586 THE STRUCTURE OF EVOLUTIONARY THEORY
In this context, I ask readers to consider two points about the tripod of
necessary support for Darwinism:
First, even during the period of current orthodoxy, beginning with the
coalescence and spread of the Modern Synthesis; the three supports have never
been particularly firm, or adequately defended. The first support—restriction of
selection to the organismal level—received little explicit defense, but rather
prevailed in a fuzzy sort of way by convention in practice. Sloppy statements
implying group selection abounded in the literature (as documented in Chapter 7,
pp. 544-556), and some disciplines, notably the classical ethology of Lorenz and
his disciples, frequently cited arguments about supraorganismal selection without
understanding their consequences for Darwinian theory. This situation changed
dramatically when Wynne-Edwards (1962) advanced his explicit argument for
group selection at a predominant relative frequency—and Williams (1966) wrote
his spirited defense of the Darwinian straight and narrow by setting out the
centrality of organismal selection so forcefully.
The second support—the validation of selection as a nearly exclusive
mechanism of evolutionary change, as embodied in the adaptationist program—
received strong verbal approbation, and elegant illustration in a few cases, but won
orthodox status largely as a bandwagon effect prompted by the urgings of a few
central figures, notably Mayr and Dobzhansky, and the subsequent acquiescence of
most professionals to the assertion of such leading figures, and not to the data of
convincing demonstrations (see Chapter 7 for a detailed defense of this claim, as
embodied in my hypothesis on the "hardening" of the Modern Synthesis). In
particular, taxonomic orthodoxy just before the Synthesis (Kinsey, 1936; Robson
and Richards, 1936) regarded most geographic variation within species as
nonadaptive. The opposite opinion triumphed as the Synthesis reached a height of
prestige and orthodoxy, but few actual cases had been overturned by data. Rather,
a shifting theoretical preference led to assertions of dominant relative frequency
based on documentation inadequate to affirm either view. (I do not regard earlier
arguments for nonadaptation as inherently more cogent. On the contrary, I am
convinced that we still have no good idea about the relative frequencies of adaptive
and nonadaptive effects in geographic variation. I only claim that the shift to
adaptationist preferences resulted more from a bandwagon effect than from direct
evidence—and that this second leg of the tripod therefore never enjoyed adequate
buttressing.)
The third support—extrapolation, explicitly discussed here in terms of the
surrogate proposition of geological uniformity, and so necessary to provide a stage
that would nurture, or at least not disrupt, Darwin's hope for explaining the entire
history of life by "pure" extension of principles derived from the small and
palpable—prevailed more by assumption than by active validation, with Kelvin's
defeat and Rutherford's proof of the earth's great age read as adequate defense (a
logically insufficient argument, by the way, because time may be long, but change
still concentrated in rare catastrophic episodes). By the largely arbitrary and
contingent sociology of disciplines, paleontology