Structural Constraints, Spandrels, and Exaptation 1223
subject, which was not treated at sufficient length in the former editions of this
work."
I will not describe Darwin's solutions to the problem of "incipient stages" at
length, because I have already done so in Chapter 2, my exegesis of the Origin of
Species. But I do need to emphasize the relevant point for this chapter—that both of
his effective arguments against Mivart's supposed trump card invoke structural
principles of constraint rooted in Nietzsche's "major point of historical method": the
discordance between historical origin and current function.
In his first argument, Darwin readily admits the "5 percent of a wing" problem,
and then presents the incisive solution that becomes enshrined in evolutionary theory
(by no fault of Darwin, who never used the term) under the most unfortunate name of
"preadaptation." Yes, five percent of a wing offers no conceivable aerodynamic
benefit, and could not therefore either be formed, or converted into a full wing, under
a smooth regime of natural selection for flight. But sequences forged by selection
only presuppose continuity in differential reproductive success, not continuity in a
single function. Thus, the incipient stages may have performed a different function,
for which their 5 percent of a wing imparted benefits. Eventually, the enlarging proto-
wing entered the domain of aerodynamic benefit, and the original function changed to
the primary utility now exploited by most birds. Current function cannot be equated
with reasons for historical origin. Mivart's cardinal objection disappears, thus
explaining why "it is so important to bear in mind the probability of conversion from
one function to another."
Darwin roots his second argument in the related, but even more generalized,
structural principle of redundancy—the inherent capacity (based on intrinsic structure
rather than current function) of most organs to work in more than one way (either at
the same time, providing dual benefits, or with one utility overtly exploited by natural
selection, and the other latent, providing unselected flexibility for future change).
Darwin presents this argument in a fascinating manner by coupling two apparently
opposite facts about redundancy: that a single function can be performed by more
than one organ, and that a single organ can perform more than one function. Thus, an
organ need not invent an entirely new function in some mysterious manner, but may
evolve by intensifying a previously minor use, or even by recruiting an inherent but
unexpressed potential. Meanwhile, the modified organ can abandon its previous
major function because other organs can continue (or intensify) their former
operation in the service of the same necessary task.
Thus, reptilian jawbones can become mammalian ear bones because they
already played some role in sound transmission while they functioned primarily to
articulate the jaw of therapsid forebears (the principle of two functions for one
structure). They then become free to move into the middle ear because the
transitional forms (as demonstrated empirically by such fossils as Diarthrognathus,
and not only as a reasonable conjecture) possessed a double jaw joint (the reciprocal
principle of two structures for one function)— and the bones of the old quadrate-
articular joint could then become the