8 Late Pleistocene climate change and Paleolithic cultural evolution in Northern China: Implications from the Last Glacial Maximum Loukas Barton1, P. Jeffrey Brantingham2 and Duxue Ji3 1 Department of Anthropology, University of California, Davis 2 Department of Anthropology, University of California, Los Angeles 3 Center for Arid Environment and Paleoclimate Research, Lanzhou University, v.4.1 September 1, 2006 Abstract importantly, our perspective is limited by an historical adherence to the recent Holocene, a period of relatively Temporal and spatial patterns in archeological data from subtle environmental change marked by rapid, cumulative Pleistocene north China suggest strong correlations between cultural evolution maintained and accelerated by large num- climate change and culture change, but only in extreme bers of continuously interacting individuals. That is not to cases. In these cases, climate has an immediate impact on say that climate has no bearing on Holocene human history, human mobility, which is severely constrained during the merely that recognizing the impact of climate change on pronounced cold/dry intervals of the Pleistocene. As high cultural evolution is exceedingly difficult. For this reason, mobility becomes incompatible with the environmental lim- we turn our attention to the Pleistocene. itations of extreme intervals, such as the Last Glacial Max- Paleolithic hunter-gatherers had a range of potential imum, previously disparate mobile human groups aggregate strategies with which to adapt to the dramatic environmental and compete for limited and spatially segregated resources. variability of the late Pleistocene. Independent invention of During such times, regional cultural variation evolves in unique tool types or modification of existing tool types are isolation and natural selection acts on group-level adapta- technological solutions to local problems. Changes in mobi- tions, facilitating the evolution of cohesive and cooperative lity, food storage, or diet-breadth are organizational adapta- social networks. The process of group selection further tions. And risk-buffering strategies such as long distance allows for the rapid diffusion of cultural and technological resource exchange, codified rules for mate selection and innovation and may explain the rapid diffusion of micro- exchange, territorial defense and land ownership, and internal blade technology throughout northeast Asia during the divisions of labor and class are institutional adaptations that post-glacial period. While climate change does present chal- require coordinated behavior from a range of individuals to lenges to human survival and may promote alternative evolve. In all likelihood, some combination of technological, adaptive strategies, rapid cultural evolution is driven pri- organizational, and institutional adaptation enabled Paleo- marily by group formation, between-group competition, and lithic foragers to cope with the often-punctuated and, by the mechanics of cultural transmission. The degree to which comparison to the Holocene, extreme, environmental change climate change mediates these interactions is the extent to characteristic of the Pleistocene. which climate should be implicated in cultural evolution. Equating cultural evolution with climate change is com- plex. Clearly, changes in the local environment evince local reactions, most of which are regular features of a very flex- 1. Introduction ible repertoire of human behavior. In a constantly variable world, flexibility itself is an adaptation that evolves to meet The influence of climate change on human cultural evolu- the challenges of a relatively brief window of evolutionary tion is regularly assumed but rarely demonstrated. In truth, relevance, perhaps on the order of only a few human genera- we have a relatively poor understanding of how individuals, tions. Overly rigid or specialized adaptations do not emerge much less groups, respond to long-term environmental in variable environments. The issue is clearly a matter of change. What we do understand, we understand on rela- scale – a question of how much and how fast environments tively short time scales and our limited experience provides change (see Madsen and Elston, this volume), and whether or us with little ability to suggest how individuals, traditions, not the mechanics of human cultural evolution generate or institutions will react or respond to sustained or punctu- behavioral adaptations capable of tracking this change. Any ated environmental change. This situation is due, in part, to adaptation that initially evolves because of its ability to cope the difficulty of collecting and identifying the appropriate with the resource stresses associated with environmental var- data sets with which to track the correlation between envir- iation is naturally trained to accommodate most subsequent onment and culture and to the difficulty of controlling the variations in climate and environment. Essentially, most many interactions that separate a suspected cultural effect environmental change will have little impact on the range from a proposed environmental cause. Perhaps more of human behaviors specifically adapted to environmental DEVELOPMENTS IN QUATERNARY SCIENCES VOLUME 9 ISSN 1571-0866 Ó 2007 ELSEVIER B.V. DOI:10.1016/S1571-0866(07)09009-4 ALL RIGHTS RESERVED 105 106 L. Barton, P.J. Brantingham & Duxue Ji variability itself. Therefore, we must look to periods marked 1976; Charnov, 1976). The deterministic ‘‘patch choice’’ by significant departures from the environmental context of model generates predictions for relocation given the declining Pleistocene cultural evolution to see if and how human value of a current resource patch, the expected value of an culture responds to environmental change. alternative patch, and the cost of travel between patches. Here, we focus on the degree to which environmental A generalized prediction of the patch choice model suggests change affects the settlement–subsistence strategies of Paleo- that costs associated with relocation in a relatively unpopu- lithic foragers. Our analysis begins with a brief discussion of lated landscape are low when compared to the costs associated human mobility as an adaptive strategy. We then present a with extracting enough energy to survive from a dwindling simple method for evaluating change in human mobility resource base. The stochastic model opens the deterministic using radiocarbon data drawn from Pleistocene archeological model to the issue of ‘‘risk’’ (Stephens, 1981; Stephens and sites across northern China. The patterns in these data enable Charnov, 1982). In optimal foraging parlance, risk is narrowly an assessment of the correlation between environmental defined as the variance around the expected mean value of the change and human biogeography. Recent archeological foraging budget; the greater the variance, the greater the risk research focused on human occupation of the western Loess associated with acquiring the expected value of the total fora- Plateau during the Last Glacial Maximum (LGM) helps to ging effort. In a perfect world (i.e., the deterministic model), illustrate the behavioral and demographic context of human the expected value is always achieved and there is no risk. In a cultural evolution during periods of extreme environmental more realistically variable or unpredictable environment, the change. We conclude by suggesting that the LGM played a variance around the mean expected payoff may be so great critical role in shaping human demographic processes and that the probability of acquiring the minimum amount of advance an hypothesis to account for the rapid and wide- resources required for survival is dangerously low. spread changes in human cultural evolution during the post- Historical and ethnographic examples of resource acqui- glacial period. The prolonged environmental deterioration sition demonstrate that optimal solutions to variable environ- characteristic of the LGM was unprecedented in the environ- ments may be reached through acquiring resources from mental history of northeast Asia, and it stands to reason that multiple sources. By combining the expected returns (with such conditions may have engendered novel behavioral solu- known means and standard deviations) of multiple sources, tions from the human populations that survived it. an individual or a small integrated group reduces the net hazards of subsistence shortfall (McCloskey, 1975; Winterhalder, 1990; Goland, 1991). In the absence of other 2. Hunter-gatherer Mobility strategies for averaging risk, human foragers move frequently to increase the number of places they acquire resources. In so The migratory patterns of Homo sapiens are the subject of doing, foragers average the risk of subsistence shortfall spa- considerable study and debate. Much of our understanding tially and temporally by incorporating more foraging patches of hunter-gatherer mobility comes from ethnographic over a given period of time. This ‘‘residential mobility’’ is observation and historical reconstruction of extant hunter- therefore a simple, effective strategy for human foragers to gatherers during the twentieth century (e.g., Steward, 1938). manage the risks and hazards of resource depression and does Inasmuch as subsistence–settlement patterns, and therefore not require the additional costs or selective pressures asso- mobility, can be considered adaptations, it may also be true ciated with complex tool technology, diet breadth expansion, that multiple adaptations