Limitations of Chaos Theory
Sometimes the first step in learning is unlearning previously learned ideas that are incorrrect. With regard to models of change, there is a prevalent social myth, started in the 1980s, that an area of study called chaos theory tells us the future of complex systems is always chaotic and unknown. This is simply untrue and we must better understand the limitations of chaos theory to place it in its proper place. Chaos describes some of the features of evolutionary systems, but very little aspects of developmental systems, whether we are talking about organisms or their environment.
In the 1960s some physicists studying the weather found a set of nonlinear differential equations (equations used to describe change in complex interacting systems, like weather, turbulence, and predator-prey interactions) that were “chaotic”, which means they produced “regularly irregular” outputs, which numerically looked a lot like the weather, with “bounded randomness” (stochasticity, for the mathematically inclined). These processes were also “sensitive to intial conditions,” meaning that small changes in the initial values of the system could produce large and unpredictable outputs as the system evolved.
One of the researchers, Edward Lorenz, called this last property the “butterfly effect.” He was noting that due to the sensitivity of his nonlinear equations to initial weather conditions (temperature, pressure, etc.), one could imagine a butterfly flapping its wings in Texas causing a storm a few weeks later in China. He never meant this literally, as there is usually a low limit to the variability generated in any real world chaotic system. Small changes almost never “blow up” as dramatically as the butterfly effect metaphor suggests. Instead, both chaotic and complex systems always have an opposing “envelope effect”, an interaction with their environment that dampens the degree of change possible due to changes in initial conditions. The complex system behaves inside a probability envelope that can be discovered by experience. Consider the weather. With good sampling, we can always find and describe the envelope of normal temperature, pressure, precipitation, and other variables in any area on Earth, and then predict that the weather will stay within that envelope at that location, and we will be right the vast majority of time, even though we can’t predict the weather there on any particular day without doing a lot more forecasting work. Even chaotic systems, then, are highly probabilistically predictable with respect to their envelope and averages.
Scientists and evolutionary theorists often use the null hypothesis of there being no relationship between observed phenomena, and the implicit assumption of randomness in system interactions as starting points in developing our statistical models and inferences. But randomness and chaos are a good starting models only for many evolutionary processes.
If Earth-like civilizations and technologies develop, not just evolve within our universe, all of their developmental features must be as isomorphic (the same in various systems) and predictable as, for example the features of galactic, stellar, and biological development are assumed to be today, even though humanity’s current math and physics are not yet good enough to model many of these predictable processes.
If you’ve seen one acorn develop into an acorn tree, you don’t need math and physics to know what another acorn will seek to do. You just need to have seen it once, and know the process is developmental.
We must also recognize that chaos is not complexity. Evo Devo Universe scholar John McCrone explains the distinction well in his book on consciousness, Going Inside, 2001:
The distinction between chaos and complexity can seem hazy at times, but essentially, chaos theory describes how simple, repetitive interaction, left alone to rub along, can produce something of rich structure, including shorelines, puddles, and weather patterns. It is about the feedback driven generation of complication. Genuine complexity is something else, however. The really interesting things in life – systems like cells, economies, ecologies, and of course, human minds, have extra properties such as an ability to adapt, to self-organize, to maintain some sort of coherence or internal integrity. These systems are not slaves to their [nonlinear] maths, passively following a trajectory through phase space. Instead, they have developed some sort of memory, [chemical, genetic, memetic, or temetic] mechanisms which allow them to fine-tune the very feedback processes which drive them. They can change the attractor landscapes in which they dwell, and so reshape their own futures. A complex system is one that has harnessed chaos, rather than one that is merely produced by it.
Many nonlinear systems are not sensitive to initial conditions, but the opposite. For example, developmental biologists use nonlinear differential equations to describe how an embryo turns into a developed organism. In those equations, the chaos of molecular movement is used to create highly specific, and experience-predictable future forms and functions in the organism. As we just said, once you’ve seen one acorn develop into an oak tree, you know what future acorn seeds will seek to do, even if you don’t know the differential equations involved. You just need to have seen it once, and know the process is developmental.
What’s more, biological development cleverly uses chaos to create future order and predictability. Molecular biologists use nonlinear equations to describe how chaotically interacting proteins impose order on the cell. Social, economic, and technological development also use chaos to create various types of predictable future order, as well will shortly see.
Our world is full of all kinds of self-stabilizing processes that work well with chaos, operating on all kinds of timescales, from the universal to the planetary to the social to the personal. For a planetary example, consider plate tectonics, the way the mantle of Earth works with its crust to create new crust at the bottom of the oceans, which moves the plates on the Earth, which get recycled down into the mantle when they reach Earth’s faultlines, causing earthquakes. That is a 100 million year cycle, and it is one of the ways Earth’s geology systems have kept CO2 levels stable on the Earth over billions of years. We only discovered it a few generations ago.
For a social example, consider an ant colony. If you open up a colony and scoop out a number of ants, a few days later the colony will have restored the proper ratio of ant subspecies, eg, soldiers, foragers, and colony-workers. This nonlinear, complex process occurs via countless small information and energy flows across to the whole collective of ants, without any central direction. This nonlinear process maintains a dynamic order, funneling the system back to the developmental average, even in a chaotic world. How many human social processes work in similar fashion?
Ian Stewart and Jack Cohen offer a nice primer on chaos in The Collapse of Chaos (1994). They make clear that all of the world’s complex systems, including social systems, are partially predictable, with predictable and unpredictable regimes. Evo devo theory calls those regimes evolution and development. They also note that intelligence is a process of simplification of the complex, via modelmaking and prediction. Such modelmaking can be divergent or convergent. Intelligence, in other words, can be either evolutionary or developmental. When you have lots of models and predictions diverging, that is an evolutionary process. When your models and predictions are converging, that is a developmental process. The act of modelmaking and prediction itself can be either evolutionary or developmental, depending on whether it is socially diverging, in search of truth, or converging on something that everyone agrees is truth. The act of prediction then is not developmental, but evo devo, a consequence of intelligence. Predictability is what is developmental. Almost all mental predictions are evolutionary. They are good for a short time, to a degree, in their contingent environments. They are diverse and socially they don’t agree, but compete and cooperate. Predictions with high predictability rise above local contingency and eventually spread to all minds. They are developmental.
Today’s scientists seek to advance complexity science, and most see chaos theory as a small subset of that science. We will eventually be learn to model many future predictable processes better with nonlinear equations and complexity science, but today most can only be described predictively with crude math and simple approximations and regularities. It is still rarely clear which nonlinear equations are most useful to describe which systems, and when, and whether those equations create bounded randomness or self-organization and other forms of predictable order.
Mostly, we make hypotheses about such systems today using a branch of philosophy known as systems theory. Systems theory attempts to understand commonalities, differences, and relationships between all our universal systems, regardless of type. It isn’t science, it is natural philosophy. But it is natural philosophy that will one day be proven either useful or incorrect, by future science. In the meantime, we look for hidden order and predictability, as best we can.