Systems Science 9 — Major Classes of Systems
I find it useful to consider systems in three major classes or types. There is a fourth type but I think of it more as a kind of an abstraction than a real system, at least at this stage of the universe's life. I'll explain that in a bit.
The three non-abstract classes are:
Below I will explain each of these, but a rough idea here will lay some groundwork. First, the "abstract" type of system can be called the Chaos-System, or Unorganized-System. Think of a completely insulated flask containing gas molecules, say hydrogen as H2. Allow the system to age for a really long time and according to thermodynamic theory inside the flask the temperature is the same everywhere and things are just bouncing around. Very boring! The reason this is abstract (even though you can imagine such a flask existing in reality) is that any attempt to peek inside to see if things are really chaotic would introduce an observer which would be at a different temperature that would cause the system to undergo changes (even a really minute measuring device would still cause a change!) Thus we can only use such systems as an abstract 'ground state' in a mathematical sense. Still it has its practical uses in being able to characterize real systems in relation to their abstract (theoretical) ground state. While an important issue, for purposes of this short version, I will leave it at that. Plenty of books on thermodynamics, statistical mechanics, thermochemistry, etc. lay it out in excruciating detail, so interested readers should refer to them.
My classification scheme (of what we should think of as 'interesting' systems) is based on several parameters or dimensions derived from the previous installment on complexity, etc. Leaving aside thermodynamic order, which is highest when the temperature is lowest, I consider four basic dimensions that help to classify systems. These are actually a pair of two dimensional spaces: competitive vs. cooperative component interactions and complexity vs. organization. I should hasten to point out that this is not the only two spaces one could conjure to elucidate system types. Indeed, it would be useful to consider multi-dimensional (hyper-)spaces to see if the analysis bears consistency when many factors are taken into account. But these two spaces should prove instructive for my present purposes. In the figures below you will see ovals representing the three types of systems. These are conceptual aggregations. The idea is if you were to take a survey of real systems according to some measures on these dimensions they would be placed within (mostly fuzzy) boundaries within the larger space. This is, after all, how we define classifications. Seemingly different systems might share similar measures on these dimensions and hence end up clustered within the 'glob' represented. I must emphasize that this is a hypothetical framework. As far as I know no one has done such an analysis rigorously. What I offer here is based on a very anecdotal survey of my own. But I think you will see the general merit of classifying systems in order to bring some organization to systems science, just as classifying plants and animals (systemics) helps organize our understanding of phylogenetic, behavioral, and physiological relationships between different flora and fauna.
Figure 1 shows a two dimensional space based on complexity (vertical axis) and organization (horizontal axis). The origin at lower left is labelled low for both dimensions, while the top and far right represent high levels of these dimensions. The three types of systems are shown as rough areas occupying somewhat overlapping regions of the space. I've included the Proto-universe (after the Big Bang), galaxies, and solar systems as kinds of Mech-Systems, which I will explain below. As represented here, Mech-Systems are toward the low end in complexity, having few components and few interactions. They can, however, be very high on organization. As we will see, human engineered machines fit this category nicely. Bio-Systems occupy the region in the middle, having medium to high organization and complexity both. Eco-Systems are generally very high in complexity, but not necessarily very ordered (we now know that even so-called 'climax' ecosystems are not long-term steady-state systems).
Figure 1. Complexity versus Organization as a way to classify system types. The overlapping regions implies that there are no 'clean' boundaries. Eco-Systems contain Bio-Systems, for example, a bee colony has aspects of both ecosystem and macro-bio-system.
The second two dimensional space relates the dominance in the tradeoff between competitive processes (component interactions) and cooperation. Bio-Systems rely more on cooperation than competition, though the latter is essential for certain biological functions. Mech-Systems are characterized by higher degrees of cooperation among the parts. Human engineered machines are the quintessence of such systems where absolute cooperation between the parts are essential. But we can also consider non-engineered, but mechanical systems, such as the Earth's crust or a mountain stream as a Mech-System with some amount of competition going on between the parts.
Eco-Systems tend to have more competitive interactions between 'components'. At the low end of an Eco-System we find populations of the same species, where individuals are in competition with one another for food and mates, but still manage to cooperate for the good of the species. At the high competition end we have a typical rocky intertidal zone where numerous species and individuals within a species are continually competing for real estate (which usually translates into food and mates!). A critical question that we need to address is: where are human societies placed in this figure? As I will argue, there are societies (examples) that run the range from relatively high internal competition to relatively high cooperation. But in general humans seem to center somewhere in the overlapping region of Bio- and Eco-Systems. I will address human society as a special case of Eco-System that may need to tend toward Bio-System if we are to survive in the future.
Figure 2. Cooperation versus Competition as a way to classify system types. Some kinds of systems based on this scheme are infeasible, they cannot have both high competition and high cooperation simultaneously. On the other hand, minimal interactions with some low cooperative and some low competitive suggests simply a disorganized, possibly chaotic system, as our abstract system mentioned in the text above.
This class of systems is characterized by the preponderance of cooperative interactions among the component parts. Additionally most components interact with very few other components directly. The most obvious members of this class are man-built machines. Machines are designed to fulfill very specific functions and their working components are designed to work together, cooperatively, to produce that function. However, if we are not too strict on the absence of competitive relations some natural systems can be fit into this class, especially atomic and molecular-scale chemical systems where quantum effects are not too noticeable. Even though DNA molecules are associated with living systems, the molecule itself is not living. It is, however, fairly stable (thermally) and rigidly structured in the right temperature and pH aqueous solution. It can be manipulated chemically in very predictable ways so that pico-machines can actually be built using molecules of DNA.
Machines are generally characterized by their degree of deterministic behavior. They can be counted on to do what they are supposed to do. The better the design and fit of all the parts, the more durable the materials used, the more reliable the machine is for a longer duration. But entropic processes still get their due. Parts wear out eventually and unevenly and can alter the performance of the machine in time. Energy is required to replace or repair the workings.
Machines are a good example of the difference between organization (in the sense of complexity) and thermodynamic order (third law of thermodynamics). Because machines are more likely to do what they are supposed to do, they can be said to have a high level of order, as if their parts are 'frozen' in their relationships to one another. In this regard they are like a crystal in which the bonds between components are 'frozen' and the number of states of the system are restricted.
Figure 3. A Mech-System is characterized by components with few interrelations with other components and those relations are "frozen" such that the machine produces predictable outputs. Here the whole machine is represented by a rounded-corner rectangle. Each component is a circle (thin line circumference). Lines between components (interactions/interrelations) are few and the overall structure is rigid, like a crystal.
As mentioned above we should not restrict the notion of Mech-Systems just to machines. Many natural systems have machine-like properties in that they have fewer interactions between components and those interactions are stable over time. In natural systems, however, many times these interactions are stochastic, that is they are probabilistic over time. One example is a waterfall. Over time the water pouring over rocks generally follows a pattern, even if the details are not specific. Waterfalls are chaotic systems with no drops of water following preordained pathways. The interactions between water flow and rock is that the latter gets gradually eroded so as to change the general pattern of the flow. So such Mech-Systems do evolve over time (actually not unlike the wearing out of mechanical parts in a human-built machine). Nevertheless, the system is fairly simple and organized. One might also argue that there is a balance between competition (erosion) and cooperation (channeling the general water flow).
Bio-Systems can be thought of as more complicated networks of quasi Mech-Systems (like the DNA example above) in which there is greater flexibility in the network of connections; they can vary over time. Clearly individual living organisms are the paradigm example of Bio-Systems, hence the name. As we will see, Bio-Systems are somewhere between machines and ecologies. In the latter case, relationships are constantly in flux and systems can only be characterized by changing network structures. In Bio-Systems there is less rigidity in interrelationships than in Mech-Systems. There are generally many more connections between components providing multiple redundant pathways for interactions. A key characteristic of Bio-Systems is the trade-off between cooperation of components (as in the machine) and competition between them under critical conditions.
Individuals, and in the case of eusocial animals, colonies, are composed of components that generally cooperate, or at least the dominant mode of operation is cooperation. But this is not a strict rule. During development there is often competition between cells as they differentiate and vie for location. In the cortical tissues of more advanced vertebrate brains, even in adulthood, there can be an on-going competition for processing power and physical space between modular functions. This is demonstrated most readily in the motor cortex where plasticity of the neural units hinges on use-it-or-lose-it rules and in the case of loss of a digit or limb results in the areas once dedicated to controlling that appendage's movements will be colonized by nearby areas controlling similar appendages.
Competition comes into play especially when the system is under stress from resource constraints or disease. Some non-critical subsystems are shut down or at least down modulated so as to allow resources to be routed to critical functions. This is true at the single cell level and at the whole individual level in multicellular organisms. In this case a logistical control subsystem is responsible for mediating the competition (regulation) to ensure that the critical function wins.
There are several other examples of competitive processes operating under normal circumstances within an individual's body but where logistic regulation prevents destructive competition. One example is opponent paired muscle sets that move joints or maintain posture. Muscles like the biceps and triceps operate in a dynamic tension to control the motion of the forearm. Neural controls, which include cross inhibition, make sure that these two muscle sets don't over-flex and produce damaging tensions around the joint or tear each other apart. Without such logistic regulation, controlled fine motion would not be possible. Another example is related to the neural circuit just mentioned. Throughout the nervous system (though generally more prevalent in brain stems) there are examples of circuits called 'central pattern generators'. There are many different versions and these circuits are responsible for rhythmic motions like walking/running gait control. They work by sets of neurons, generally through a sequence of other neurons, mutually inhibiting or exciting one another. Their outputs tend to resemble oscillatory excitations that are sent to specific effectors, like the logistic regulators for opponent paired muscles. They are especially useful for undulatory motion such as peristalsis in the digestive tract. In these circuits the neurons appear to be competing with one another for excitation (say through cross inhibition), but the phase differences created by the effective distance between cells and their terminating synapses allow the circuits to produce a cooperative output. [see also: Mobus, G.E. and Paul S. Fisher, "Foraging Search at the Edge of Chaos", in D.S. Levine, V. R. Brown and V. T. Shirey (Eds.), Oscillations in Neural Systems, Lawrence Erlbaum Associates, Publishers, Mahwah, New Jersey.]
Evolution has worked to produce individuals in which cells and tissues generally cooperate with one another in a well choreographed dance. Complexity is quite high compared with machines since many component parts interact in multiple ways with many other parts, the interactions are most often context dependent. The range of possible states is huge by comparison with machines, but it is not, technically, infinite. States are defined by molecular-level interactions within cells, so at a certain level of observation the system appears to be continuous rather than discrete. But if one examines the micro-states of cells, one discovers more distinct transitions as biochemical processes proceed. Still, for generally practical purposes the systems can be treated as continuous without (usually) loss of information. However, caution is advised in treating this too literally. A single micro-mutation in a DNA molecule in a stem cell or a germ cell can lead to considerable shifts in the future states of individuals.
Competition between certain components in a Bio-System are conserved through evolution because competition is useful at critical times in the life of individuals. Cooperative processes, via logistic regulation, dominate so that the individual acts much like a machine most of the time, at least during its adult, sexual stage. The parts work in concert until age or disease cause them to falter or fail. But, at the same time, their normal function is stochastic in nature, meaning that the variability of behavior of parts is higher than in a strict machine. Such variability stems from retention of the plasticity of sub-cellular and tissue-level components. This plasticity is what allows Bio-Systems to adapt to non-standard environmental conditions when necessary (and in brain tissue it is the basis for learning). The downside of maintaining such plasticity is that sometimes parts don't behave exactly as they should and because they are interconnected to so many other parts this can cause intrinsic (that is, not from the outside) malfunction. The parts may get 'out of synch' with one another leading to oscillatory fluctuations. Generally this is temporary and correctable. But with more complex Bio-Systems, more parts means more potential for things going wrong. Many genetic-based diseases fall in this category.
Bio-Systems are, thus, highly organized and the parts are highly interconnected making life a very complex kind of system. At the same time they are intermediate in terms of thermodynamic order between machines and Eco-Systems. Evolution of Bio-Systems has been described as keeping species on the boundary between order and chaos. They are just organized enough, and sufficiently controlled so that they are successful in meeting the challenges of living and procreating. But they are flexible enough and subject to just enough intrinsic fluctuation that they are not strict automata. They can deviate from the norm as a form of 'exploration' of the environment and of useful behavior space.
Figure 4. A Bio-System is characterized by more interconnectivity (greater organization) between components. Here components are represented as Mech-Systems but real Bio-Systems are more often hierarchies of Bio-Systems within Bio-Systems. The whole Bio-System is here represented by a thick line circle. More cogent characteristics, not shown in this figure, are the relationships in time between competitive and cooperative interactions.
The third class of systems is characterized by a high degree of complexity (generally) and a low degree of thermodynamic order. These systems are aggregates of multiple Bio-Systems from multiple 'species', often multiple Mech-Systems, and non-living components, having many more interactions, with much higher variation in both kinds of interactions and degrees of coupling strength in those interactions. More importantly there is often a much higher degree of competitive types of interactions, even between conspecifics, relative to cooperative ones.
Figure 5. An Eco-System is characterized by somewhat less, but highly dynamic and variable interconnectivity between components. Components are of every other type of system, Mech- and Bio-, and can also include nested Eco-Systems.
Here I look at two basic kinds of Eco-Systems, the standard idea of an ecosystem as studied in ecology, and a Social System, an ecology of social organisms. Of course the one we tend to be most interested in is the human social system.
Ecology is the study of natural environments and their component biological subsystems. In fact many ecologists have eschewed the systems concept in describing their units of interest because the term too often is associated with Mech-Systems (machines) and ecological systems are anything but like a Mech-System. However, as this scheme of classification implies, ecological systems include Mech-Systems, the non-living substrate and general physical environment, as well as Bio-Systems from viruses to megafauna and plants.
Ecologies are often studied from the standpoint of systemics (what species are present), dynamics (how they interact with one another — behavior), population dynamics (how species grow and contract), evolution (how some species diminish over time and other species invade, changing the dynamics), and, very importantly, the thermodynamics or energy flows from inputs (via primary producers) through trophic webs (primary consumers, carnivores, etc.) to microbial recyclers.
I don't propose to go into any greater detail in this section since so much has been written about ecology and there are still many fine systems theoretical perspectives in play in the field. If this ever becomes a book on systems science I would probably provide some of these examples plus relate various principles from prior installments to specific ecological principles to show how systems science (and not just Mech-Systems science) should still be the overarching framework for this field. This is especially important because of the relationship between the non-human ecology and human social systems. Though many wise people throughout recent history have pointed out that the human ecology has to be regarded as a subset of the larger ecology (of the Ecos) it is only now that we are about to disrupt that ecology by not recognizing this fact that truth dawns.
My argument is simple. Human societies and cultures, including all artifacts, behavioral conventions, institutions, and especially agriculture, are just as natural as any wild forest or coral reef. The only difference is that humans are clever enough to overcome any of the constraints of the larger Ecos that might have kept us from overrunning the world. In the process of expanding our territory out of Africa we have exploited other ecologies, other resources, and often destroyed them in the process. Humans have now succeeded in appropriating over 40% of the primary productivity of the planet for their own trophic uses, and there are no natural predators waiting to check our population.
Social Eco-Systems depend on the strong cognitive and emotional interactions between individuals making up the social group. Ultimately that is what binds us. As in a general Eco-System these interactions may be restricted to a few other individuals (at least at a time), but through chains of interactions with other groups, and elements of our culture (e.g. TV) we have evolved an extremely complex and generally organized ecology.
Eco-Systems are, however characterized as much by competitive processes as cooperative ones. In the case of biological ecologies, the shifting dynamics of a region allow for an evolution of natural feedback loops that constrain the components from 'taking over'. Human societies have also evolved to introduce some elements of regulatory constraints to help coordinate activities and keep competition from causing too much damage. But this evolution is really just begun. Governance and economic market activities are still relatively young inventions and even though we would like to believe we have developed the perfected form of both, in the form of democratic capitalism, there are still many flaws and unrestrained aspects as we have recently witnessed in the economic crises we find ourselves in now.
Humans continue to evolve, and assuming that there are human survivors from any potential cataclysm due to climate change and net energy decline I fully expect that it will be in the area of strategic thinking capacity (an actual development in the brain). Along with that evolution I expect that the social structures involving logistical and tactical coordination (in other words governance and markets) will also develop. For one thing we have learned that political process breaks down into ideology and dishonesty (with money involved). Politics has turned out to be a horrible way to form and run a government. And markets without appropriate regulatory oversight (along with foolish market participants) are disasters in the making.
The human social Eco-System is actually evolving toward being more of a Bio-System in which coordination resolves the issues of competition that cause Eco-Systems to be so dynamic and in a state of constant flux. Bio-Systems are stable over time whereas Eco-Systems can go through drastic changes from time to time. As an individual component in either of these two kinds of systems, which would you rather be? A Bio-System provides greater predictability, stability, assurance, etc. An Eco-System is a grab bag. As human evolve to have greater moral fortitude, altruism, empathy, and strategic perspective, they will work to evolve social systems that function more like Bio-Systems than Eco-Systems. But that is a long way off. The current species of human has only just awoken to the desire for Bio-System status. It will take a long time for this species to give way to the next, more sapient species that can complete that process.
* I use this hyphenated form in Mech-, Bio-, and Eco- systems to denote a generalized type of system. In fact other non-living systems might fit into the Bio-System class, and social systems can be analyzed as a form of ecosystem, hence the Eco-System class.