almost never studied explicitly because prevailing theory treated
stasis as uninteresting nonevidence for nonevolution. Evolution was defined
as gradual transformation in extended fossil sequences, and the overwhelming
prevalence of stasis became an embarrassing feature of the fossil record,
best left ignored as a manifestation of nothing (that is, nonevolution).
My
own thesis advisor had mastered statistics in the hopes of detecting a
subtle gradualism that was not visually evident in fossil sequences. He
applied his techniques to some fifty brachiopod lincages in Silurian rocks
of the Michigan Basin, found no evidence for gradual change (but stasis
in all lineage with one ambiguous exception), considered his work a disappointment
not even worth publishing, and left the field soon thereafter (for a brilliant
career in another domain of geology, so our loss was their gain).
But
Eldredge and I proposed that stasis should be an expected and interesting
norm, and that evolution should be concentrated in brief episodes of branching
speciation. Under our theory, stasis became interesting and worthy of
documentation--as the norm that rare events of change disrupt. We
took as the motto of punctuated equilibrium: stasis is data. (One might
quibble about the grammar, but I think we won the conceptual battle.)
Punctuated equilibrium is still a subject of lively debate, and some
(or most) of its claims may end up on the ash heap of history, but I take
pride in one success relevant to Cordelia's dilemma: our theory has brought
stasis out of the conceptual closet. Twenty-five years ago, stasis was
a non subject--a "nothing" under prevailing theory. No one
would have published, or even proposed, an active study or lineages known
not to change. Now such studies are routinely made and published, and
we have a burgeoning literature to document the character and extent of
stasis in quantitative terms.
Punctuated equilibrium is a theory about the origin and history of species.
That is, the stability of individual species represents the "nothing"
that our theory emphasized to attract the attention of researchers. A
different kind of "nothing" permeates, and also biases, our
consideration of the next most inclusive level of evolutionary stories--the
history of phyletic bushes, or groups of species sharing a common ancestry:
the evolution of horses, of dinosaurs, of humans, for example. This literature
is dominated by the study of trends--directional changes through time
in average characteristics of species within the bush. Trends surely exist
in abundance, and they do form the stuff of conventional good stories.
Brain size does increase in the human bush; and toes do get fewer, and
bodies bigger, as we move up the bush of horses.
But the vast majority of bushes display no persistent trends through
time. All paleontologists know this, but few would ever think of actively
studying a bush with no directional growth. We accept that the history
of continents and oceans presents no progressive pattern most of the time--"the
seas come in and the seas go out" in an old cliché of geology
teachers from time immemorial. But we expect life's bushes to grow toward
the light, to tell some story of direction change. If they do not, we
do not feature them in our studies--if we even manage to see them
at all. We cannot accept for life the preacher's assessment of earthly
time (Ecclesiastes |