Whole Systems Thinking

Notes on introductory readings

by Karl North

Summary: The systems thinking paradigm is not new: Marx, Leibig, and Darwin were 19th century antecedents. But the shift to this paradigm is finally changing the way science is done.

Particularity and separability are infirmities of the mind, not characteristics of the universe - Dee Hock, Birth of the Chaordic Age

To point out the key concept that has been lost, Fred likes to cite the subtitle of a little book by Craig Holdrege: The Forgotten Factor of Context. Context consists of two parts, both essential to understanding the whole under investigation: the spatial context, and the dynamical or historical context. In regard to space, attention to context takes seriously what I call the Russian Dolls nature of the world under study: wholes nested within wholes. A scientist studying the digestive system of the intestinal parasite of a sheep may call himself a systems analyst, but that is not whole systems science. And such a scientist may be doing perfectly good science. The trouble comes when scientists draw conclusions that require understanding of a larger context beyond the boundaries of their research. For example:

A popular parasitology textbook written by a Cornell scientist focuses on the immediate systemic environment of the parasites of common domestic animals, and on no time frame beyond the life cycles of the parasites. Most of the information the book provides is scientifically well-founded since it requires no knowledge of systems beyond the author's focus. Then the scientist offers parasite control instructions, limited mainly to extinction of the population in the intestine via parasiticides. Suddenly we are no longer in the realm of science, but rather religion. For the author has asked us take a great leap of faith. His conclusions about parasite control require knowledge of vast systems that the scientist has not bothered to describe or relate to the data of his research:

• The farmer's management system of the domestic animal and its forage,
• The farmer's management system of the domestic animal and its forage,
• The biological community of the agroecosystem in its various and potential interactions with both the parasite population and the parasiticides,
• The market system in which the farmer is enmeshed as a buyer of parasiticides and a seller of domestic animal products,
• The political economy of the pharmaceutical industry and its governmental regulators,
• And not least, the genetic evolution of parasite populations under long-term contact with parasiticides.

A cynic would suggest that the scientist may have made a Faustian deal, sacrificing his scientific integrity in return for research funding from the pharmaceutical industry. Whatever the motive, the reader is suddenly confronted with bad science. But let us be kind, and assume that the scientist made an unintentional mistake, perhaps traceable to the reductionist framework of conventional scientific training. Is this one that he might have avoided had he a better understanding of systems thinking? The most fundamental lesson of seeing the world through this lens is that there are important understandings that emerge from the study of wholes that are unobtainable from the study of parts. As one systems thinker has put it:

"At each level of complexity, entirely new properties appear. And at each stage, entirely new laws, concepts, and generalizations are necessary requiring inspiration and creativity to just as great a degree as in the previous one. Psychology is not applied biology, nor is biology applied chemistry."

Thus the sciences themselves are nested whole systems that come in a kind of logical hierarchy that descends roughly like this:

Ecology

     Biology

           Chemistry

                Physics

Such is the logic that if one tries to explain a higher category entirely in terms of a lower, something is lost. Ecosystems are not just an aggregation of life forms. And a flower studied at the level of biology has characteristics that 'emerge' at only at that level of study, and are not accessible at the level of chemistry. This seems obvious on the face of it, and yet the passion of science to reduce reality to its parts has led to constant disrespect of the hierarchy. But perhaps we can learn from history, from the holism of earlier times, what has been lost, and gain insights on how to regain whole systems perspective in the context of today's scientific knowledge.

Historical Revelations

There was a time, before dissective science began in earnest (with Newtonian physics, say), when the holistic nature of the world was taken for granted. In the Middle Ages, church theology dictated how all things fit together in the universe. Even as late as the 19th century great scientists like Darwin, Liebig, and Marx assumed that their quest for understanding required the mastery of numerous disciplines.

Liebig has been caricatured as a proponent of a narrow chemical approach to soil fertility, but his broad education extending from soil ecology through agricultural political economy led him to denounce use of his contributions to the chemistry of plant nutrition in ways that broke natural cycles that he considered imperative to maintaining a healthy metabolism between soil and plant, and man and the rest of nature. Liebig insisted that returning the nutrients contained in urban sewage to the soil is an indispensable part of a rational agricultural system. He further contended that British imperialism robbed all colonies of their fertility, and pointed to Ireland as an extreme example.

Darwin could not have made sense of the diversity of species without visualizing their complex metabolic relationships as the result of a long coevolution. Marx read deeply in Liebig and Darwin. Darwin's work confirmed for Marx the importance of history to systems analysis: patterns that emerge from the long view of a system over time are critical to understanding the system. From Liebig Marx took the concept of metabolic exchange and widened it into a socio-ecological concept. A "metabolic rift" could occur, he said, at both the human/nature level in the alienation of town and country, and the human/human level, in the alienation of labor. His holistic approach to science, that simultaneously considered historical, ecological, political, economic, and social dimensions, led him eventually to conclude that "all progress in capitalist agriculture is a progress in the art, not only of robbing the worker, but of robbing the soil."

Today leading systems ecologists as well as biochemists employ the concept of metabolism to refer to all biological levels, starting within the single cell and ending with the ecosystem. Understood as a regulatory process that governs complex material and energy exchanges between systems components and their environments, the concept has wide application in systems thinking.

But today also, after reductionist science has been delivering ever more powerful technologies attended by insufficient understanding of their consequences for the systems they are deployed in, top scientists are beginning to have second thoughts about the way science is done, and are occasionally deserting their native disciplinary habitats and creating extra-academic environments which stimulate and support integrative work. An example that is noteworthy for its intellectual synergy is the Santa Fe Institute, www.santafe.edu, where a mix of visiting and more permanent participants from a wide array of physical and social sciences, aided by the massive computer power of nearby Los Alamos, began to discover the universal properties of systems which populate and structure the landscape of each of their areas of research. Mitchell Waldrop's Complexity: The Emerging Science at the Edge of Order and Chaos tells the intriguing story of the Santa Fe Institute. What seems to emerge from this learning community is not a merely a new science, but something more revolutionary: a new paradigm for all of science. The next part of this essay will test this hypothesis as it outlines the universals that we need to recognize in order to do good systems thinking.

Systems Universals: Through the Looking Glass of Systems Thinking

Perhaps the most important property of most of the systems we must deal with in life is complexity. Compared to most man-made systems, even rocket science, most natural systems exhibit complexity of a higher order of magnitude. The consequences for science are revolutionary.

First, because of the multiplicity of variables and interdependencies, changes imposed on these systems yield multiple outcomes.

Second, in complex systems components don't just interact; the interactions are organized into feedback loops of two kinds: reinforcing and balancing. Reinforcing loops are like compound interest; they commit some part of the system to exponential growth, expressed as a percentage. The growth will accelerate until it approaches a limit. In healthy systems runaway growth is controlled by balancing loops, where initial behavior in the system is checked, not accelerated. The reproduction cycle in most species is such a reinforcing loop, checked as a species regularly becomes lunch for predator populations.

When recognition sinks in that complex systems consist most importantly of feedback loops, the certainty of the linear logic of reductionist science flies out the window, because we realize that what appear to be short straight lines of cause-and-effect are most likely only arcs in a loop where the effect can change the causal behavior.

An article in a prominent farm magazine by even more prominent scientists concluded that soils on the prairies were gradually becoming more acidic due to the use of large amounts of nitrogen fertilizers, and that acidification reduced agricultural productivity. They went on to say that really there was no problem, though, because the effect on crops could be compensated for by adding even more nitrogen fertilizer. Quite a stunning bit of logic, I thought, and I wonder where they thought that might end. I suppose we should soon find out, as reinforcing feedback loops have a way of going off the rails quite quickly.

Failure to take into account reinforcing feedback loops has caused some scientists to under-estimate two of the gravest threats to civilization in the next half century: global warming and energy deficits likely at the imminent end of the petroleum era. The result is possibly disastrous for future generations.

Multiple outcomes interacting over time, and networks of feedback loops escalating and checking each other, easily generate unpredictable system changes exemplified in the famous 'butterfly effect': such is the nature of complex, dynamical systems that a butterfly flapping its wings in Tokyo could cause a tornado in Tampa. Thus the vaunted predictive power of conventional science stands revealed as a short run affair, almost inevitably altered through time, as changes in one place ripple through larger wholes. As contemporary holist Dick Richardson states it:

"Predictability is limited to early results. To manage a complex system means that a manager must be able to detect early deviations from expectations quickly, and take actions to correct their undesired trends. If one waits too long, then the deviation is greater, and increasingly going awry, which requires a strong action for adjustment. This creates an 'overshoot' of the desired condition and an increasingly violent action-reaction cycle is set up. In electronic control systems terminology, there is positive feedback that causes increasing oscillations deviating from the average (desired) state. In a nutshell, the system quickly becomes unmanageable."

One possible consequence of all this for science is a relative loss of status of quantitative methods and a gain in the legitimacy of qualitative techniques.

The attention to system behavior over time that served Darwin and Marx so well, has led modern systems thinkers to more universal properties of complex systems. Given energy inputs, all such systems are capable of spontaneous self-organization in the direction of increasing diversity and complexity. Even relatively 'inert' mixtures of chemicals self-organize, as has been demonstrated with autocatalytic sets of polymers. This process is adaptive, entails learning, and thus causes the system to evolve. Some scientists have proposed a search for laws of the construction of complexity, which dynamic is counterintuitive to the entropy and randomization effect of the Second Law of Thermodynamics.

A first step in this direction was the discovery of a common dynamic of many small changes and occasional rare large ones that appears naturally in phenomena as disparate as species population demographics, politics, markets like the stock market, nuclear fission, and piles of sand. In the sand pile metaphor a trickle of sand dropping on the pile causes many small changes leading to a state of criticality, where avalanches of change can occur. Strategies of change in social institutions (like landgrants) depend on this concept when they propose to build a critical mass of opinion. Work with systems dynamics is beginning to suggest that the optimal state of health that allows adaptive evolution (learning) for survival is near criticality: too far in one direction is chaos and system breakdown; too far in the other is increasing rigidity and death. The pattern of organization that locates system behavior 'on the edge of chaos' and best facilitates adaptive coevolution of components, subsystems, etc. is not a hierarchal structure of central command and control, but control dispersed in the competitive and cooperative relationships of many agents.

Conclusions?

Part of why science has become a powerful tool is its universality: its ability to cut across the diverse worldviews of a pluralistic humanity. Systems thinking continues this pattern: all complex systems everywhere have the same fundamental characteristics. Thus systems thinking becomes an essential tool of science in all fields. Systems thinking is delivering a revolutionary new perspective on how our universe is organized, and how it evolves. Systems thinking is not merely one scientific tool among others; it is a new over-arching and controlling paradigm that will shape the way research must be done if it is to live up to a scientific standard. But the training of research scientists must reflect the new paradigm. The core curriculum of the first interdisciplinary graduate degree program in sustainable agriculture in the USA, created at Iowa State University, is an example of a step in that direction:

Core Curriculum

· Sus Ag 509. Agroecosystem Analysis. (3-0) Cr. 3. SS. Prereq: 6 credits in social sciences, 6 credits in natural, biological or engineering sciences and senior or above classification. Field study of commercial farming systems within the context of global energy flows and biogeochemical cycles, including ecological, agronomic, and social perspectives. (Co-listed as Agron 509, Soc 509, and Anthr 509) · Sus Ag 515. Integrated Crop and Livestock Production Systems. (3-0) Cr. 3. Alt F, offered 2001. Prereq: SusAg 509. Managing productivity and minimizing ecological impacts of agricultural systems by understanding nutrient cycles, crop residue and manure management, grazing systems, and multispecies interactions. Consideration of crop and livestock production within landscapes and watersheds. (Co-listed as A E 515, Agron 515, and AnSci 515). · Sus Ag 530. Ecologically Based Pest Management Strategies. (3-0) Cr. 3. Alt F, offered 2002. Prereq: SusAg 509. Durable, least-toxic strategies for managing weeds, pathogens, and insect pests, with emphasis on underlying ecological processes. (Co-listed as Agron 530, Ent 530, and Pl P 530). · Sus Ag 543. Organizational Strategies for Diversified Farming Systems. (3-0) Cr. 3. Alt S, offered 2002. Prereq: SusAg 509. The day-to-day operation and social relations of the complex, diversified farm. Alternative organizational strategies for the diversified and sustainable farm. Farm family dynamics and goal setting. Cooperation between farmers. The social relations of alternative marketing, including green labeling, community supported agriculture, farmers' markets, and relationship marketing. (Co-listed as Soc 543, Hort 543, and Agron 543). · Sus Ag 600. Sustainable Agriculture Colloquium. (1-0) Cr. 3. F,S. Weekly seminar for graduate students in the Sustainable Agriculture program. · Sus Ag 610. Society and Technology in Sustainable Food Systems. (3-0) Cr. 3. Alt S, offered 2003. Prereq: SusAg 509. Social and technological dimensions of sustainability in food systems. Emphasis on strategies and ethics for evaluating existing and emerging options. (Co-listed as Soc 610 and A E 610, cross-listed as Anthr 610). In addition to courses in the core curriculum and the colloquium, credit may be obtained for work in the following courses: · Sus Ag 599. Creative Component. Cr. var. F.S.SS. Pre-enrollment contract required. Advanced topic for creative component report in lieu of thesis. · Sus Ag 699. Research. Cr. var. F.S.SS. M.S. and Ph.D. thesis and dissertation research.

-- Karl North

URL: www.geocities.com/northsheep/systems.html

Top