Natural Selection 117and then be favored by natural selection for another reason (dominance displays).
Gould and Vrba [1982] refer to this phenomenon as an “exaptation” and argue
that traits such as a long neck are really only adaptations for why they were
firstfavored by natural selection, and are merely exaptations for the subsequent
behavior. I see no reason, however, to restrict the term adaptation in such a way.
Surely a trait can be an adaptation both for why it was first selectedandfor why
it is maintained by selection [Sterelny and Griffiths, 1999]. At the same time,
Gould and Vrba are right to remind us that one cannot simply assume that the
reason that a trait is favorednowor in the recent past is the same reason that it
was favored by selection in the more distant past.
The central debate about adaptationism concerns thepowerandscopeof nat-
ural selection. Everyone agrees that traits like eyes and wings were favored by
natural selection for some reason, though there are disputes about exactly why
such traits were favored. Biologists also agree that natural selection is an impor-
tant process in evolution. There is considerable disagreement, however, about the
significance of non-selective factors such as random genetic drift and the role of
mechanical, developmental, genetic, and other sorts of constraints. Is natural se-
lection powerful enough so that it can overcome constraints and other non-selective
processes? Can model builders effectively ignore factors such as drift, at least in
most circumstances?
Although debates about the importance of natural selection go back to Darwin
and Wallace, the contemporary discussion is in large part a reaction to Gould and
Lewontin’s [1979] famous paper “The spandrels of San Marco and the Panglossian
paradigm”. In their paper, Gould and Lewontin raise two main kinds of objections
to adaptationism — empirical and methodological. On the empirical side, they
claim that many biologists tend to confuse adaptiveness (current utility) with
adaptation by assuming that a trait must have been selected in the past for what
it does well now. They also charge that adaptationists tend to “atomize” organisms
— thinking that each trait of an organism can be assumed to evolve independently
from its other traits. The number of digits on one hand should not be understood
as a trait that evolves independently from the number of digits on our other hand.
This is an extreme case, of course, but anti-adaptationists tend to think that there
are many such constraints and correlations between traits, whereas adaptationists
tend to think there are not.
In addition to disagreement about empirical issues, adaptationists and their
critics have methodological disputes as well. Gould and Lewontin charge that
adaptationists, when faced with the failure of one adaptive explanation, tend to
immediately come up with an alternative adaptationist explanation rather than
look to a non-adaptationist explanation. In particular, Gould and Lewontin want
to encourage biologists to pursue the idea of a body plan (bauplan) and to take
seriously physiological constraints that limit the power and scope of natural selec-
tion.
Gould and Lewontin’s most important methodological criticism of adaptation-
ism is that adaptationists tend to make up “just-so stories” — how-possibly ex-