An activity of the Primer Group


A Special Integration Group (SIG) of the
International Society for the Systems Sciences (ISSS)
originally SGSR, Society for General Systems Research.






What is a System?

by Bill Shireman

There are two ways to approach a problem: by attacking it with a program, or end-running it with a system.

A program is a single action or set of actions intended to influence something outside itself. Typical programs might include: a crime enforcement program, a jobs program, an environmental program.

A system is a set of interrelating parts that performs functions inside itself. It is a set of ideas or objects that must work together to perform a particular function. Typical systems: ecosystems, economies, and within them subsystems such as biosystems, communities, businesses, transportation networks, watersheds, health care, communications, animals, automobiles, human bodies, and so on.

Programs are simple and direct. They confront a problem head-on: a program is instituted which is intended to resolve a narrowly-targeted problem. Politicians like them because they appear to be serious, tough, decisive, and sound-bitable.

Systems are complex and indirect. They are vast fabrics sewn of individual causes-and-effects, which when taken together perform a function which overcomes their individual limitations. Systems deal with problems indirectly; the problem tends to be solved automatically, as a natural consequence of the existence of the system.

For example, under a crime prevention program, crime might be stopped by enforcement agents that intervene directly, at a cost to taxpayers. Under a crime prevention system, crime might be reduced automatically, as a byproduct of another action a system of potential penalties, or perhaps a public education system that provides better opportunities reducing crime because people have less desire to commit it.

Similarly, under a forest protection program, deforestation might be stopped by hiring police to arrest violators, or land might be purchased to take it out of the marketplace. Under a forest protecting system, the costs or penalties to the deforester would be generally sufficient to prevent the action, and the land might be of such little value as a source of poorly-harvested timber that no successful business could engage in deforesting practices for long.

Both programs and systems are processors of resources. They process three types of resources: energy, matter, and information. But systems are distinct from programs in that they also create, or internally generate, at least one kind of resource: information.

Programs are intended to accomplish one thing. Systems accomplish many things. A forest protection program is intended to do that only (although it always produces expected and unexpected consequences as well). A system does many other things; forest protection is an indirect but necessary byproduct.

Programs require that money be diverted from something else. They are entropic they require constant inputs of resources from the outside. Systems can be self-supporting, though perhaps over politically and culturally inconvenient timetables. They are synergetic they are more than the sum of their parts. A forest program requires that someone spend money on something from which they don't get a return it requires involuntary sacrifice by someone. A forest protecting system requires no sacrifice only tradeoffs.

Programs are reactive they happen after the fact. Systems are preventive.

Programs are usually divisible. Systems are not. A program divided in half is about half as useful as it otherwise would have been. A system divided in half is fundamentally different, and far less useful than as a distinct whole. A log divided in half is half as useful; a tree divided in half is at least half-dead; a human body divided in half loses its life.

Programs operate within systems. Systems are catalysts and enablers to programs. Many policies are a mix of program and system. A program of road-building is initiated to institute a system of transportation. A program of school construction and management is instituted to provide a system of education.

A system is less efficient at doing just one thing than a program. But it is more efficient at doing many things. Example: a forest system is less efficient at growing board feet of marketable timber than a plantation program. But it is much more efficient at providing for a variety of needs timber, habitat, genetic diversity. A plantation program is profitable because it harnesses a biological system that grows fiber. A school system is less efficient at crime prevention than a jail, but it is more efficient at growing well-rounded citizens with a variety of positive attributes.

Programs exert influence because of their physical dimensions: money spent, power exerted. The more money spent, the bigger their impact. Systems exert influence because of the intelligence of their design. The better their design, the less money needs to be spent.

When systems are optimized, their programs are individually suboptimized. When programs are optimized, the systems of which they are a part are suboptimized. From a programmatic perspective, reducing crime by improving schools is considered inefficient far less direct and more expensive than punishing criminals. But from a systemic perspective, reducing crime by improving education is most efficient, because it happens automatically as a consequence of a system that accomplishes many other things as well. While the route is long and indirect, the destination is generally reached more surely and at lower cost with a system than a program.

Programs within healthy systems are always constrained from explosive growth by competition from other programs. Example: If an economy optimizes for one program, like weapons production, then the system as a whole is less productive. If a forest is changed to optimize output of a single species of plant or wood, then the forest as a whole is generally less diverse, resilient, and productive over the long term.

Systems are non-physical; they are ideas or designs. Programs are physical. Because of this, systems can be sustainable. Programs cannot be. They are constrained by the laws of thermodynamics. Systems, by contrast, operate by laws of ecodynamics or system dynamics:

Systems provide the resources that fund programs. Systems are the only source of resources. They are the only source of profit and wealth. Programs are merely channels for the expenditure of resources generated by systems. Systems can be self-regulating. For example, consider cybernetic systems like the human body such systems rely upon a limited number of inputs and adjust their operations to optimize the use of these inputs. Programs always require external regulation, and there are no inherent mechanisms to regulate programmatic operations.

Systems sustain themselves by following three principles: individual elements of systems become more specialized. As they do, the system as a whole naturally becomes more internally diversified. And the diverse elements of the system become more complex and interdependent.

These three are the operating principles of all sustainable systems. They are the means by which sustainable systems overcome the chase of entropy. For example, as the organisms within ecosystems reach physical limits or experience competition, those with a comparative advantage a specialty naturally survive and excel. They fill narrow functional niches to which they are particularly well-suited, thus outlasting or avoiding direct competitors and helping the overall system to operate more efficiently. As the individual components so specialize, the overall system naturally comes to be comprised of a more diverse array of organisms or components, each performing a niche function with particular adeptness. This increases system energy efficiency, which counteracts entropy. And as the system becomes more internally diverse, it becomes more complex and its elements more interdependent more reliant on one another for mutual gain. Simple economies or ecosystems, for example, are dominated by a few highly competitive organizations or organisms; like monocultures, they are highly vulnerable to change. More complex, sustainable ones are comprised of a richer mix of components; because of their diversity, complex systems adapt more readily to change.

Systems gain attributes qualitatively. Programs gain attributes quantitatively. A healthy program grows by getting physically bigger. A healthy system evolves by becoming more complex and "smarter." It passes certain thresholds, after which it is qualitatively different. In a simple physical system, matter is inert. As the system evolves, matter reshapes into more and more complex forms which at first become replicating, then alive.

Systems learn and change. Programs don't, except when compelled systemically. Living systems tend to be self-correcting through a variety of teaching instruments, like:

Systems evolve by using ever-increasing increments of information. Programs evolve by using ever-increasing amounts of physical resources.

Systems range from the least open to the most open, and from the least complex to the most complex. Open systems receive inputs (energy, matter, information) from outside; closed systems do not. Simple systems consist of few elements in interaction; complex systems consist of many.

Simple, closed systems are spent quickly. Because they are closed, the laws of thermodynamics dictate that their energy content will pass from the more dense to the less dense or, in terms of heat content, from the hotter to the colder. Their relative simplicity means that they will not have the creative means to withstand challenge or counteract entropy for long.

Complex, open systems like Earth are spent slowly, if they are spent at all. Because they are open, they are part of larger systems which create or distribute new resources to them, in the form of energy, matter or design and information. More important, because they are complex, they have a higher capacity to be creative, taking full advantage of the attributes of a more diverse array of components, and therefore are more fleet in escaping the approach of entropy.

Programs are simple open systems in physical form: highly entropic, vulnerable, and non-sustainable.

The most sustainable systems are those which are most complex and open. The least sustainable are those which are most simple and closed. Simple, closed systems are subject quickly to the limits imposed by entropy, and do not have the internal complexity to escape through negative entropy or synergy.

Open systems may lose system integrity if they are overwhelmed with resources from the outside. By losing integrity, they lose their individual distinctiveness. Simple open systems are most easily subject to a loss of integrity, and assimilation into an outside system. Complex open systems are relatively more immune to loss of integrity, because the resources which enter from outside are processed through and changed as a function of their complex design. System integrity is maintained, but in a new, more complex form. The system, in fact, couples with the system which provided the windfall of new resources, yet it maintains its distinctiveness; it is a new system evolved from its prior form, but modified by influence from the outside. For example, the emergence of industrialization led to a massive infusion of resources into virtually every global culture. Simple open cultures were often overwhelmed by the money, resources, or ideas of industrialism, and many of those cultures seem to have essentially disappeared, having been assimilated into the mass. Complex open cultures managed to maintain their traditional practices and beliefs within a new context, and often coupled with the industrial system to evolve their own cultures into even more complex forms. Now, many of those complex traditional cultures are injecting their own resources, in the form of perennial belief systems and views of nature, back into the industrial culture, leading to its further complexification.

System coupling is a function of how closely two systems operate in time and space, how similar or complementary they are, and how open they are. Traditional cultures within the United States and Europe were virtually decimated by the wealth and power of industrialization. Traditional cultures which were more closed and operated further from centers of commerce were more protected from external challenge. Those traditional cultures, whose attributes include elements lacking in industrial culture, are now often seen as complementary to industrial culture, and are beginning to influence it as the latter enters a period of challenge, self-doubt and change.

No system, except perhaps the system of all (the "universe" or whatever encompasses it), is completely closed, since maximum entropy would lead to its dissolution.

Global Futures Foundation, 3090 Fite Circle, Suite 203, Sacramento, California USA 95827
Voice (916) 368-5200 Fax (916) 368-5250 Email

Send e-mail to

Return to home page

Return to map page