Innovation and Population 249
Remember that the essential feature of Darwin’s theory of evolution is population thinking.... All
of the large-scale features of life—its beautiful adaptations and its intricate historical patterns—can
be explained by the events in individual lives that cause some genetic variants to spread and others
to diminish. (Richerson and Boyd 2005: 59)
Population thinking, on this view, is really nothing more than what we might call “aggre-
gative thinking.” It is the kind of thinking one engages in when one explains the behavior
of a unit composed of varied parts in terms of the properties of those parts and their inter-
actions. An approach of this sort seeks to explain population-level phenomena in terms of
individual-level, rather than population-level, properties:
The processes that cause... cultural change arise in the everyday lives of individuals as people
acquire and use cultural information.... In the short run, a population-level theory of culture has to
explain the net effect of such processes on the distribution of beliefs and values in a population
during the previous generation. Over the longer run, the theory explains how these processes,
repeated generation after generation, account for observed patterns of cultural variation. The heart
of this book is an account of how the population-level consequences of imitation and teaching work.
(Richerson and Boyd 2005: 6)
So population thinking for Richerson and Boyd is all about explaining how population-
level patterns emerge from the collective behavior of the diverse individuals that make up
the population. Some of their broad methodological statements might appear to exclude
attributing causal powers to populations in their own right; however, population-level
properties can feed back to the individual. An example from a traditional organic selection
model might be when the reproductive success of an individual with trait T depends on
the frequency of T in the population. An example from a cultural evolutionary model might
be when the chance of an individual coming to believe that P depends on the proportion
of people in the population at large who believe that P.
A similar population-based methodology features in the work of evolutionary econo-
mists Richard Nelson and Sidney Winter. They have much in common with so-called
behavioral economists (see, once again, Kahneman, Slovic, and Tversky 1982), specifi -
cally regarding the rejection of the rationality assumptions of classical economics. However,
they explain how they differ from behavioral economists in the following way:
We diverge from the behavioral theorists in our interest in building an explicit theory of industry
behavior, as contrasted with individual fi rm behavior. This means on the one hand that our charac-
terizations of individual fi rms are much simpler and more stylized than those employed by the
behavioral theorists, and on the other hand that our models contain a considerable amount of appa-
ratus linking together the behavior of collections of fi rms. (Nelson and Winter 1982: 36)
It is because of the potentially counterintuitive nature of the aggregation of individual-level
events that population thinking of this sort has value. This general moral underlies the
work of Boyd and Richerson as well as Nelson and Winter. Consider this simple illustra-
tive example, which Richerson and Boyd draw from the domain of technological change