120 Christopher Stephens
If the model makes precise predictions, it is no longer a trivial matter to make up
a theory or model that fits the facts. Furthermore, one can often build into one’s
model various kinds of constraints, thereby dealing with various concerns about
certain sorts of mechanical and engineering constraints. It should also be noted
thatnon-adaptationistexplanations are easy to make up too, which means that
precise models are needed to testalternativesto adaptationism as well.
Of course, critics of adaptationism point out that this still leaves room for
variousad hocmaneuvers that making testing adaptationist claims more difficult.
Dawkins discusses such a case in considering an appeal to a “time lag” between
present and past conditions to explainaway data that does not fit a model. Lack
[1966] used this approach when discussing optimal clutch size. In a case where
the data didn’t match up to his theory, Lack appealed to the claim that the
conditions under which the clutch size evolved were different from the present
circumstances, and so there had not yet been time for the trait to evolve to meet
the new environmental conditions. Dawkins claims that although it is post hoc, it
is at least still testable in principle. Information about the food conditions in the
time period during which the trait evolved would help resolve the issue.
The second major tool that adaptationists avail themselves to is the compara-
tive method [Harvey and Pagel, 1991; Sterelny and Griffiths, 1999; Sober, 2000].
The main idea behind the comparative method is to examine correlations in many
species between two more traits or between a trait and some environmental vari-
able. For instance, if increasing neck length is correlated with environments that
have tall trees, this correlation can be studied in close relatives of the giraffe to
get more data. In some cases, one discovers historical facts that undermine an
otherwise plausible adaptationist hypothesis. Some biologists thought that the
low birth weight found in bears was a by-product of the evolution of hiberna-
tion. It turns out, however, that low birth weight evolvedbeforehibernation, and
appears on branches of the evolutionary tree in which hibernation never evolved
[McKitrick, 1993; Sterelny and Griffiths, 1999].
Peter Godfrey-Smith [2001] argues that there is a third kind of adaptationism,
in addition to its empirical and methodological versions. According to Godfrey-
Smith,explanatory adaptationismis the claim that the most important question
in evolutionary biology is how to explain apparent design, and that natural selec-
tion is the “big answer” to this question. This is distinct from empirical adapta-
tionism because one could think that natural selection is the answer to the “big
question” about apparent design without thinking that natural selection is partic-
ularly pervasive or powerful in overcoming constraints. Godfrey-Smith claims that
explanatory adaptationism is the kind of adaptationism responsible for the most
difficult conceptual problems surrounding the adaptationism debate. He suggests
that Dawkins and Dennett are clearly explanatory adaptationists and may not be
empirical adaptationists. Dawkins [1983, 16], for instance, defends explanatory
adaptationism when he says that explaining adaptive complexity is the main task
of any theory of evolution. He also says that “I would not presume to try to con-
vert any of these people to my point of view” (Dawkins 1983, 16), which suggests