courses/complexity/index.html
20004.02.18
This collection of short papers explores how we can teach and learn to make decisions in complex systems. Even if we were working on a problem in environmental science for which the "science" or "technology" portion were deterministic and linear, the larger context of any decision made to implement this solution would be in the complex contexts of the economy or society. Since all of our work will eventually lead to a complex decision, it may be valuable to address our scientific and technological contributions in a more general framework from the very beginning.
In these papers, the complex system approach is assumed to be more general than the deterministic/equilibrium approaches. What we know about deterministic systems serve as useful examples and, in fact, the deterministic approach may provide a perfectly suitable solution to some of the problems that we face. My claim is that we won't really know whether the deterministic solution is appropriate until we set that in the larger context, i.e. until we test the assumptions that limit the parameters against the larger parameter space encompased by complex systems.
There are six or seven articles or sets of notes in this collection. They have been prepared as independent papers that are interelated.
status = draft
Game theory provides a broad context for making decisions. The precautionary principle is invoked in environmental science as a way to make decisions to protect the environment against unintended consequences. As a specific algorithm for making a choice however, the precautionary principle misses some of the alternative strategies that could be considered. In particular, the precautionary principle is based on a causal model that links particular events to outcomes.
**comparison of IE to committed environmentalist perspective **
IE: system, causal, economic tools
CE: network, emergent, unmeasurable values
date = 2004.02.18
status = short draft
Actual decisions in complex systems will have to be driven more by empirical data and inductive reasoning than a known set of possible events and outcomes. The scale of the decision is a critical part of the decision itself; if the system is either small enough to be dominated by the decision maker or large enough to be insensitive to the decision maker's choice, the decisions can be simplified into a game style choice. If however the system is at an intermediate scale there are many actions that the decision maker can take that will actually help determine the fate of the system. For example, an actor could choose to demonstrate that a particular process is feasible even if it isn't economically sound. Such actions can lead to alternative paths including some that might have positive feedback effects.
status = short draft
date = 2004.02.19
The five principles of teaching about complexity are reviewed and the explored as guidelines for collecting data. It is assumed that the decision maker will never be able to have complete information about the system and thus will have to work in a bounded rationality mode. One of the ways to deal with limited information is to determine what decisions are possible and limit the information collection to address these decision choices. As strange as it might seem at first, the seemingly nebulous and certainly complex area of social values and culture form a strong set of constraints to possible decisions. Additionally, valid management strategies for complex systems are also limited. The combination of social and management constraints provide substantial guidance to scientific and engineering efforts.
status = draft
date = 2004.02.18
There are five principles of teaching and learning about complex systems; metaphors, experience, simulations, inductive data collection, and inferential analysis. The teacher or self-directed learner needs to establish both a wide set of metaphors for complex systems. In traditional science education that focuses more on deterministic processes, there is a gap between concepts and associations and the application of this knowledge with scientific tools. The reason for this is that most of the analytical tools used in the traditional context are based on deductive approaches and the power that comes from that generality. Instead of having to jump this chasm, complex system students need to wade through the swamp of rich, personal exposure to some complex systems. From this experience they should be clear that the simulations and resulting analysis can only possibly capture some of the features of the system, and that it will be up to them to craft which parameters and features are used. Simulations are powerful tools for exploring the response of complex systems to changes in parameters, structure or initial conditions. Practice with appropriate simulations is crucial for students to develop the ability to spot patterns. Collection of information can be guided by experience from the simulations but shouldn't be constrained by the invocation of natural laws. Finally, the data should be analyzed with appropriate tools that search for patterns. These inferential tools can be applied to simulation output for the student to gain experience at detecting and rejecting patterns. As with any learning activity these principles need to be simultaneously incorporated, not performed as discrete steps. Three examples are provided that demonstrate the principles, incorportation and iteration.
status = draft with incomplete added material
There are many types of patterns in the information that we use in environmental science. These patterns range from deterministic, cause and effect patterns in the data to complex and emergent patterns that can only be understood using more complex underlying models. We can learn about and describe these patterns in two ways. First we can make a catalog of as many processes and patterns as we can. This catalog can be used in both directions, from process to typical patterns and to sort patterns for possible underlying mechanisms. The second, and more general way to describe patterns, is to use a grammar as suggested by Alexander et al. (1977) that includes the context, forces and resolution of the pattern. In addition to these general approaches there is a list of complex patterns. Because of the importance of energy and matter dissipation in the human/nature interface dissipative structures and patterns formed by these structures are discussed.
status = notes
The concepts of ecosystem health (from Costanza) and sustainability are combined to examine possible management approaches. The chosen definition of health is that the system will recovery from a disturbance will result in some type of benefit. In these systems, we should employ techniques which manipulate the purpose of individual components within a narrow range of the systems ability to handle the stress. The final component of this approach is to determine what people really want and how the desires for longer time scale "stability" can be exploited.
exploring how humans solve complex landscape problems and whether these skills can be directed at other environmental problems
exploring many other native skills humans have for solving complex problems including; social networks, stigmergy, grammar, kinship, ..
some problems may be ammenable to applying human skills
some other problems may be totally foreign to our way of thinking. I want to explore whether these novel complex problems are a threat to sustainability, i.e. if we can't make sense of some issues.