Species selection is, as we saw in Chapter 4, one kind of species sorting (compare also Vrba and Gould 1986). Trivially, species sorting is either random (unbiased) or biased. It is just an empirical question whether species sorting is biased or unbiased in any given instance. But it's plausible to think that species sorting is at least sometimes biased. The alternative would be to say that species sorting is always random, and that the MBL model accurately represents macroevolution. Although Schopf (1979) sympathizes with that view, few paleontologists would go so far. Thus, if we define species selection as biased species sorting, it should be relatively easy to document cases of species selection in nature. This is the broadest possible conception of species selection. Some theorists, however, argue that we need a narrower conception of species selection, and that it would be a mistake simply to equate species selection with biased species sorting.

One theme of this chapter is that scientists and philosophers who are interested in species selection confront a trade-off. The more we relax our view of what is required for species selection, the easier it will be to show that species selection really occurs in nature. That means that species selectionists must feel an empirical pull toward thinking of species selection in the broadest possible terms.