118 Christopher Stephens
planations that aren’t supported by much evidence. It is easy to make up a story
about how a traitmighthave been favored by natural selection and much harder
to get evidence in favor of a particular adaptationist explanation. It may seem
obvious that long necks in giraffes are an adaptation for reaching leaves high in
trees, but if it turns out that long necks evolved in an environment without tall
trees, this explanation would be disconfirmed.
Some adaptationists, such as Richard Dawkins [1982; 1983] and Daniel Dennett
[1995], defend their view by pointing out that natural selection is the only process
capable even in principle of explaining adaptive complexity. Dawkins argues, for
instance, that Lamarckian methods of use and disuse plus the inheritance of ac-
quired characteristics cannot explain even in principle how a trait such as the eye
could have evolved, since there is no way that the use of a proto-eye could change
its properties to make it more transparent and hence better adapted. It should be
noted that however indispensable natural selection might be in order to explain
adaptive complexity, this response does not answer the methodological challenges
raised by Gould and Lewontin.
There is considerable dispute about how exactly to formulate adaptationism
[Maynard Smith, 1978; Lewontin, 1979; Dawkins, 1982; Orzack and Sober, 1994;
Dennett, 1995; Brandon and Rausher, 1996; Sober, 1996; Amundson, 2001; Godfrey-
Smith, 2001]. The adaptationism debate has often involved caricatures and mis-
representations of the opposing side. Lewontin (1979), for instance, describes
adaptationism as “that approach to evolutionary studies which assumes without
further proof that all aspects of the morphology, physiology and behavior of organ-
isms are adaptive optimal solutions to problems.” As described, this is probably
not a position held by anyone, and makes adaptationism trivially false. There is,
however, broad agreement that there are at least two different kinds of adaptation-
ism. On the one hand there is what is sometimes called Empirical Adaptationism
[Sterelny and Griffiths, 1999; Sober, 2000; Godfrey-Smith, 2001].Empirical Adap-
tationismis usually defined as the view that natural selection is the most or only
important cause of most traits in most populations. The thesis may be stated
more or less strongly, depending on how powerful and how pervasive one claims
that natural selection is. Anti-adaptationists, by contrast, often emphasize the
role of constraints in limiting the power of natural selection.
Consider a case of heterozygote superiority. Suppose we have a diploid organism
that can have either aAoraallele at a given locus. That is, each organism in
the population has either aaaorAAhomozygote or aAaheterozygote. Suppose
further that the heterozygote codes for the fittest trait. Sickle-cell anemia is a
famous case of this phenomenon. Humans with one copy of theaallele acquire
a resistance to malaria but still have healthy cells. People with two copies of the
aallele develop a serious disease known as sickle-cell anemia. AAhomozygotes
have neither the resistance to malaria nor the anemia. The heterozygote is the
fittest trait, but heterozygotes don’t “breed true”, since half of the time when
two heterozygotes mate their offspring willnotbeAa. This means that natural
selection cannot cause the fittest trait to sweep to 100% of the population because