Demography and Cultural Evolution
Media
Part of Demography and Cultural Evolution
- Title
- Demography and Cultural Evolution
- extracted text
-
Demography and Cultural Evolution
STEPHEN SHENNAN
Abstract
Trying to explain the increase in cultural complexity over the long term of human
history has long been an interest of anthropology and of historical social sciences
more generally. In recent years, interest has grown rapidly in the idea that a key
factor in accounting for it might be the size of the human population itself and the
extent of interaction between people, because of the effect these have on the innovation rates in populations and on the success with which innovations are transmitted.
An important driver of this growth of interest has been the emergence of the new
interdisciplinary field of cultural evolution, which makes extensive use of mathematical techniques, especially methods derived from population genetics. The result has
been the development of a range of analytical and computer simulation models that
make various predictions about the way in which population size influences cultural
change, and in particular the growth of cumulative culture, including the processes
that have led from the very simple forms of culture possessed by other great apes
to those characteristic of Homo sapiens. The aim of this review is to distinguish them,
so that future work can focus on evaluating their strengths and weaknesses and the
circumstances in which they are useful.
INTRODUCTION
The past decade or so has seen an enormous increase in interest in the relationship between population patterns and processes on the one hand and
patterns of cultural change and adaptation, on the other; in particular, the
extent to which population size and interaction rates have an effect on innovation rates and on the growth of cultural complexity or cumulative culture.
Cumulative culture refers to the idea that over time the number of cultural
elements present in human societies tends to increase, and that the presence
of complex cultural or technological innovations, for example, the internal
combustion engine, requires the prior existence of a number of other traits.
A major driver of this new interest has been the development of the interdisciplinary field of cultural evolution (Boyd & Richerson, 1985; Cavalli-Sforza
& Feldman, 1981), which takes a Darwinian theoretical perspective on cultural variation and change, seeing it in terms of innovation, transmission, and
Emerging Trends in the Social and Behavioral Sciences. Edited by Robert Scott and Stephen Kosslyn.
© 2015 John Wiley & Sons, Inc. ISBN 978-1-118-90077-2.
1
2
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
selection processes. Associated with this perspective has been the application
to studies of cultural data of methods derived from evolutionary biology,
especially population genetics. More specifically, the development of this
new field has led to the growth of interest in exploring the existence or otherwise of cultural traditions in animals, not least our closest living relatives, the
great apes (e.g., Whiten et al., 1999). Much if not most of the work on the relationship between population and cultural complexity has been of a generalizing nature, based on the construction of relatively abstract models, though
there has been a particular interest in the long-term evolutionary question
of how the cultural complexity of present-day human societies could have
emerged from the much simpler culture of other nonhuman species.
There is now a range of different models and suggestions in the literature relating to the role of population processes in cultural accumulation
and change, and the aim of this review is to distinguish them, so that future
work can focus on evaluating their strengths and weaknesses and the circumstances in which they are useful. An initial distinction can be made between
studies that have placed particular emphasis on the detailed processes of
cultural transmission and innovation and those that pay less attention to
mechanisms and look more at general trends. I will begin with the latter,
which have a longer history, but the vast majority of research on this topic
has only been published in the twenty-first century and is still the subject of
intense debate, hence the distinction between foundational and cutting-edge
research is a difficult one to make.
FOUNDATIONAL RESEARCH
MACROSCALE ECONOMIC MODELS
At the end of the eighteenth century, Thomas Malthus proposed that the
growth of human populations rapidly outstrips available resources and that
they are held in check by disease, famine, and warfare. As many have pointed
out since, while this was true in the past, Malthus was writing at a time when
the situation was changing. The Industrial Revolution and subsequent technological developments have meant that resources have been able to keep
up with population, which has risen to unprecedented levels. Marx criticized Malthus on these grounds, and much more recently, Boserup (1965)
proposed that the deterioration in living conditions produced by “population pressure” created incentives in human societies to innovate to overcome
existing limits. Kremer’s (1993) model of long-term technological evolution
and its relation to population size is in the spirit of Malthus and Boserup.
Population is limited by the carrying capacity of the environment given the
availability of a specific technology. Technological innovation can increase
Demography and Cultural Evolution
3
the carrying capacity, and population will then rise to a new limit. However,
the rate of innovation is not independent of population size. Other things
being equal, a larger population will contain more innovators, so there will
be a feedback relationship between technology and population size, with one
dependent on the other. This further implies that the rate of growth of a population at a given time is proportional to its size at that time, in contrast to
models where technological improvement is independent of population size,
determined by other factors. Kremer shows that, at least until recently, the
predictions of this model are consistent with long-term human population
history, and that it can be extended to a more general version that factors
in variations in research productivity, which can additionally account for
the recent decline in population growth rates with increased incomes. He
also considers the geographical implications of his model, pointing out that
for the same level of technology and population density, regions with larger
areas and therefore larger starting population sizes will have faster rates of
technological change and over a given period will reach higher levels of population and technology. Again, he shows that this prediction is met by data
on estimated population sizes of different regions of the world at 1500 AD
just before European expansion into the New World. Finally, he notes that
his model does not include the possibility of technological loss, as appears to
have occurred in prehistoric Tasmania after it was cut off from the Australian
mainland, potentially leading to a negative spiral in technology and population. Kremer ascribes the relationship between technological complexity and
population size to the “nonrivalry” of technology, the fact that one person’s
use of it does not preclude the use by others.
Richerson, Boyd, and Bettinger (2009) also develop an economic model
within the Malthus–Boserup framework, proposing, like Boserup, that as
population approaches a given carrying capacity and individual income
goes down, people will innovate and intensify, but also suggesting, in line
with the diet-breadth model in behavioral ecology (e.g., Kaplan & Hill,
1992), that if incomes increase, for example, as a result of the appearance of
new resources, perhaps as a result of climate change, people can also “deinnovate” or deintensify, at least for short periods until population catches up.
Apart from such short periods, “population pressure” is always present; so
long-term population trends depend on the rate of innovation that increases
the environmental carrying capacity. Interestingly, the earliest evidence we
have for pressure on resources is only 50,000 years ago, with indications
of the overexploitation of tortoises in the Middle East (Stiner, Munro, &
Surovell, 2000), and Richerson et al. (2009, pp. 223–224) suggest that before
this, the limits to human population size might have come from competition
with other carnivores, postulating social and technical innovations that
could have made modern humans more successful competitors. They go on
4
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
to suggest, following Marx, that, at least in the Holocene, the limiting factor
on cultural evolution may have been the rate of social rather than technical
innovation, and conjecture that at the broadest social and temporal scale, a
model in which the innovation rate is dependent on the population size will
lead to exponential growth in technological sophistication (2009, p. 228).
CUTTING-EDGE RESEARCH
MACROSCALE INNOVATION-TRANSMISSION MODELS
Ghirlanda, Enquist, and Perc (2010) take a rather different approach to the
relation between population and culture by focusing more closely on the
processes involved. To understand the relationship between population and
culture, we need to consider the balance between individual creativity and
the effect of cultural innovations on environmental carrying capacity, on the
one hand, and the rate of cultural loss, on the other. If the former outstrips
the latter, then cultural accumulation or complexity, together with the population size, can increase. This can arise as a result of an increased effect of
culture on carrying capacity, increased innovation or increased transmission
fidelity. Ghirlanda et al. also introduce the possibility that an innovation that
initially increases carrying capacity may not continue to do so, for example,
the use of firearms for hunting may initially increase returns but then lead to a
decrease as resources are overexploited. They show that varying the effect of
culture on carrying capacity and the impact of what they call the corruption
rate, the rate at which adaptive features become nonadaptive, can result in
very different dynamics, ranging from stability to unlimited growth to population extinction. Unsurprisingly, the introduction of a parameter that allows
for “adaptive filtering,” the ability to detect and discard features that have
become maladaptive, increases the space in which unlimited growth occurs.
Finally, they return to the question of the hyperbolic rate of human population increase over the long term, generally assumed, as in Kremer’s model,
to arise from the fact that innovations increase both the environmental carrying capacity and the rate of population increase, because larger populations
have more innovations, and suggest a possible alternative explanation that
the rate of increase in innovations with population size may be faster than linear. The relevance of this suggestion is borne out by the study of Bettencourt,
Lobo, Helbing, Künhert, and West (2007) showing that the rate of increase
in innovations in cities is not a linear function of the population size but of
population size to the power 1.2, which leads to growth that is faster than
exponential. On the other hand, it is clear that until very recent times, cities
were population sinks, with low birth rates and high death rates, maintained
only by large-scale immigration from the countryside [Knauft, 1986, cited by
Demography and Cultural Evolution
5
Richerson et al. (2009), cf. Barnes, Duda, Pybus, & Thomas, 2001]. How the
accumulation of innovations and increased population would have emerged
from these different forces remains to be explored.
The long-term mutual dependency of technology and population size is
clear from all the studies outlined so far; however, the view of technology, or
adaptive culture, represented, important though it is, is a one-dimensional
one and does not consider what we might call the internal dynamics of
cumulative culture itself. As pointed out earlier, cumulative culture refers to
the idea that, beyond the very simplest level, cultural attributes generally
depend on one another, in the sense that the presence of one cultural attribute
affects the likelihood that another will be present; the strongest cases being
those where one is a prerequisite for the other, and, at the other extreme,
where one precludes the other. Enquist, Ghirlanda, and Kimmo Eriksson
(2011, pp. 415–418 and Figure 1) show that there is variety of ways that such
dependencies can arise: directional change in a single dimension, where to
get from state (a) to state (c) you have to go through state (b); branching
differentiation, where a given state can give rise to two or more states, which
then change independently of one another; pairwise combinations, where
the state of one attribute is affected by the state of two others; and finally,
“systems of cultural elements” in which there may be multiple influences,
positive and negative, between attribute states, including reciprocal ones.
Unsurprisingly, simulation studies show that “systems of cultural elements”
result in cultures whose number of elements increases at a far faster rate than
any of the others and also produce corresponding increases in the histories of
individual elements, that is to say, “the number of evolutionary events that
created the element, starting from a cultural seed” (Enquist et al., 2011, p. 418,
Figure 5). Systems characterized by these complex interdependencies are
strongly path-dependent, both because of the very large number of possible
links and because there is a strong stochastic element as the links are not
deterministic ones. The authors argue that there is strong evidence in many
domains for the predicted exponential increase in the diversity and complexity of culture (Enquist, Ghirlanda, Jarrick, & Wachtmeister, 2008, Figure 1). In
summary, the rate of accumulation of culture at a given point depends on its
rate at the previous point purely as a result of the increase in cultural interdependencies. However, as Enquist et al. emphasize, this is a very abstract
model and it does not address transmission processes at the individual
level, nor does it consider the demographic aspects that they and others
have shown elsewhere to be important, as we have seen earlier. The role
of selection in winnowing out ineffective combinations of traits is also not
considered.
6
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
MODELS OF SPECIFIC INNOVATION-TRANSMISSION PROCESSES
The best-known model that looks at the implications of specific innovation
and transmission processes and their link to demography is that of Henrich
(2004). This is based on a model of learning in which individuals attempt to
imitate the most skilled individual available to them. Most individuals will
fail to reach that level, so there will be a distance between the modal skill
level in the population and the best. However, there will be a certain probability that an individual will exceed the current best and a new, higher, most
skilled level will result. In these circumstances, given a constant gap between
the best level and the modal level, the result will be an improvement in the
modal level. The probability of improvement or decline in the population
modal level depends on the ratio of the gap between the mode and the best,
on the one hand, and the dispersion of the distribution of skill levels on the
other. If the gap is small and the dispersion is large, then improvement will
be far more probable than in the opposite case. For a given ratio, though,
Henrich demonstrates that what happens will depend on population size, as
improbable events will occur more often in larger populations. For very complex technologies where the gap between the modal level and the best is large
and the variation between individuals is small, then a large population will
be required to maintain or improve the technology. Thus, if external forces
cause a decline in the population size, the most complex technologies are
likely to be lost. Simpler ones, even if they are lost, have a much higher probability of being reinvented. Henrich showed that this process could account for
the well-known and apparently deeply puzzling cultural impoverishment
of the aboriginal Tasmanians after Tasmania was cut off from the Australian
mainland at the end of the past Ice Age. More generally, “It is the selective
transmission of lucky errors and occasional experiments that drives much of
the evolution of adaptive technology, skills, beliefs, and practices” (Henrich,
2004, p. 202), and makes it strongly dependent on population size. Henrich’s
result has subsequently been confirmed by Kobayashi and Aoki (2012), who
showed that a version of the model with the more satisfactory assumption
of overlapping rather than discrete generations produced an even stronger
population effect.
The substantive assumption in Henrich’s argument that real improvements
are difficult to make because of the sheer amount of cultural knowledge
embedded and embodied in human artifacts and cultural practices is
supported by the ethnoarchaeological work of Roux (2010), who makes the
point that training in any production process that involves the acquisition
of high levels of expertise that take a long time to acquire is likely to
restrict innovation, because the whole process of learning is designed to
fix particular sets of skills and knowledge; this may be particularly the
Demography and Cultural Evolution
7
case with physical expertise and motor skills. Only the most expert, those
who have complete control of all aspects of a process and its associated
knowledge, are likely to transcend the limits of what they have learned and
invent something new. Roux cites (2010, p. 224) the example of inventions
in the field of pyrotechnology by traditional craftsmen in India, which were
only made by those who were most skillful, individuals “exceptional as
much for their skills as for their rarity.”
It is important to note that the population size in such cases refers to the
“effective population size,” not the total local population but the relevant
interacting population of learners and teachers. Thus, in the case of Henrich’s
proposal for Tasmania, the interacting population size decreases because of
the loss of contact with the mainland, not because of local population decline,
though this probably occurred. Where there is craft specialization, the effective population size may be tiny and the skills are correspondingly vulnerable to loss. Thus, Roux (2010) shows that despite the productivity of the
technique, the wheel-coiling of pottery was lost on two occasions in the societies of the ancient Levant because it was a highly specialized activity with
only a small number of practitioners. When the socioeconomic conditions
that sustained it collapsed, a small number of transmission links were broken. It was only later, as the number and spatial extent of transmission links
increased, that this technical system became less vulnerable to the effects of
external historical events.
Powell, Shennan, and Thomas (2009) simulated a version of Henrich’s model and effectively ran it in reverse, to explore the effects of
increased population density and increased migration between groups in
a meta-population on increases in cultural complexity. They showed that
differentials between regions in terms of population density or migration
rate could sustain very substantial differences in the skill level as measured
in terms of Henrich’s model, and proposed that these factors could account
for the puzzling appearance and disappearance of features of the so-called
behavioral modernity in the African Middle Stone Age, features that had
previously been accounted for in terms of biological cognitive evolution, if
human populations fluctuated in size or contact rates as a result of changing
climate patterns. However, others have emphasized the importance of
evaluating the costs and benefits of more and less complex technologies
in different environments (Mackay & Marwick, 2011). In any case, it is
worth emphasizing that neither Henrich’s original model and, Powell et al.’s
version, nor Kobayashi and Aoki’s modification includes any feedback
from the improving or devolving technology itself to the population size.
Population size is simply an independent variable affecting the innovation
and transmission of novel improved ways of doing things.
8
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
A different model of the relationship between cultural complexity and population size was explored by Shennan (2001). He modified a genetic model
developed by Peck, Barreau, and Heath (1997) on the basis of the ideas of
Fisher (1930). Individuals have a number of different attributes or traits that
are subject to selection depending on the values of their attribute states. The
model starts with a perfectly fit population. Cultural transmission occurs,
with the parental value of a cultural trait having a certain probability of not
being passed on to its offspring; when this occurs, the individual chosen to
be the cultural model is the one with the highest adaptive value chosen from
a random selection of individuals in the parental generation; otherwise, the
offspring takes the parent’s value. In the course of transmission innovations
can occur, which can be either beneficial or deleterious. The probability that
any given attribute state adopted by an offspring will have just undergone an
innovation is represented by the attribute innovation rate. An innovation is
assumed to change the adaptive value associated with an attribute by some
amount, drawn from a distribution such that the majority of innovations
have very little effect but some have a more significant one. The results of
the simulations carried out using this model showed that larger populations
can evolve to a higher average fitness than smaller ones, because they carry
a smaller drift load of deleterious cultural traits, though again there was no
feedback to population size.
A similar model has recently been used by Lehmann and Wakano (2013) to
look at the effect of the multidimensionality of cultural traits; in other words,
where the best payoff depends on getting the right combination of a number
of different features, for example, different aspects of size and shape of an
artifact. In Henrich’s model, everything is collapsed down to a single dimension. With increasing numbers of dimensions, there are far more ways of
making things worse than making them better. Where the number of dimensions is very high, then even a learner with very high computational ability
going through different combinations and their effects in their imagination
will not find the optimum, but a population of payoff-biased learners making
uncorrelated errors will be able to reach it if it is sufficiently large. An excellent example of the kinds of complexity and multidimensionality involved is
provided by Lombard and Haidle’s (2012) construction of what they call the
“effective chain” of activities for the production of bows and arrows, based
on evidence from Middle Stone Age Africa.
If we think about the different kinds of dependencies between cultural elements described by Enquist et al. (2011) and discussed earlier, then clearly the
case where there are “systems of elements” in which there may be multiple
influences, positive and negative, between attribute states, and whose numbers of elements increase very rapidly as a result, will also be the case where
finding an optimum is the most difficult. Once again then, from a different set
Demography and Cultural Evolution
9
of premises, successful innovation, and therefore adaptive cumulative culture, depends on population size. The more complex the system of elements
becomes, the steeper is the rate of increase in cultural complexity, as Enquist
et al. (2011) show; correspondingly, the larger the population size needs to be
in order to have a high probability of arriving at the optimum combination
of trait values.
The general argument about the importance of population size for cultural
complexity is supported by Kline and Boyd’s (2010) study of marine fishing
technology in Oceania. This showed a correlation between the number and
complexity of marine fishing tools and the size and connectedness of island
populations, with smaller and more isolated islands having fewer and less
complex tools than larger and less isolated ones. Importantly, they point out
that the pattern is not entirely consistent with the type of Malthus–Boserup
model proposed by Kremer in which improved technology produces
increased carrying capacity and therefore increased population because this
model does not explain why the connectedness of the islands, not just their
local population size, also has an effect. They proposed that the pattern
could be accounted for by either a drift model of the kind proposed by
Shennan or Henrich’s process (which they call the treadmill model); however,
in any case, a model in which the transmission process has a significant role.
ALTERNATIVE MODELS OF LINKS BETWEEN CULTURAL COMPLEXITY AND POPULATION
Of course, the models described earlier that show the importance of the effective population size in the process of cumulative cultural evolution are based
on the assumption of differential payoffs in influencing outcomes, even if, as
we have seen there is no feedback to population size. This seems appropriate
when the aim is to understand the adaptive role of culture in human evolution. However, it will not necessarily apply to all or even most cultural traits,
many of which will be under very weak selection or indeed completely neutral, that is to say having no fitness or performance differentials. In the latter
case, there may still be cumulative evolution in the sense that, for example,
a particular decoration pattern on a ceramic vessel may have been arrived at
via a series of prior steps, perhaps combining a number of existing variants,
but it will not be governed by the same rules. In such cases, we cannot make
any predictions about the history of particular variants but the turnover of
neutral variants will be determined by a combination of the innovation rate
and the population size (e.g., Bentley, Hahn, & Shennan, 2004), so the latter
remains important.
The models outlined earlier, although among the best known, are by no
means the only ones that have postulated a role for demographic factors
in the evolution of more complex human culture in prehistory. Still within
10
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
the same family as the innovation-transmission models described earlier,
Premo and Kuhn (2010) developed a spatially explicit agent-based model
that explored the extent to which frequent local extinctions of subpopulations in a meta-population could produce an appearance of stability
in the archaeological record. Their model showed that such an extinction
process would indeed affect a range of patterns potentially visible in the
archaeological record, including the rate of cumulative change and the
degree of differentiation between different local groups, because when
there are frequent local extinctions, more cumulative change is lost from the
meta-population. Importantly, this process affects neutral traits as well as
adaptive ones.
However, Moore (2013) has offered a completely different role for demography in the case of increased cultural complexity in prehistoric Australia.
He examined the change from simple stone flaking to more complex
sequences there and suggested that there was no evidence that they were
more efficient or led to the production of more efficient toolkits. By correlating their appearance with other developments, he postulated that their
adoption was a consequence of population increase that led to restrictions
on group mobility and the growth of closed territories. The increasing role
of material-based symbolic communication was part of the development of
relations between groups based on ceremony and exchange and involving
complex signaling, of which elaborate symbolically loaded stone tools
formed a part. As such, the processes generating cultural elaboration here
are completely different from those in the models described earlier as they
are not based on the relationship between demography and transmission
processes but on selection for investment in more complex signaling arising
from the consequences of increased population density. In this sense, the
model fits into the Malthus–Boserup framework discussed earlier because
the socio-technical innovations result in increased carrying capacity.
TRANSMISSION FIDELITY AND THE CONTRAST BETWEEN HUMAN AND NONHUMAN CULTURES
The previous discussion has explored a range of factors affecting cumulative
culture and its relationship to demographic processes but it has effectively
assumed that the detailed mechanisms affecting the fidelity of transmission
are a constant. Of course, this cannot be assumed when we are trying to
understand the transition from great ape culture to human culture and there
has been much discussion of the relative importance of creativity and transmission fidelity in accounting for the difference. A simulation of a simple
cultural transmission model by Lewis and Laland (2012) showed that the
most significant factor affecting the buildup of cumulative culture was cultural loss and that a major difference was made by minor improvements
Demography and Cultural Evolution
11
in the fidelity of transmission, which led to increased longevity of cultural
traits, while forms of innovation involving the combination of existing traits
also played a role. In keeping with the central importance of transmission
fidelity, they cited a study showing that superior achievement in a sequential problem-solving task was linked to verbal instruction and imitation in
children though not in chimpanzees.
Pradhan, Tennie, and van Schaik (2012), on the other hand, reject the idea
that the differences between great ape and human culture are explicable
by the fact that the former cannot imitate, because in their view, there is
evidence that they can transmit complex techniques. On the basis of their
knowledge of the social behavior of great ape groups, the authors propose
that the key dimension affecting the growth of complexity is differences
between the species in terms of sociability and they show by modeling that
greater sociability leads to increased technical accumulation. They conclude
that “[cultural accumulation] in hominins was induced by changes in social
organization that led to higher sociability, brought about by cooperative
hunting or scavenging, followed by the adoption of full terrestriality and
teaching elicited by systematic food sharing and provisioning, which further
improved social transmission of skills” (Pradhan et al. 2012, p. 186).
Although the authors do not appear to see it in this way, their model is
another version of the demographic arguments outlined earlier, in this case,
the trend to increased sociability is a trend toward increased effective population size as far as the acquisition and accumulation of skills is concerned;
the total connected population size in their model has no effect on cultural
accumulation but there is no reason why we should expect it to. In any event,
their proposal makes the important point that structures of social interaction
play an important role in affecting the fidelity of transmission. The same is
true of the nature of social practices, as Barth (1990) and Whitehouse (1992)
have shown in the case of the transmission of ritual in New Guinea societies.
KEY ISSUES FOR FUTURE RESEARCH
Cultural propensities are species-wide but cultures and cultural complexity
are not—they are specific to particular populations. Recent work has shown
that effective population size—the number of people interacting with respect
to a particular learned and transmitted activity—has a major impact on a
wide range of cultural evolutionary processes and their outcomes, including
cultural accumulation. Different models postulate different mechanisms as
responsible for this and an important task for the future must be to create
formal models including demography of processes for which they do not yet
exist and to compare the predictions of the various models. Introducing feedback from cultural states to population size via considerations of fitness is
12
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
also important, as is an investigation of the role of population size in relation
to attributes where selection is weak to nonexistent.
The emphasis in this review does not mean to imply that effective population size is the only relevant factor affecting cultural evolution, especially
the persistence of cultural traits. Costs and benefits, that is to say, selective
forces, are relevant and are likely to change through time with respect to any
given cultural feature. Transmission forces are also relevant, for example,
the importance of conformist transmission bias in affecting the fidelity of
transmission. Further psychological experiments and anthropological field
studies that address the role of different factors will also be required, so that
the models are based on growing knowledge of the psychology of innovation and transmission processes on the one hand and their operation in real
social contexts on the other.
REFERENCES
Barnes, I., Duda, A., Pybus, O. G., & Thomas, M. G. (2001). Ancient urbanisation
predicts genetic resistance to tuberculosis. Evolution, 65, 842–848. doi:10.1111/
j.1558-5646.2010.01132.x
Barth, F. (1990). The guru and the conjurer: Transactions in knowledge and the shaping of culture in Southeast Asia and Melanesia. Man (N.S.), 25, 640–653.
Bentley, R. A., Hahn, M. W., & Shennan, S. J. (2004). Random drift and culture change.
Proceedings of the Royal Society Biological Sciences, 271, 1443–1450. doi:10.1098/
rspb.2004.2746
Bettencourt, L. M. A., Lobo, J., Helbing, D., Künhert, C., & West, G. B. (2007). Growth,
innovation, scaling, and the pace of life in cities. Proceedings of the National Academy
of Sciences of the USA, 104, 7301–7306. doi:10.1073_pnas.0610172104
Boserup, E. (1965). The conditions of agricultural growth. Chicago, IL: Aldine Publishing
Company.
Boyd, R., & Richerson, P. J. (1985). Culture and the evolutionary process. Chicago, IL:
Chicago University Press.
Cavalli-Sforza, L., & Feldman, M. W. (1981). Cultural transmission and evolution: A
quantitative approach. Princeton, NJ: Princeton University Press.
Enquist, M., Ghirlanda, S., Jarrick, A., & Wachtmeister, C. A. (2008). Why does
human culture increase exponentially? Theoretical Population Biology, 74, 46–55.
doi:10.1016/j.tpb.2008.04.007
Enquist, M., Ghirlanda, S., & Kimmo Eriksson, K. (2011). Modelling the evolution
and diversity of cumulative culture. Philosophical Transactions of the Royal Society B,
366, 412–423. doi:10.1098/rstb.2010.0132
Fisher, R. A. (1930). The genetical theory of natural selection. Oxford, England: Clarendon Press.
Ghirlanda, S., Enquist, M., & Perc, M. (2010). Sustainability of culture-driven population dynamics. Theoretical Population Biology, 77, 181–188.
Demography and Cultural Evolution
13
Henrich, J. (2004). Demography and cultural evolution: How adaptive cultural processes can produce maladaptive losses—the Tasmanian case. American Antiquity,
69, 197–214.
Kaplan, H., & Hill, K. (1992). The evolutionary ecology of food acquisition. In
E. A. Smith & B. Winterhalder (Eds.), Evolutionary ecology and human behavior
(pp. 167–202). New York, NY: Aldine de Gruyter.
Kline, M. A., & Boyd, R. (2010). Population size predicts technological complexity in Oceania. Proceedings of the Royal Society B, 277, 2559–2564. doi:10.1098/
rspb.2010.0452
Knauft, B. M. (1986). Divergence between cultural success and reproductive fitness
in preindustrial cities. Cultural Anthropology, 2, 94–114.
Kobayashi, Y., & Aoki, K. (2012). Innovativeness, population size and cumulative cultural evolution. Theoretical Population Biology, 82, 38–47. doi:10.1016/
j.tpb.2012.04.001
Kremer, M. (1993). Population growth and technological change: One million B.C. to
1990. The Quarterly Journal of Economics, 108, 681–716.
Lehmann, L., & Wakano, J. Y. (2013). The handaxe and the microscope: Individual and social learning in a multidimensional model of adaptation. Evolution and
Human Behavior, 34, 109–117. doi:10.1016/j.evolhumbehav.2012.11.001
Lewis, H. M., & Laland, K. N. (2012). Transmission fidelity is the key to the
build-up of cumulative culture. Philosophical Transactions of the Royal Society B, 367,
2171–2180. doi:10.1098/rstb.2012.0119
Lombard, M., & Haidle, M. N. (2012). Thinking a bow-and-arrow set: Cognitive
implications of middle stone age bow and stone-tipped arrow technology. Cambridge Archaeological Journal, 22, 237–264. doi:10.1017/S095977431200025X
Mackay, A., & Marwick, B. (2011). Costs and benefits in technological decision making under variable conditions: Examples from the late Pleistocene in southern
Africa. In B. Marwick & A. Mackay (Eds.), Keeping your edge: Recent approaches to
the organisation of stone artefact technology (pp. 119–134). Oxford, England: Oxbow
Books.
Moore, M. W. (2013). Simple stone flaking in Australasia: Patterns and implications.
Quaternary International, 285, 140–149. doi:10.1016/j.quaint.2011.09.030
Peck, J. R., Barreau, G., & Heath, S. C. (1997). Imperfect genes, Fisherian mutation
and the evolution of sex. Genetics, 145, 1171–1199.
Powell, A., Shennan, S. J., & Thomas, M. G. (2009). Late Pleistocene demography and the appearance of modern human behavior. Science, 324, 1298–1301.
doi:10.1126/science.1170165
Pradhan, G. R., Tennie, C., & van Schaik, C. P. (2012). Social organization and the evolution of cumulative technology in apes and hominins. Journal of Human Evolution,
63, 180–190. doi:10.1016/j.jhevol.2012.04.008
Premo, L. S., & Kuhn, S. L. (2010). Modeling effects of local extinctions on culture change and diversity in the paleolithic. PLoS One, 5(12), e15582. doi:10.1371/
journal.pone.0015582
Richerson, P. J., Boyd, R., & Bettinger, R. L. (2009). Cultural innovations and demographic change. Human Biology, 81, 211–235. doi:10.1353/hub.0.0054
14
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
Roux, V. (2010). Technological innovations and developmental trajectories: Social factors as evolutionary forces. In M. J. O’Brien & S. J. Shennan (Eds.), Innovations in
cultural systems: Contributions from evolutionary anthropology (pp. 217–234). Cambridge, MA: MIT Press.
Shennan, S. J. (2001). Demography and cultural innovation: A model and some implications for the emergence of modern human culture. Cambridge Archaeological Journal, 11, 5–16.
Stiner, M. C., Munro, N. D., & Surovell, T. A. (2000). The tortoise and the hare small-game use, the broad-spectrum revolution, and paleolithic demography. Current Anthropology, 41, 39–73.
Whitehouse, H. (1992). Memorable religions: Transmission, codification, and change
in divergent Melanesian contexts. Man (N.S.), 27, 777–797.
Whiten, A., Goodall, J., McGrew, W. C., Nishida, T., Reynolds, V., Sugiyama, Y., … ,
Boesch, C. (1999). Cultures in chimpanzees. Nature, 399, 682–685. doi:10.1038/
21415
STEPHEN SHENNAN SHORT BIOGRAPHY
Stephen Shennan studied Archaeology and Anthropology at the University
of Cambridge, UK, where he also completed his PhD. He spent the earlier
part of his professional academic career at the University of Southampton,
UK, before moving to the Institute of Archaeology, University College London (UCL). He is currently Professor of Theoretical Archaeology and Director
of the Institute of Archaeology, UCL. His main research interest is the study
of cultural evolution on the basis of archaeological evidence, using cultural
evolutionary theory to develop and test hypotheses about cultural change
in prehistory, and using archaeology to throw light on postulated cultural
evolutionary processes. From 2001 to 2005, he was the Director of the Centre
for the Evolutionary Analysis of Cultural Behaviour at UCL and he currently
holds an Advanced Grant from the European Research Council for the project
“Cultural Evolution of Neolithic Europe.”
RELATED ESSAYS
Cultural Neuroscience: Connecting Culture, Brain, and Genes (Psychology),
Shinobu Kitayama and Sarah Huff
Below-Replacement Fertility (Sociology), S. Philip Morgan
Limits to Human Longevity (Sociology), Samuel H. Preston and Hiram
Beltrán-Sánchez
Production of Culture (Sociology), Vaughn Schmutz and Candace N. Miller
Recent Demographic Trends and the Family (Sociology), Lawrence L. Wu
-
Demography and Cultural Evolution
STEPHEN SHENNAN
Abstract
Trying to explain the increase in cultural complexity over the long term of human
history has long been an interest of anthropology and of historical social sciences
more generally. In recent years, interest has grown rapidly in the idea that a key
factor in accounting for it might be the size of the human population itself and the
extent of interaction between people, because of the effect these have on the innovation rates in populations and on the success with which innovations are transmitted.
An important driver of this growth of interest has been the emergence of the new
interdisciplinary field of cultural evolution, which makes extensive use of mathematical techniques, especially methods derived from population genetics. The result has
been the development of a range of analytical and computer simulation models that
make various predictions about the way in which population size influences cultural
change, and in particular the growth of cumulative culture, including the processes
that have led from the very simple forms of culture possessed by other great apes
to those characteristic of Homo sapiens. The aim of this review is to distinguish them,
so that future work can focus on evaluating their strengths and weaknesses and the
circumstances in which they are useful.
INTRODUCTION
The past decade or so has seen an enormous increase in interest in the relationship between population patterns and processes on the one hand and
patterns of cultural change and adaptation, on the other; in particular, the
extent to which population size and interaction rates have an effect on innovation rates and on the growth of cultural complexity or cumulative culture.
Cumulative culture refers to the idea that over time the number of cultural
elements present in human societies tends to increase, and that the presence
of complex cultural or technological innovations, for example, the internal
combustion engine, requires the prior existence of a number of other traits.
A major driver of this new interest has been the development of the interdisciplinary field of cultural evolution (Boyd & Richerson, 1985; Cavalli-Sforza
& Feldman, 1981), which takes a Darwinian theoretical perspective on cultural variation and change, seeing it in terms of innovation, transmission, and
Emerging Trends in the Social and Behavioral Sciences. Edited by Robert Scott and Stephen Kosslyn.
© 2015 John Wiley & Sons, Inc. ISBN 978-1-118-90077-2.
1
2
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
selection processes. Associated with this perspective has been the application
to studies of cultural data of methods derived from evolutionary biology,
especially population genetics. More specifically, the development of this
new field has led to the growth of interest in exploring the existence or otherwise of cultural traditions in animals, not least our closest living relatives, the
great apes (e.g., Whiten et al., 1999). Much if not most of the work on the relationship between population and cultural complexity has been of a generalizing nature, based on the construction of relatively abstract models, though
there has been a particular interest in the long-term evolutionary question
of how the cultural complexity of present-day human societies could have
emerged from the much simpler culture of other nonhuman species.
There is now a range of different models and suggestions in the literature relating to the role of population processes in cultural accumulation
and change, and the aim of this review is to distinguish them, so that future
work can focus on evaluating their strengths and weaknesses and the circumstances in which they are useful. An initial distinction can be made between
studies that have placed particular emphasis on the detailed processes of
cultural transmission and innovation and those that pay less attention to
mechanisms and look more at general trends. I will begin with the latter,
which have a longer history, but the vast majority of research on this topic
has only been published in the twenty-first century and is still the subject of
intense debate, hence the distinction between foundational and cutting-edge
research is a difficult one to make.
FOUNDATIONAL RESEARCH
MACROSCALE ECONOMIC MODELS
At the end of the eighteenth century, Thomas Malthus proposed that the
growth of human populations rapidly outstrips available resources and that
they are held in check by disease, famine, and warfare. As many have pointed
out since, while this was true in the past, Malthus was writing at a time when
the situation was changing. The Industrial Revolution and subsequent technological developments have meant that resources have been able to keep
up with population, which has risen to unprecedented levels. Marx criticized Malthus on these grounds, and much more recently, Boserup (1965)
proposed that the deterioration in living conditions produced by “population pressure” created incentives in human societies to innovate to overcome
existing limits. Kremer’s (1993) model of long-term technological evolution
and its relation to population size is in the spirit of Malthus and Boserup.
Population is limited by the carrying capacity of the environment given the
availability of a specific technology. Technological innovation can increase
Demography and Cultural Evolution
3
the carrying capacity, and population will then rise to a new limit. However,
the rate of innovation is not independent of population size. Other things
being equal, a larger population will contain more innovators, so there will
be a feedback relationship between technology and population size, with one
dependent on the other. This further implies that the rate of growth of a population at a given time is proportional to its size at that time, in contrast to
models where technological improvement is independent of population size,
determined by other factors. Kremer shows that, at least until recently, the
predictions of this model are consistent with long-term human population
history, and that it can be extended to a more general version that factors
in variations in research productivity, which can additionally account for
the recent decline in population growth rates with increased incomes. He
also considers the geographical implications of his model, pointing out that
for the same level of technology and population density, regions with larger
areas and therefore larger starting population sizes will have faster rates of
technological change and over a given period will reach higher levels of population and technology. Again, he shows that this prediction is met by data
on estimated population sizes of different regions of the world at 1500 AD
just before European expansion into the New World. Finally, he notes that
his model does not include the possibility of technological loss, as appears to
have occurred in prehistoric Tasmania after it was cut off from the Australian
mainland, potentially leading to a negative spiral in technology and population. Kremer ascribes the relationship between technological complexity and
population size to the “nonrivalry” of technology, the fact that one person’s
use of it does not preclude the use by others.
Richerson, Boyd, and Bettinger (2009) also develop an economic model
within the Malthus–Boserup framework, proposing, like Boserup, that as
population approaches a given carrying capacity and individual income
goes down, people will innovate and intensify, but also suggesting, in line
with the diet-breadth model in behavioral ecology (e.g., Kaplan & Hill,
1992), that if incomes increase, for example, as a result of the appearance of
new resources, perhaps as a result of climate change, people can also “deinnovate” or deintensify, at least for short periods until population catches up.
Apart from such short periods, “population pressure” is always present; so
long-term population trends depend on the rate of innovation that increases
the environmental carrying capacity. Interestingly, the earliest evidence we
have for pressure on resources is only 50,000 years ago, with indications
of the overexploitation of tortoises in the Middle East (Stiner, Munro, &
Surovell, 2000), and Richerson et al. (2009, pp. 223–224) suggest that before
this, the limits to human population size might have come from competition
with other carnivores, postulating social and technical innovations that
could have made modern humans more successful competitors. They go on
4
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
to suggest, following Marx, that, at least in the Holocene, the limiting factor
on cultural evolution may have been the rate of social rather than technical
innovation, and conjecture that at the broadest social and temporal scale, a
model in which the innovation rate is dependent on the population size will
lead to exponential growth in technological sophistication (2009, p. 228).
CUTTING-EDGE RESEARCH
MACROSCALE INNOVATION-TRANSMISSION MODELS
Ghirlanda, Enquist, and Perc (2010) take a rather different approach to the
relation between population and culture by focusing more closely on the
processes involved. To understand the relationship between population and
culture, we need to consider the balance between individual creativity and
the effect of cultural innovations on environmental carrying capacity, on the
one hand, and the rate of cultural loss, on the other. If the former outstrips
the latter, then cultural accumulation or complexity, together with the population size, can increase. This can arise as a result of an increased effect of
culture on carrying capacity, increased innovation or increased transmission
fidelity. Ghirlanda et al. also introduce the possibility that an innovation that
initially increases carrying capacity may not continue to do so, for example,
the use of firearms for hunting may initially increase returns but then lead to a
decrease as resources are overexploited. They show that varying the effect of
culture on carrying capacity and the impact of what they call the corruption
rate, the rate at which adaptive features become nonadaptive, can result in
very different dynamics, ranging from stability to unlimited growth to population extinction. Unsurprisingly, the introduction of a parameter that allows
for “adaptive filtering,” the ability to detect and discard features that have
become maladaptive, increases the space in which unlimited growth occurs.
Finally, they return to the question of the hyperbolic rate of human population increase over the long term, generally assumed, as in Kremer’s model,
to arise from the fact that innovations increase both the environmental carrying capacity and the rate of population increase, because larger populations
have more innovations, and suggest a possible alternative explanation that
the rate of increase in innovations with population size may be faster than linear. The relevance of this suggestion is borne out by the study of Bettencourt,
Lobo, Helbing, Künhert, and West (2007) showing that the rate of increase
in innovations in cities is not a linear function of the population size but of
population size to the power 1.2, which leads to growth that is faster than
exponential. On the other hand, it is clear that until very recent times, cities
were population sinks, with low birth rates and high death rates, maintained
only by large-scale immigration from the countryside [Knauft, 1986, cited by
Demography and Cultural Evolution
5
Richerson et al. (2009), cf. Barnes, Duda, Pybus, & Thomas, 2001]. How the
accumulation of innovations and increased population would have emerged
from these different forces remains to be explored.
The long-term mutual dependency of technology and population size is
clear from all the studies outlined so far; however, the view of technology, or
adaptive culture, represented, important though it is, is a one-dimensional
one and does not consider what we might call the internal dynamics of
cumulative culture itself. As pointed out earlier, cumulative culture refers to
the idea that, beyond the very simplest level, cultural attributes generally
depend on one another, in the sense that the presence of one cultural attribute
affects the likelihood that another will be present; the strongest cases being
those where one is a prerequisite for the other, and, at the other extreme,
where one precludes the other. Enquist, Ghirlanda, and Kimmo Eriksson
(2011, pp. 415–418 and Figure 1) show that there is variety of ways that such
dependencies can arise: directional change in a single dimension, where to
get from state (a) to state (c) you have to go through state (b); branching
differentiation, where a given state can give rise to two or more states, which
then change independently of one another; pairwise combinations, where
the state of one attribute is affected by the state of two others; and finally,
“systems of cultural elements” in which there may be multiple influences,
positive and negative, between attribute states, including reciprocal ones.
Unsurprisingly, simulation studies show that “systems of cultural elements”
result in cultures whose number of elements increases at a far faster rate than
any of the others and also produce corresponding increases in the histories of
individual elements, that is to say, “the number of evolutionary events that
created the element, starting from a cultural seed” (Enquist et al., 2011, p. 418,
Figure 5). Systems characterized by these complex interdependencies are
strongly path-dependent, both because of the very large number of possible
links and because there is a strong stochastic element as the links are not
deterministic ones. The authors argue that there is strong evidence in many
domains for the predicted exponential increase in the diversity and complexity of culture (Enquist, Ghirlanda, Jarrick, & Wachtmeister, 2008, Figure 1). In
summary, the rate of accumulation of culture at a given point depends on its
rate at the previous point purely as a result of the increase in cultural interdependencies. However, as Enquist et al. emphasize, this is a very abstract
model and it does not address transmission processes at the individual
level, nor does it consider the demographic aspects that they and others
have shown elsewhere to be important, as we have seen earlier. The role
of selection in winnowing out ineffective combinations of traits is also not
considered.
6
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
MODELS OF SPECIFIC INNOVATION-TRANSMISSION PROCESSES
The best-known model that looks at the implications of specific innovation
and transmission processes and their link to demography is that of Henrich
(2004). This is based on a model of learning in which individuals attempt to
imitate the most skilled individual available to them. Most individuals will
fail to reach that level, so there will be a distance between the modal skill
level in the population and the best. However, there will be a certain probability that an individual will exceed the current best and a new, higher, most
skilled level will result. In these circumstances, given a constant gap between
the best level and the modal level, the result will be an improvement in the
modal level. The probability of improvement or decline in the population
modal level depends on the ratio of the gap between the mode and the best,
on the one hand, and the dispersion of the distribution of skill levels on the
other. If the gap is small and the dispersion is large, then improvement will
be far more probable than in the opposite case. For a given ratio, though,
Henrich demonstrates that what happens will depend on population size, as
improbable events will occur more often in larger populations. For very complex technologies where the gap between the modal level and the best is large
and the variation between individuals is small, then a large population will
be required to maintain or improve the technology. Thus, if external forces
cause a decline in the population size, the most complex technologies are
likely to be lost. Simpler ones, even if they are lost, have a much higher probability of being reinvented. Henrich showed that this process could account for
the well-known and apparently deeply puzzling cultural impoverishment
of the aboriginal Tasmanians after Tasmania was cut off from the Australian
mainland at the end of the past Ice Age. More generally, “It is the selective
transmission of lucky errors and occasional experiments that drives much of
the evolution of adaptive technology, skills, beliefs, and practices” (Henrich,
2004, p. 202), and makes it strongly dependent on population size. Henrich’s
result has subsequently been confirmed by Kobayashi and Aoki (2012), who
showed that a version of the model with the more satisfactory assumption
of overlapping rather than discrete generations produced an even stronger
population effect.
The substantive assumption in Henrich’s argument that real improvements
are difficult to make because of the sheer amount of cultural knowledge
embedded and embodied in human artifacts and cultural practices is
supported by the ethnoarchaeological work of Roux (2010), who makes the
point that training in any production process that involves the acquisition
of high levels of expertise that take a long time to acquire is likely to
restrict innovation, because the whole process of learning is designed to
fix particular sets of skills and knowledge; this may be particularly the
Demography and Cultural Evolution
7
case with physical expertise and motor skills. Only the most expert, those
who have complete control of all aspects of a process and its associated
knowledge, are likely to transcend the limits of what they have learned and
invent something new. Roux cites (2010, p. 224) the example of inventions
in the field of pyrotechnology by traditional craftsmen in India, which were
only made by those who were most skillful, individuals “exceptional as
much for their skills as for their rarity.”
It is important to note that the population size in such cases refers to the
“effective population size,” not the total local population but the relevant
interacting population of learners and teachers. Thus, in the case of Henrich’s
proposal for Tasmania, the interacting population size decreases because of
the loss of contact with the mainland, not because of local population decline,
though this probably occurred. Where there is craft specialization, the effective population size may be tiny and the skills are correspondingly vulnerable to loss. Thus, Roux (2010) shows that despite the productivity of the
technique, the wheel-coiling of pottery was lost on two occasions in the societies of the ancient Levant because it was a highly specialized activity with
only a small number of practitioners. When the socioeconomic conditions
that sustained it collapsed, a small number of transmission links were broken. It was only later, as the number and spatial extent of transmission links
increased, that this technical system became less vulnerable to the effects of
external historical events.
Powell, Shennan, and Thomas (2009) simulated a version of Henrich’s model and effectively ran it in reverse, to explore the effects of
increased population density and increased migration between groups in
a meta-population on increases in cultural complexity. They showed that
differentials between regions in terms of population density or migration
rate could sustain very substantial differences in the skill level as measured
in terms of Henrich’s model, and proposed that these factors could account
for the puzzling appearance and disappearance of features of the so-called
behavioral modernity in the African Middle Stone Age, features that had
previously been accounted for in terms of biological cognitive evolution, if
human populations fluctuated in size or contact rates as a result of changing
climate patterns. However, others have emphasized the importance of
evaluating the costs and benefits of more and less complex technologies
in different environments (Mackay & Marwick, 2011). In any case, it is
worth emphasizing that neither Henrich’s original model and, Powell et al.’s
version, nor Kobayashi and Aoki’s modification includes any feedback
from the improving or devolving technology itself to the population size.
Population size is simply an independent variable affecting the innovation
and transmission of novel improved ways of doing things.
8
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
A different model of the relationship between cultural complexity and population size was explored by Shennan (2001). He modified a genetic model
developed by Peck, Barreau, and Heath (1997) on the basis of the ideas of
Fisher (1930). Individuals have a number of different attributes or traits that
are subject to selection depending on the values of their attribute states. The
model starts with a perfectly fit population. Cultural transmission occurs,
with the parental value of a cultural trait having a certain probability of not
being passed on to its offspring; when this occurs, the individual chosen to
be the cultural model is the one with the highest adaptive value chosen from
a random selection of individuals in the parental generation; otherwise, the
offspring takes the parent’s value. In the course of transmission innovations
can occur, which can be either beneficial or deleterious. The probability that
any given attribute state adopted by an offspring will have just undergone an
innovation is represented by the attribute innovation rate. An innovation is
assumed to change the adaptive value associated with an attribute by some
amount, drawn from a distribution such that the majority of innovations
have very little effect but some have a more significant one. The results of
the simulations carried out using this model showed that larger populations
can evolve to a higher average fitness than smaller ones, because they carry
a smaller drift load of deleterious cultural traits, though again there was no
feedback to population size.
A similar model has recently been used by Lehmann and Wakano (2013) to
look at the effect of the multidimensionality of cultural traits; in other words,
where the best payoff depends on getting the right combination of a number
of different features, for example, different aspects of size and shape of an
artifact. In Henrich’s model, everything is collapsed down to a single dimension. With increasing numbers of dimensions, there are far more ways of
making things worse than making them better. Where the number of dimensions is very high, then even a learner with very high computational ability
going through different combinations and their effects in their imagination
will not find the optimum, but a population of payoff-biased learners making
uncorrelated errors will be able to reach it if it is sufficiently large. An excellent example of the kinds of complexity and multidimensionality involved is
provided by Lombard and Haidle’s (2012) construction of what they call the
“effective chain” of activities for the production of bows and arrows, based
on evidence from Middle Stone Age Africa.
If we think about the different kinds of dependencies between cultural elements described by Enquist et al. (2011) and discussed earlier, then clearly the
case where there are “systems of elements” in which there may be multiple
influences, positive and negative, between attribute states, and whose numbers of elements increase very rapidly as a result, will also be the case where
finding an optimum is the most difficult. Once again then, from a different set
Demography and Cultural Evolution
9
of premises, successful innovation, and therefore adaptive cumulative culture, depends on population size. The more complex the system of elements
becomes, the steeper is the rate of increase in cultural complexity, as Enquist
et al. (2011) show; correspondingly, the larger the population size needs to be
in order to have a high probability of arriving at the optimum combination
of trait values.
The general argument about the importance of population size for cultural
complexity is supported by Kline and Boyd’s (2010) study of marine fishing
technology in Oceania. This showed a correlation between the number and
complexity of marine fishing tools and the size and connectedness of island
populations, with smaller and more isolated islands having fewer and less
complex tools than larger and less isolated ones. Importantly, they point out
that the pattern is not entirely consistent with the type of Malthus–Boserup
model proposed by Kremer in which improved technology produces
increased carrying capacity and therefore increased population because this
model does not explain why the connectedness of the islands, not just their
local population size, also has an effect. They proposed that the pattern
could be accounted for by either a drift model of the kind proposed by
Shennan or Henrich’s process (which they call the treadmill model); however,
in any case, a model in which the transmission process has a significant role.
ALTERNATIVE MODELS OF LINKS BETWEEN CULTURAL COMPLEXITY AND POPULATION
Of course, the models described earlier that show the importance of the effective population size in the process of cumulative cultural evolution are based
on the assumption of differential payoffs in influencing outcomes, even if, as
we have seen there is no feedback to population size. This seems appropriate
when the aim is to understand the adaptive role of culture in human evolution. However, it will not necessarily apply to all or even most cultural traits,
many of which will be under very weak selection or indeed completely neutral, that is to say having no fitness or performance differentials. In the latter
case, there may still be cumulative evolution in the sense that, for example,
a particular decoration pattern on a ceramic vessel may have been arrived at
via a series of prior steps, perhaps combining a number of existing variants,
but it will not be governed by the same rules. In such cases, we cannot make
any predictions about the history of particular variants but the turnover of
neutral variants will be determined by a combination of the innovation rate
and the population size (e.g., Bentley, Hahn, & Shennan, 2004), so the latter
remains important.
The models outlined earlier, although among the best known, are by no
means the only ones that have postulated a role for demographic factors
in the evolution of more complex human culture in prehistory. Still within
10
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
the same family as the innovation-transmission models described earlier,
Premo and Kuhn (2010) developed a spatially explicit agent-based model
that explored the extent to which frequent local extinctions of subpopulations in a meta-population could produce an appearance of stability
in the archaeological record. Their model showed that such an extinction
process would indeed affect a range of patterns potentially visible in the
archaeological record, including the rate of cumulative change and the
degree of differentiation between different local groups, because when
there are frequent local extinctions, more cumulative change is lost from the
meta-population. Importantly, this process affects neutral traits as well as
adaptive ones.
However, Moore (2013) has offered a completely different role for demography in the case of increased cultural complexity in prehistoric Australia.
He examined the change from simple stone flaking to more complex
sequences there and suggested that there was no evidence that they were
more efficient or led to the production of more efficient toolkits. By correlating their appearance with other developments, he postulated that their
adoption was a consequence of population increase that led to restrictions
on group mobility and the growth of closed territories. The increasing role
of material-based symbolic communication was part of the development of
relations between groups based on ceremony and exchange and involving
complex signaling, of which elaborate symbolically loaded stone tools
formed a part. As such, the processes generating cultural elaboration here
are completely different from those in the models described earlier as they
are not based on the relationship between demography and transmission
processes but on selection for investment in more complex signaling arising
from the consequences of increased population density. In this sense, the
model fits into the Malthus–Boserup framework discussed earlier because
the socio-technical innovations result in increased carrying capacity.
TRANSMISSION FIDELITY AND THE CONTRAST BETWEEN HUMAN AND NONHUMAN CULTURES
The previous discussion has explored a range of factors affecting cumulative
culture and its relationship to demographic processes but it has effectively
assumed that the detailed mechanisms affecting the fidelity of transmission
are a constant. Of course, this cannot be assumed when we are trying to
understand the transition from great ape culture to human culture and there
has been much discussion of the relative importance of creativity and transmission fidelity in accounting for the difference. A simulation of a simple
cultural transmission model by Lewis and Laland (2012) showed that the
most significant factor affecting the buildup of cumulative culture was cultural loss and that a major difference was made by minor improvements
Demography and Cultural Evolution
11
in the fidelity of transmission, which led to increased longevity of cultural
traits, while forms of innovation involving the combination of existing traits
also played a role. In keeping with the central importance of transmission
fidelity, they cited a study showing that superior achievement in a sequential problem-solving task was linked to verbal instruction and imitation in
children though not in chimpanzees.
Pradhan, Tennie, and van Schaik (2012), on the other hand, reject the idea
that the differences between great ape and human culture are explicable
by the fact that the former cannot imitate, because in their view, there is
evidence that they can transmit complex techniques. On the basis of their
knowledge of the social behavior of great ape groups, the authors propose
that the key dimension affecting the growth of complexity is differences
between the species in terms of sociability and they show by modeling that
greater sociability leads to increased technical accumulation. They conclude
that “[cultural accumulation] in hominins was induced by changes in social
organization that led to higher sociability, brought about by cooperative
hunting or scavenging, followed by the adoption of full terrestriality and
teaching elicited by systematic food sharing and provisioning, which further
improved social transmission of skills” (Pradhan et al. 2012, p. 186).
Although the authors do not appear to see it in this way, their model is
another version of the demographic arguments outlined earlier, in this case,
the trend to increased sociability is a trend toward increased effective population size as far as the acquisition and accumulation of skills is concerned;
the total connected population size in their model has no effect on cultural
accumulation but there is no reason why we should expect it to. In any event,
their proposal makes the important point that structures of social interaction
play an important role in affecting the fidelity of transmission. The same is
true of the nature of social practices, as Barth (1990) and Whitehouse (1992)
have shown in the case of the transmission of ritual in New Guinea societies.
KEY ISSUES FOR FUTURE RESEARCH
Cultural propensities are species-wide but cultures and cultural complexity
are not—they are specific to particular populations. Recent work has shown
that effective population size—the number of people interacting with respect
to a particular learned and transmitted activity—has a major impact on a
wide range of cultural evolutionary processes and their outcomes, including
cultural accumulation. Different models postulate different mechanisms as
responsible for this and an important task for the future must be to create
formal models including demography of processes for which they do not yet
exist and to compare the predictions of the various models. Introducing feedback from cultural states to population size via considerations of fitness is
12
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
also important, as is an investigation of the role of population size in relation
to attributes where selection is weak to nonexistent.
The emphasis in this review does not mean to imply that effective population size is the only relevant factor affecting cultural evolution, especially
the persistence of cultural traits. Costs and benefits, that is to say, selective
forces, are relevant and are likely to change through time with respect to any
given cultural feature. Transmission forces are also relevant, for example,
the importance of conformist transmission bias in affecting the fidelity of
transmission. Further psychological experiments and anthropological field
studies that address the role of different factors will also be required, so that
the models are based on growing knowledge of the psychology of innovation and transmission processes on the one hand and their operation in real
social contexts on the other.
REFERENCES
Barnes, I., Duda, A., Pybus, O. G., & Thomas, M. G. (2001). Ancient urbanisation
predicts genetic resistance to tuberculosis. Evolution, 65, 842–848. doi:10.1111/
j.1558-5646.2010.01132.x
Barth, F. (1990). The guru and the conjurer: Transactions in knowledge and the shaping of culture in Southeast Asia and Melanesia. Man (N.S.), 25, 640–653.
Bentley, R. A., Hahn, M. W., & Shennan, S. J. (2004). Random drift and culture change.
Proceedings of the Royal Society Biological Sciences, 271, 1443–1450. doi:10.1098/
rspb.2004.2746
Bettencourt, L. M. A., Lobo, J., Helbing, D., Künhert, C., & West, G. B. (2007). Growth,
innovation, scaling, and the pace of life in cities. Proceedings of the National Academy
of Sciences of the USA, 104, 7301–7306. doi:10.1073_pnas.0610172104
Boserup, E. (1965). The conditions of agricultural growth. Chicago, IL: Aldine Publishing
Company.
Boyd, R., & Richerson, P. J. (1985). Culture and the evolutionary process. Chicago, IL:
Chicago University Press.
Cavalli-Sforza, L., & Feldman, M. W. (1981). Cultural transmission and evolution: A
quantitative approach. Princeton, NJ: Princeton University Press.
Enquist, M., Ghirlanda, S., Jarrick, A., & Wachtmeister, C. A. (2008). Why does
human culture increase exponentially? Theoretical Population Biology, 74, 46–55.
doi:10.1016/j.tpb.2008.04.007
Enquist, M., Ghirlanda, S., & Kimmo Eriksson, K. (2011). Modelling the evolution
and diversity of cumulative culture. Philosophical Transactions of the Royal Society B,
366, 412–423. doi:10.1098/rstb.2010.0132
Fisher, R. A. (1930). The genetical theory of natural selection. Oxford, England: Clarendon Press.
Ghirlanda, S., Enquist, M., & Perc, M. (2010). Sustainability of culture-driven population dynamics. Theoretical Population Biology, 77, 181–188.
Demography and Cultural Evolution
13
Henrich, J. (2004). Demography and cultural evolution: How adaptive cultural processes can produce maladaptive losses—the Tasmanian case. American Antiquity,
69, 197–214.
Kaplan, H., & Hill, K. (1992). The evolutionary ecology of food acquisition. In
E. A. Smith & B. Winterhalder (Eds.), Evolutionary ecology and human behavior
(pp. 167–202). New York, NY: Aldine de Gruyter.
Kline, M. A., & Boyd, R. (2010). Population size predicts technological complexity in Oceania. Proceedings of the Royal Society B, 277, 2559–2564. doi:10.1098/
rspb.2010.0452
Knauft, B. M. (1986). Divergence between cultural success and reproductive fitness
in preindustrial cities. Cultural Anthropology, 2, 94–114.
Kobayashi, Y., & Aoki, K. (2012). Innovativeness, population size and cumulative cultural evolution. Theoretical Population Biology, 82, 38–47. doi:10.1016/
j.tpb.2012.04.001
Kremer, M. (1993). Population growth and technological change: One million B.C. to
1990. The Quarterly Journal of Economics, 108, 681–716.
Lehmann, L., & Wakano, J. Y. (2013). The handaxe and the microscope: Individual and social learning in a multidimensional model of adaptation. Evolution and
Human Behavior, 34, 109–117. doi:10.1016/j.evolhumbehav.2012.11.001
Lewis, H. M., & Laland, K. N. (2012). Transmission fidelity is the key to the
build-up of cumulative culture. Philosophical Transactions of the Royal Society B, 367,
2171–2180. doi:10.1098/rstb.2012.0119
Lombard, M., & Haidle, M. N. (2012). Thinking a bow-and-arrow set: Cognitive
implications of middle stone age bow and stone-tipped arrow technology. Cambridge Archaeological Journal, 22, 237–264. doi:10.1017/S095977431200025X
Mackay, A., & Marwick, B. (2011). Costs and benefits in technological decision making under variable conditions: Examples from the late Pleistocene in southern
Africa. In B. Marwick & A. Mackay (Eds.), Keeping your edge: Recent approaches to
the organisation of stone artefact technology (pp. 119–134). Oxford, England: Oxbow
Books.
Moore, M. W. (2013). Simple stone flaking in Australasia: Patterns and implications.
Quaternary International, 285, 140–149. doi:10.1016/j.quaint.2011.09.030
Peck, J. R., Barreau, G., & Heath, S. C. (1997). Imperfect genes, Fisherian mutation
and the evolution of sex. Genetics, 145, 1171–1199.
Powell, A., Shennan, S. J., & Thomas, M. G. (2009). Late Pleistocene demography and the appearance of modern human behavior. Science, 324, 1298–1301.
doi:10.1126/science.1170165
Pradhan, G. R., Tennie, C., & van Schaik, C. P. (2012). Social organization and the evolution of cumulative technology in apes and hominins. Journal of Human Evolution,
63, 180–190. doi:10.1016/j.jhevol.2012.04.008
Premo, L. S., & Kuhn, S. L. (2010). Modeling effects of local extinctions on culture change and diversity in the paleolithic. PLoS One, 5(12), e15582. doi:10.1371/
journal.pone.0015582
Richerson, P. J., Boyd, R., & Bettinger, R. L. (2009). Cultural innovations and demographic change. Human Biology, 81, 211–235. doi:10.1353/hub.0.0054
14
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
Roux, V. (2010). Technological innovations and developmental trajectories: Social factors as evolutionary forces. In M. J. O’Brien & S. J. Shennan (Eds.), Innovations in
cultural systems: Contributions from evolutionary anthropology (pp. 217–234). Cambridge, MA: MIT Press.
Shennan, S. J. (2001). Demography and cultural innovation: A model and some implications for the emergence of modern human culture. Cambridge Archaeological Journal, 11, 5–16.
Stiner, M. C., Munro, N. D., & Surovell, T. A. (2000). The tortoise and the hare small-game use, the broad-spectrum revolution, and paleolithic demography. Current Anthropology, 41, 39–73.
Whitehouse, H. (1992). Memorable religions: Transmission, codification, and change
in divergent Melanesian contexts. Man (N.S.), 27, 777–797.
Whiten, A., Goodall, J., McGrew, W. C., Nishida, T., Reynolds, V., Sugiyama, Y., … ,
Boesch, C. (1999). Cultures in chimpanzees. Nature, 399, 682–685. doi:10.1038/
21415
STEPHEN SHENNAN SHORT BIOGRAPHY
Stephen Shennan studied Archaeology and Anthropology at the University
of Cambridge, UK, where he also completed his PhD. He spent the earlier
part of his professional academic career at the University of Southampton,
UK, before moving to the Institute of Archaeology, University College London (UCL). He is currently Professor of Theoretical Archaeology and Director
of the Institute of Archaeology, UCL. His main research interest is the study
of cultural evolution on the basis of archaeological evidence, using cultural
evolutionary theory to develop and test hypotheses about cultural change
in prehistory, and using archaeology to throw light on postulated cultural
evolutionary processes. From 2001 to 2005, he was the Director of the Centre
for the Evolutionary Analysis of Cultural Behaviour at UCL and he currently
holds an Advanced Grant from the European Research Council for the project
“Cultural Evolution of Neolithic Europe.”
RELATED ESSAYS
Cultural Neuroscience: Connecting Culture, Brain, and Genes (Psychology),
Shinobu Kitayama and Sarah Huff
Below-Replacement Fertility (Sociology), S. Philip Morgan
Limits to Human Longevity (Sociology), Samuel H. Preston and Hiram
Beltrán-Sánchez
Production of Culture (Sociology), Vaughn Schmutz and Candace N. Miller
Recent Demographic Trends and the Family (Sociology), Lawrence L. Wu