252 Tim Lewens
this does not make it trivial, because it reminds us that there are explanations of different
levels of innovative success in different nations, or among different ethnic groups, that
need not appeal to differences in culture, environment, level of investment, or social
institutions.
Jared Diamond’s explanation for the higher rate of innovation in Europe compared to
America is a little less intuitive, but it makes use of a similar populational perspective. As
Richerson and Boyd (2005: 54) put it:
Diamond argues that the greater size of the Eurasian continent, coupled with its east-west orientation,
meant that it had more total innovations per unit time than smaller land masses, and that these
innovations could easily spread throughout long east-west bands of ecologically similar territory.
The Americas are not only smaller but are oriented north-south, making it diffi cult to diffuse useful
cultivars, like maize from (say) temperate North America to temperate South America, or domesti-
cated animals in the opposite direction. As a result, the set of adaptations necessary to support
complex urbanized societies was assembled more slowly in the Americas.
In a similar vein, Boyd and Richerson explain the disappearance of important technologies
on Tasmania by reference to declining population size alone. Drawing on the work of
anthropologist Joseph Henrich, they suggest that the maintenance of technologies and the
associated behaviors required to produce and operate them may require a population that
is large enough for the rate of innovation to offset the degradation that results from error-
prone imitation (Richerson and Boyd 2005: 138). Again these hypotheses are by no means
obvious, and they arise from aggregative thinking, which prompts us to ask what number
of individuals, with fallible imitative abilities and limited innovative abilities, is required
to sustain complex technical know-how.
14.8 Population Thinking and Memes
The claim that Sober’s challenge naïvely assumes we are good intuitive population think-
ers works well as a defense of Boyd and Richerson’s views. It works far less well as a
defense of so-called memetic theories of cultural change. Here fi tnesses are assigned
directly to ideas, and sometimes to techniques or even artifacts, according to the expected
growth rate of those entities over time within a population. Suppose, for example, we
decide that Rollerblades are fi tter than roller skates, meaning that the former have a higher
long-run growth rate in the population than the latter. Once we have assigned a higher
fi tness to Rollerblades than to roller-skates it is hard to see how intuitive population think-
ing might fail us. Again the explanatory interest lies not in seeing how the consequences
of these different fi tnesses play out at the population level; it lies in seeing what makes
Rollerblades fi tter than roller-skates.
Boyd and Richerson are not defenders of memetics. They do not attempt to assign
reproductive fi tnesses directly to entities like ideas or artifacts. As we have seen, they are