Complex Adaptive Systems
- Complex, dynamic, opportunistic, self-organising systems operating at every level from the cellular to the galactic, characterised by the property of emergence, living on the edge of chaos and always moving into adjacent possibilities. Human beings, societies and ecosystems are all examples of CAS.
The following account of CAS is taken from Jack Huber's 2013 book: The Future of the Mind.
"There is as yet no comprehensive theory of these complex systems. But like the elephant partially explored and described by the blind men, these systems have been described in various ways by different scientific disciplines: non-linear systems, chaos theory, complex adaptive systems, network systems, and emergent systems are some of the perspectives given to them. I will use the term ‘complex adaptive system’ and for the sake of brevity, I will abbreviate the term to CAS.
But be warned! CASs can seem counter-intuitive if not incongruous. We have an innate confidence in the regulated processes of life—cause and effect flow everywhere. Perhaps, as Douglas Hofstadter suggested, CASs are a little unsettling. Rudderless, without direction or objective—except for an unceasing urge to reshape themselves, to self-organize and self-organize and self-organize... ad infinitum. Yet, they are where this story begins. CASs are all around us. They are us and they are the environment. They adapt to change in their environment while becoming the changing environment of other CASs. They are a fundamental part of this story and the future of the mind.
CASs are distinct from systems that are only complicated. I may cut the grass with a lawn mower—a relatively complicated machine of numerous interconnected, interacting systems and parts, including starter, fuel, ignition, exhaust, propulsion, cutting, and grass handling. No matter what the conditions or the weather—hot or cold, snow or rain, or the terrain—smooth, flat, or filled with gullies—the machine will attempt to cut. No matter what is encountered—grass or weeds, bricks or twigs, flowers or garden hose—the lawn mower will do the same thing. It will attempt to cut whatever I put in its path. The lawn mower just sits there through rain, snow, heat, or cold, unless I direct it. It does not adapt to change in its environment. But the grass, also complicated in nature, is a self-organizing system. It changes with no outside or centralized direction. It adapts to the stimulation of seasonal change. Grass withers in the face of dryness to preserve its future, grows with nutrients, reproduces, and goes dormant in the cold, saving itself for recovery in the warmth. Over eons of time, grass has adapted to changing environments.
The lawn mower is complicated. The grass is a complex adaptive system.
So why are we interested in CASs? These systems can be found at the cellular level and every other level all the way up to ourselves... and beyond! They are hierarchical. What do I mean by that? We normally think of hierarchies as levels of cooperation or collaboration within an organization like a church, or a firm, or an educational institution. In the case of CASs, the hierarchies come about through an interesting aspect of their organization. A CAS, once formed, is then available to be used as an element in the formation of a higher-order CAS. These higher order levels can carry on without limit. Thus we have cells organized into systems into organs into people. We have individuals organizing into consumers or producers, groups into markets, markets into economies. They are all around us! And we need to understand a few things about this CAS phenomenon before proceeding.
So what makes up these systems? How do they work? If they are not fully understood by science, how can they be useful? There are several known aspects of these phenomena that are of interest to us.
- They are comprised of independent, basic elements—ranging from cells to consumers to galaxies.
- These elements self-organize in response to changes in the environment.
- The self-organizations balance between stability and chaos—on the edge of chaos.
- They only self-organize into possibilities adjacent to their initial conditions.
- They harbor an unpredictable property: emergence.
The Basic Elements
All CASs are comprised of building blocks, independent elements interacting with one another for the benefit of the whole. You scratch my back and I’ll scratch yours? Well, it is not quite that simple. Together, the interacting elements are better able to interact with their environment—better than any individual element could do on its own, and they collaboratively develop a capability for manipulating that environment for the good of the CAS and for the good of each of the elements. Even so, CASs are not necessarily optimal. Typically, they are just good enough, good enough to persist, good enough to survive, and good enough to multiply. I will explore this characteristic of CASs throughout the book.
The basic elements comprising a CAS need not be identical so long as they relate on a common basis. This characteristic of the element is illustrated by the object-oriented construct used in computer programming, where objects belong to a substitutable class no matter what their other characteristics, so long as they retain the characteristics of the class. They may differ in size, scale, or properties. So long as they relate with the same ‘language,’ function together in a common interest and maintain their integrity, the elements may be of any mix. They need not be ‘living,’ or mineral, or even sentient, (though they may be all of these and more). The elements comprising economies and societies—both CASs—demonstrate this aspect of the elements of CASs. CASs can include many disparate elements—even other CASs.
The interfaces among elements determine the manner of exchange—the modality of interaction among the elements in a system. In an economy it may be currency; in a brain, dopamine. This modality both supports and constrains the relationship of the elements and contributes to the simplicity of the relationships. These interfaces support feedback for two-way interaction and exchange. The most familiar example of this modality constraint is our senses. We smell odors, not light.
These interactions can be seemingly chaotic. Elements may participate in multiple CASs simultaneously. It’s like belonging to a poker club, a bridge club, and a pinochle club… on the same night, in the same room, at adjacent tables. Elements need only adhere to the protocol of each of the CASs in which they participate. This property is evident in economies where individuals can be manufacturing workers, service providers, consumers, or facilitators of the interaction of others. Multiple roles are also apparent in societies, and in the central nervous system. Disintegration, or perhaps disengagement, occurs when participation in a CAS leads to conflict among the functions of the elements. We see this in bankruptcies, corporate buyouts, revolutions, mental disturbances, and the ultimate cellular disengagement—death.
There’s more. Since any element in an environment is eligible to participate in a CAS, the CASs themselves participate in relationships with other CASs, self-organizing into an endless potential of layers or hierarchies. In higher order systems within hierarchies, the interfaces may extend deep into the systems and be comprised of other CASs. The hierarchies of interaction can be limitless—important to the future of the mind.
A World of Elements Responding to Change
These systems are all around us. They nurture us. They shape us. They comprise us.
CASs are opportunistic, adapting to change and to stimulation from without. Each CAS functions in the context of its own environment. It is subject to the influence of that environment and self-organizes in response to changes in that environment—but on its own terms. The availability of new or changed elements, the disappearance or modification of elements by the environment, all represent opportunities or threats stimulating self-organization.
Since CASs self-organize as elements of the environment, comprised of elements of the environment, and in response to disturbances in the environment, this self-organization reflects the environment. Self-organizations are consequences of stimulation by the environment; they are an integral part of the environment. The current state of a CAS reflects the culmination of all previous self-organizations, all previous responses to the environment. The structure of the CAS represents its response to the environment.
As a CAS changes in response to stimulation or opportunities, it also becomes the changed environment encountered by other CASs. It is the environment to other CASs. The relationships and patterns of self-organized systems change as the environmental circumstances do, and environmental circumstances change even as CASs do. In every sense, the self-organizations of CASs reflect their environment and are part of the environment of other CASs. They co-evolve.
Self-organization at one level of a hierarchy may or may not cause reorganization in other levels of the hierarchy, depending on whether or not the elements in other levels change. We see this in economies and societies. There may be changes within a business, or an industry, or a new set of laws to cope with societal changes, yet the economy—or societal relationships—persist.
Living on the Edge
Chaos is a term encountered in the study of CASs. Complex adaptive systems encompass both stability and change. They exist at the transition between orderly systems with stability and those that are chaotic in behavior. They truly skirt the edge between stability and chaos. CASs are restless, seemingly without discipline. But there are limits to this restlessness.
CASs are governed by their own emergent rules. The rules for each arise from within and are used by the elements participating for the benefit of the CAS as a whole and the individual elements comprising it. Yet CASs need integrity. When an environmental influence provokes change that is not within the scope of the existing system, the system disintegrates. Poof goes the CAS! The independent elements are then freed into the environment, available for new self-organizations, new CASs. The assets, competitive advantages, skilled workers, trade secrets of commercial organizations in bankruptcy, are dispersed to other organizations capable of using them in their systems. Mergers and acquisitions often ‘spin-off’ unwanted or unusable functions. The chemicals in the cellular makeup of our bodies return to the environment in death.
This brings us to autopoiesis. Autopoiesis, defined by Maturana and Varela, is the process of becoming. According to Maturana and Varela, autopoetic systems literally pull themselves up by their bootstraps and are continually self-producing. The only product of their organization is themselves and there is no distinction between the producer and the product. Sound crazy? Well, as we used to say, “you are one.” This process, and every CAS, is constrained both by past structures—by the history of previous self-organizations—and by the need to maintain ongoing structural integrity from moment to moment. Otherwise there is an end to becoming. Stability—no change—is death. Chaos is the collapse of system integrity—also death. Self-organizations and reorganizations that take place within a system while maintaining the integrity of that system operate at the edge of chaos and are said to have the property of autopoiesis: becoming is a process without an end.
This can have unexpected consequences. On the one hand, CASs can have the attributes of both stability and chaos. Well, maybe not simultaneously, but almost. On the other hand, tracing the path of a CAS through history might give the impression that it is seeking a direction—an objective—when it is simply following the path of least resistance through a series of opportunistic self-organizations, moving from one set of environmental conditions to another, and to another, and to another.
These systems can also spontaneously self-organize into sub-systems—into a division of labor. If we consider that these systems are constantly seeking new possibilities at all levels of a hierarchy, these self-organizations into sub-systems are simply moves into adjacent possibilities.
Initial Conditions, Adjacent Possibilities, and Adaptation
What are initial conditions and adjacent possibilities? Initial conditions are just that. They are the conditions before a change takes place—in whatever. After a change, the new conditions become the initial conditions for the next change. As a CAS self-organizes repeatedly, new initial conditions will arise repeatedly. Change can take place only into adjacent possibilities. They are possible only because they are one step away, adjacent to the initial conditions.
Imagine you are standing on a busy street corner with a traffic light, walk-wait signs, numerous pedestrians, and lots of traffic. Those are your initial conditions. That’s where you are. You can obey the traffic signals, walk into the traffic, or walk in another direction. You can bump into or avoid the pedestrians, or dodge the cars and jay-walk. Or you can just stand there. Those are your adjacent possibilities. That’s what you can do given where you are. Whatever you do, that becomes your new set of initial conditions for your next move: a move into a set of adjacent possibilities surrounding your new set of initial conditions..
The possibilities adjacent to the current state further limit the self-organizing options available to any CAS, substantially reducing the operation of randomness.
Another aspect of CASs is of particular interest. There is no going back. A CAS self-organizing into an adjacent possibility then resides in that new possibility with its own initial conditions for the next self-organization: the residual structure of the self-organization constitutes the beginnings for the next one. The adjacent possibilities presented from that vantage point do not include the previous state. If I cross the street, I am no longer on the same corner. If I remain on the corner, the traffic conditions will have changed. My adjacent possibilities will have changed. Been there, done that! The environment has moved on. That street corner has a completely different set of opportunities once a choice is made—or not. Traffic and pedestrians change. The signals change. The interaction of all the participants is at a different point. The opportunities are different. The previous state is not within reach. As a consequence, the sequence of initial conditions and adjacent possibilities traversed by self-organizations may not lead to an optimal organization. Suboptimal self-organizations are more likely than optimal ones. An illustration of this characteristic is the man standing among several hills. He is simply told to walk uphill. What are the chances of his reaching the highest peak among all the hills? Once he starts up a low rising hill, there is no turning back… just walk uphill.
Surprisingly, a hierarchy of simple if-then rules can lead to complex behavior. An example given early in the study of CASs in biology cited the operation of bacteria. If bacteria encounter an increasing flow of glucose, then they move toward it. In a noted software program, ‘Boids’, three simple rules lead to behavior emulating the practice of birds when they flock together. In a hierarchy of decisions, if-then rules can be very powerful. As these systems constantly self-organize in response to environmental disturbances, they are adapting to those disturbances.
This same characteristic is evident in the operation of the human central nervous system. The brain is composed of a multitude of sub-systems operating in almost limitless relationships and hierarchies, each subject to the influence of its own environment—including the body and other parts of the brain.
Of particular note—and of some mystery—is the fact that these systems have an unpredictable property that emerges from the connections and interactions of the elements in the relationship. It is as if the relationship reaches a critical mass, and boom—a phase transition occurs. Something emerges that is qualitatively different from the sum of the parts. This emergent property is not predictable from an examination of the individual elements, nor of the connections among those elements.
Consider the movement of automobiles in an urban area. The elements interacting are the individual automobiles, the rules are manifested in the signage and signaling, and in the actions of the drivers. An examination of the elements alone would not disclose the property that emerges from the system—traffic, sometimes stable sometimes chaotic. It is this property of emergence that is most important to the future of the mind.
Today, many basic elements available for self-organization are created by ourselves—the very systems self-organizing. These elements are as primitive as generation-specific adornment and as sophisticated as online social media and search engines. These elements are systems in themselves, with their own environmental stimulation and self-organization—systems we are creating. While it is unusual for a species to create an environment that then brings about subsequent change in that species, it is not new. Beavers build dams to create lakes for their homes. Corals build reefs for their footing. What is new is that the elements influencing the change are not just material. They are cyberous: machine-oriented, electronic in nature. Underlying them are the multiple dimensions of the Internet and the continuing connectivity of ever more sophisticated mobile devices. We are always in touch. Most importantly, these electronic elements are complementary to the three eons-long paths of adjacent possibilities that I mentioned earlier. These three sets of possibilities are still being explored by the CAS that is made up of us humans: pattern recognition, vision, and post-birth development. These three ‘trajectories’ in evolution need some exploration before examining the future of the mind and the contribution these forces make to that future.
Jack Huber: The Future of the Mind
Maturana and Varela