Natural Selection 115giraffe descendents.
Matthen [1999; 2002; 2003] offers an independent argument in favor of the
claim that natural selection can explain the traits of individual organisms. In his
scenario, there is also a two-stage process but not one that involves cumulative
selection. A single selection event can influence the composition of future gener-
ations by increasing the chances that a particular individual will have offspring
with a particular trait because a selection event can change the frequencies of
genes available. In Matthen’s [1999] example, a selection process favors a certain
traitCin a population of sheep. As a result, this increases the proportion of sheep
with traitCin the population, and makes it more likely that any given sheep will
mate with a sheep with traitC. In particular, if we consider a particular sheep
P (parent), thepriorselection event in favor of traitCincreases the probability
thatPwill mate with a sheep that has traitC. This means that there could be a
particular sheep — let’s call it Dolly — that is the offspring ofP, whose possession
of traitCis in part explained by the prior selection event.
Matthen argues that in order to defend his view, Sober would have to rely on a
contentious ontological assumption about origin essentialism — in particular, he
must assume that a particular sheep such as Dolly must come from the parents it
did come from. Sober needs to distinguish between an individuallikePolly coming
into existence with the traits that Polly has and Pollyherselfcoming into existence
with the traits that she has. If we help ourselves to this distinction, then someone
such as Sober could argue that the prior selection event in Matthen’s sheep example
only increases the probability that a sheeplikePolly will have traitC; it does not,
however, increase the probably thatPollyhas traitC. This is because in order for
it to be Polly the sheep, she must have been born to the particular parents that she
had [Kripke, 1972]. Once you have fixed on Polly’s particular parents, you have
screened off the influence of the prior selection process. Matthen (correctly) doubts
that this assumption about origin essentialism is crucial to population genetics, and
so concludes that natural selection can explain (in part) why individuals have the
traits that they do. Pust [2001] in contrast, argues in favor of origin essentialism.
Forber [2005] argues that there are two related questions that need to be dis-
tinguished. First, what is the role of selection in explaining the origin of biological
traits and second, what are the implications of this for explaining why particular
organisms have the traits that they do? He argues that selection can explain why
a trait originates if the trait is a result of a combination of factors. Forber has us
imagine two scenarios where there is a haploid organism with two relevant loci. At
one locus are alternative allelesAandaand at the other locus are alternativesB
orb. In one case, we suppose that theABcombination is the fittest and that the
aBandAbcombinations are less fit, and theabcombination the least fit. If the
population starts out at 100%ab, we can then appeal to selection to explain why
theABcombination arises in the first place. If mutations fromatoAand from
btoBare relatively rare, then selection can help explain how the population gets
toAB. This is because there is a selective advantage to eitheraBorAboverab.
so selection increases the probability that an individual (and eventually the entire