The science of organized complexity is interdisciplinary and best integrated by generalists, but its problems are often oversimplified by mistakenly - or deliberately - applying the wrong tools to understand those problems.
By Ted Mitchell
Published December 14, 2004
Scientific problems can be sorted into three types, with a tool for each type. People often fail to identify the correct type of problem and are then misguided by thoughtlessly using the wrong tools.
The first type of problem is found in physical sciences, such as Newton's mechanics. The appropriate tool is two-variable analysis: change one variable and observe the response in another.
The second type of problem arises when there are far too many variables to consider. Statistics can find predictive value out of apparent disorder. Major advances in quantum physics and chemistry have resulted from this approach. In this model, individual behaviours need not be measurable or even noticeable, but the whole behaves in a predictable fashion.
The third problem, between the two extremes, is what Dr. Warren Weaver calls "organized complexity." Here, several variables contribute to a problem in interconnected but not clearly defined ways. One commonly finds this type of problem in such diverse areas as biology, medicine, urban planning, and public policy. In this case, applying the first and second tools can be helpful, but reliance on these methods alone will be insufficient to understand fully and address these fundamentally different problems.
The appropriate tool relies on the unique integrating abilities of the human brain. The science in this kind of analysis tends to be "empirical", namely, derived from real world observation. Because of this, it is absolutely critical to be objective, or the result can be seriously misguided. Conflicts of interest or preconceived ideas are toxic to this type of science.
Jane Jacobs raises the issue in The Death and Life of Great American Cities. Dig up Jacobs' analysis of how organized complexity describes the types of problems that cities encounter, and don't be surprised if this 1961 work is still relevant. We have made little progress in this realm of thinking, likely because the practitioners of this science are undermined by strong political influences.
A good example of the third type of problem is new suburban development. There is ample evidence that such development is more costly in the long term, requiring more asphalt and water pipe to maintain per assessment dollar. Sprawl means isolation, and physical activity loses out to car-dependence. Medical studies have shown that suburban residents are less healthy in a number of areas. It should be obvious that these low-density areas are largely devoid of economic activity, much less economic diversity. The same is true of the social environment, as any teenager will tell you. These factors are not unrelated or unpredictable.
Hamilton has been a leader in approaching such problems with the correct tools, as evidenced by Vision 2020. But understanding has not informed action. Politicians who are evidently bad at math (or have developers for friends) still sing the economic praises of a new sprawling assessment base while ignoring the related degradation of the city core. They are oblivious to recent history, which repeatedly shows this kind of development to be harmful as well as a poor long-term investment.
The Kyoto Accord is another organized complexity problem where governments have plainly applied the wrong tool. This proposal reduces a fundamentally complex issue down to one variable: carbon dioxide (CO2). The theory is that climate change should be less likely or severe if we reduce our total output of CO2, but there are many interdependent variables, not just CO2.
Altering the ways we produce CO2 is fundamentally intertwined with a variety of other factors, the foremost of which is air pollution. Simple measures directed only at reducing CO2, like car-pooling, might also reduce pollution. But other narrowly applied measures, like switching from gasoline to fuel-efficient diesel, reduces CO2 but markedly increases other pollutants.
Other factors related to Kyoto are fossil fuel reserves, short and long term economics, the physical form of our cities and dwellings, human physical activity and social interaction, rates of disease and death, and global political instability, to name but a few factors that affect humans directly. Politicians and corporations have made no meaningful attempts to explore these interrelated factors. They appear unable to look beyond self-interest.
Health care also suffers from overuse of the wrong scientific tool. Medicine is fundamentally a multi-variable problem, and is suffering from excess statistical analysis at the expense of common sense. (Hamilton is actually the birthplace of "evidence based medicine," or the rigorous application of statistics to treatments.)
Physicians cannot treat patients solely with statistics. If this were the whole story, a visit to your family doctor would involve recording symptoms and making a diagnosis based on probability. She would give you treatment options, each with a unique probability of working or not, how much benefit you could expect, and the risk of harm and what form that could take.
But then you mention stress at work and with the children, and all the unique aspects of your individual health. The statistical method breaks down, because you are no longer comparable with the people in the study. For this new, unique problem, the doctor has no relevant studies, no tests, and no specialist to fall back on. This is a problem that only "thinking on your feet" and integrating many different factors can resolve.
These and other problems of organized complexity appear to be consistently addressed with the wrong tools. Perhaps because the appropriate science is fundamentally imprecise, it is an easy target for self-interest groups calling for "more evidence." This predictably leads to inappropriate simplification.
The science is interdisciplinary and best integrated by generalists, but our society expresses disrespect for "Jacks of all trades, masters of none." Consequently, try to find a family doctor, when specialists earn more money and respect. Finally, this approach requires honesty and freedom from bias, attributes that appear to be growing scarcer with time.
Problems are what they are, and artificial simplification is really scientific deception. Progress is made only when the type of question is matched to the proper tool. Any other approach is bound to fail.
You must be logged in to comment.