http://web.pdx.edu/~rueterj/ia/Appendix1-pattern-catalog/index.html

Appendix 1: Catalog of patterns of interaction

A. Introduction: Simple and Complex Patterns

One of the first steps in solving any problem is identifying that something is wrong, i.e. that what you observe doesn't fit with a set or acceptable pattern. As humans, we are keenly in tune with some patterns and unaware or oblivious to other patterns. Even as we experience our environment in our non-scientific, day-to-day, life we see patterns that we act on. For example, we see the weather change over the day and we expect rain.

The scientific method builds on our observation skills by formalizing the process to hypothesize an underlying mechanism for the pattern we see. From the pattern (observation) we draw on our experience to make a prediction about what will happen (hypothesis). Some problems that we deal with have simple cause and effect relationships. For example if we see a clump of dead trees by a newly excavated house lot, the simplest cause is probably that a bulldozer drove over the roots and caused damage. This would be easy to check by examining the soil around the trees. Other observations might be based on data that is collected over time or space or a range of concentrations of some pollutant. From these data we might see a relationship that is a simple correlation, the change in one factor is a major predictor of the change in the dependent variable. Again, a simple example could be that the number of dead plants might be correlated to measured levels of a pollutant at each spot. You will encounter many of these correlations in environmental science. Our discipline has well developed protocols for making observations that can be studied using statistics to test the validity of the prediction. We have made progress using these methods but there are limitations.

Some of the processes that we should be observing are not the result of a simple cause and effect, but rather a more complex interaction of parameters. In a complex process, components of the system interact continually and any change in one part is accompanied by changes in other parts of the process. It is important that we learn how to identify these complex patterns. The purpose of this catalog is to introduce you to a variety of patterns.

If you are studying an environmental problem that is built on simple, cause and effect, relationships it is very valuable to create strong hypotheses based on general laws of chemistry, physics and biology. These laws will help you determine what parameters should be measured. While this process is very efficient, it tends to pre-bias the student to observe some parameters and ignore other details. To learn about and study complex systems the student needs to start by considering a wider range of alternatives and pay attention to the details. This process may be much less efficient in terms of the progress made toward a coherent description of the system.

In another paper I describe how to learn about complex systems (Teaching about complexity). I claim that the study of complex systems requires the following five components:

1. Students need to develop a wide repertoire of metaphors for systems that have complex interactions.
2. Students need to have first hand experience with the details of complex systems.
3. Students should explore the range of outcomes available for an active complex system using simulations.
4. Students should collect observations and data from these systems without being pre-biased with a set of generalized laws.
5. Students should create meaning for their observations by constructing explanations that highlight the relationships they observed and extend that by using simulations and more sophisticated analytical tools (statistical methods).

This catalog is intended to help you build a wider repertoire of metaphors for how complex systems work. In an introductory course like this, we expect that students will make progress on steps 1, 2 and 3. Steps 4 and 5 are require much more time than is available here.

B. Classes of Patterns

This catalog is divided into eleven classes of patterns ranging from simple to complex to multiple-complex systems. The purpose of this categorization is to facilitate learning and teaching. Each category has similarities that can be discussed and leads to the next levels of complex systems. A particular category should share a general metaphor. These general metaphors contain the salient features of the system. For example, "swarm" patterns are all related to a group of bees or any colony. There are simulations and analysis tools that are appropriate for understanding and describing these patterns.

The catalog contains the following patterns that can be related to their dominant metaphor.

C. Table 2 - links to images