Adaptive Modeling
Modeling experiments during the 1970s raised the issues of global interdependencies and universal interactions between human, technological, and natural systems for the first time. But due to a lack of dynamic data and limited knowledge of the nature of these interdependencies and interactions, models became imprinted with the normative assumptions of their creators. This text, published in support of the seminar Modeling Wicked Problems, which took place during Anthropocene Campus 2014, considers the work of Canadian ecologist Crawford S. Holling, who developed adaptive modeling with the aim of circumventing this recurring problem.
The models attained scientific effect just from the use of computers—a critique that has been formulated throughout the scientific community. Another approach to modeling—again from the 1970s—not only seems to acknowledge the criticism that has been provoked by these world models, but also appears as a reaction to some general “neo-catastrophic”1 trends in scientific theory of the 1960s. For example in evolutionary biology, paleontology, astronomy, or the climate sciences, theories were proposed that featured violent change and shock, or emphasized unexpected surprises and fundamental uncertainties in relation to complex dynamic systems. Edwards Lorenz’ Chaos Theory or René Thom’s Catastrophe Theory are just two examples with broad applications across different disciplines. One conclusion drawn from these theoretical foundations was the recognition of some general limits to planning and prediction and the “wicked” nature of some problems. The Canadian ecologist Crawford S. Holling explicitly acknowledged this situation both in his theoretical approach to ecosystems as well as in his methods to manage them. In his widely cited paper “Resilience and Stability of Ecosystems” he developed a perspective that no longer focused on notions of “equilibrium” or “balance of nature,” but in which, instead, he defined the complex measure of resilience that can account for the ability of an ecosystem to remain cohesive even while undergoing strong perturbations. In Holling’s systems thinking, “the constancy of [the systems’] behavior becomes less important than the persistence of the relationships.”2
Within such a framework, the primary objective is not the maintenance of either a certain status quo or the adherence to global goals but, in line with other pioneers of the Anthropocene, Holling understood development as arising from the coevolutionary interplay of environmental and social systems. Thus, while previous models and environmental assessments have provided mainly static “snapshots” of the world, modeling, in contrast, should enable actors to actively engage with (eco)systems to allow for coevolutionary development. As professor and director of the Institute of Animal Resource Ecology at the University of British Columbia, and especially as head of the Ecology Project at the International Institute for Applied Systems Analysis (IIASA), Holling invented processes of adaptive environmental assessment and management, which earned him a wide professional following beyond that of the environmental sciences. The adaptive management process consists of a carefully designed series of alternating workshops and research periods in which scientists from different disciplines, policymakers, and other stakeholders come together. Systems analysts helped to identify the various aspects of a problem, translated them into variables and integrated them into a model, with which the participants play in order to explore the potential impacts of different policy options, the sensitivity of the system under observation, and the sensitivity of the simulation model itself. Models used to foster such learning experience were understood as evolving devices of self-instruction. The value of the model “is not so much to give answers as to generate better questions,” Holling and his colleague wrote.3 During the 1970s, adaptive management was tested in numerous different case studies addressing complexities that systems analysts tended to ignore by then—multiple conflicting decision-makers, multiple conflicting objectives, intertemporal and intergenerational trade-offs, and the design of strategies that deal with the irreducible uncertainty of the real world, for example by avoiding premature foreclosure of options and by developing policies that are as robust, viz. resilient, as ecosystems. One study was conducted in cooperation with the Alpine Research Center of Innsbruck University, the UNESCO Man and the Biosphere Program (MAB), and the inhabitants of Obergurgl, a Tyrolean Alpine village that the researchers presented as a microcosm for the problem of economic growth in relation to limited ecological and social resources.
Major components of the Obergurgl model. Dotted lines show areas of workshop subgroup responsibility. Taken from the Alpine Areas Workshop, IIASA Conference Proceedings CP-74-2, 6.
The illustration shows the basic identified components—from hay production to parking spaces to birth rates of humans and animals—and how these components interact in the Obergurgl model. The model structured the workshop series and stood at the center of negotiations between the different stakeholders. Holling summarized the process:
“Its purpose was to examine the likely consequences of several options available for this high alpine region of Austria: zoning changes, building subsidy or taxation, ski-lift construction. In a 5-day workshop a model was built, and the alternative futures under the different options were examined. The results of this exercise became a topic of major consideration in the region, and we believe they made a significant impact on decision-making. After a 1-day planning meeting, a core group of 5 methodologists and 15 participants met for a 1-week workshop. [. . .] Several important problems were defined and clarified by the Obergurgl model. The initial concerns about environmental quality receded to minor significance. Of more concern was the obvious inability of the village to maintain its current style of life, which is associated with continued growth of the hotel industry. The land will run out; subsidization, taxation, and zoning changes can only alter the date. When the Obergurglers returned to their village after the workshop, they initiated a series of public discussions about the future of the village. […] The need for a change in lifestyle and expectations became obvious to many of the villagers; the search for a solution began.”4
In this participatory approach to adaptive local management, models are essentially embedded in a collective process. Models are not just representations of the world, but platforms for an adaptive engagement with changing environments and understood as tools of self-instruction and collective learning about complex systems and their possible future trajectories. They structure language, coordinate people, and help to make robust decisions that help ecosystems to retain resilience.
