Selecting for Sex

Under natural selection, species adapt to their environments. When
the environment refers to a species' physical habitat, this seems
simple enough. If a species lives in the Arctic, it had better evolve
some warm fur. Under sexual selection, species adapt too, but they
adapt to themselves. Females adapt to males, and males adapt to
females. Sexual preferences adapt to the sexual ornaments
available, and sexual ornaments adapt to sexual preferences.
This can make things quite confusing. In sexual selection, genes
do not code just for the adaptations used in courtship, such as
sexual ornaments. They also code for the adaptations used in
mate choice, the sexual preferences themselves. What the physicalenvironment is to natural selection, sexual preferences are to
sexual selection. They are not only the tastes to which sexual
ornaments must appeal, but the environment to which they must
adapt.
With sexual selection, genes act as both the fashion models
and the fashion critics, both the apostates and the inquisitors.
This creates the potential for the same kind of feedback loops
that drive progress in high fashion and modern theology. These
feedback loops are the source of sexual selection's speed,
creativity, and unpredictability. Yet they also raise the classic
problem of runaway corruption in autarchies: who watches the
watchmen? How can mere genes be trusted as both selectors and
selectees in evolution under sexual selection? The world of mate
choice plays by its own rules, and though survival is a
prerequisite for mating (as it is for scholarship, fashion, and
faith), the principles of sexual selection cannot be reduced to the
principles of survival. The biologist seems to have no point of
entry into this protean wonderland where genes build brains and
bodies, which pick the genes that build the next generation's
brains and bodies, which in turn pick the genes that pick the
genes. . .
Imagine the headaches if natural selection worked that way.
Organisms would select which environments exist, as well as
environments selecting which organisms exist. Strange, unpredictable
feedback loops would arise. Would the feedback loop
between polar bears and Arctic tundra result in a tundra of
Neptunian frigidity where bears have fur ten feet thick, or a
tundra of Brazilian sultriness where bears run nude? Would
migratory birds select for more convenient winds, lower gravity,
and more intelligible constellations? Or just an ever-full moon that
pleasingly resembles an egg? Evolutionary prediction seems
impossible under these conditions. Yet this is just what happens
with sexual selection: species capriciously transform themselves
into their own sexual amusements.
Introducing sexual selection in this way is more than just an
attempt to encourage you to share my belief that it is one of the
weirdest and more wonderful of nature's phenomena. That I could achieve simply by presenting the standard catalog of sexual
selection's "greatest hits": the peacock's tail, the nightingale's song,
the bowerbird's nest, the butterfly's wing, the Irish elk's antlers, the
baboon's rump, and the first three Led Zeppelin albums. By
presenting sexual selection as a strange world of genes selecting
other genes, I have tried to provoke a different question: How
could one ever make a science out of sexual selection? Darwin
showed that sexual selection exists and documented its effects, but
it took another century before biologists had the scientific tools for
explaining why sexual selection produces certain kinds of traits
and not others. To understand how sexual selection shaped human
mental evolution, we need to become familiar with this new
toolbox of ideas and models. Let's first have a better look at
Fisher's runaway process. It is the best example of how sexual
selection exercises a power distinct from natural selection.