Using Classical Chinese Concepts of Pattern Identification and Action to Study and Manage Complex Adaptive Systems?

In this post, I’d like to explore the Chinese concepts of Yin-Yang and correlative cosmology, and attempt to apply it to the context of the study of complex adaptive systems.

[Here’s some background for those who aren’t familiar with complex adaptive systems: Complex adaptive systems are one of two broad categories of complex systems, where the constituents (commonly referred to as “agents”) are constantly adapting in response to their interactions with other agents. As a result of this, not only do the agents themselves change over time, but the very process by which they interact with each other changes as well. Sometimes, we may even go as far as to include the environment, the context in which the agents are situated in our consideration: that too can be subject to change over time. What sets complex systems (like complex adaptive systems) apart from merely complicated (or even just messy) systems, is the fact that the aggregation of the system as a whole is greater than the sum of its mere parts. In other words, the product of the agents in their context/environment, and in relation with other agents results in emergent phenomena that you would not have discovered/seen if you were to study the respective agents in isolation from each other (and attempt to construct an understanding of the system based on such isolated studies).

As such, the study of such complex adaptive system is a new field in the sciences, and it demands new ways of studying the systems. Because the agents and their interactions are constantly evolving, it is not possible to reduce the interactions to a kind of universal principle.

The modern scientific method that we have now works best at explaining linear causality, under certain fixed (or perfect/ideal) situations, isolated from its context. This same method, however, is somewhat deficient when dealing with complex systems. It does not function optimally when trying to explain non-linear causality (many-causes to many-effects) where agents are not isolated but remain situated in their context.

People who have done research on complex systems have found that Chinese philosophy shares many parallels with how researchers would like to approach the study of complexity. This is where, I’d like to speculate on what I know, and hopefully contribute to the study of complexity.]

One of the common characteristics of complex adaptive systems is the regularity of patterns. Though agents and their processes of interaction evolve over time, there are still, nonetheless, regular patterns that emerge in their interactions with one another. I think this observation is a very promising lead in trying to develop more sophisticated conceptual tools for the study of complexity.

As the saying goes, “A picture is worth a thousand words.” Patterns are like pictures in the sense that they convey a lot more than can be described merely by words. And like pictures, patterns convey a host of information on multiple levels in a non-linear way. It is a kind of short-hand that we can use to denote a complex/complicated phenomena. Here is an example to demonstrate this: a counsellor is able to identify whether a child suffered from domestic abuse at home based on certain behavioural (and even physical) patterns. The identification is not based merely on one action, isolated from its context, but sets of actions in relation to their context, paying attention to the conditions that give rise to certain sets of actions. It is through these sets of actions (in their context) that one is able to identify it as a pattern indicating domestic abuse.

Patterns are useful, because we can speak about them in a linear way, but actually refer to non-linear causality involving many elements/agents. For example, if we say, “carry out Pattern X to achieve Pattern Y”, we are actually saying: do all these things that will bring about Pattern X, and in doing so, we can induce Pattern Y to emerge as a result. So, while we speak as if it is a linear causal relation (e.g. Pattern X causes Pattern Y), we mean something more than just this linear relationship.

How does Yin-Yang and Chinese correlative cosmology come into the picture? Well, Yin-Yang and correlative cosmology seem to hold the potential for us to further develop this kind of non-linear thinking involving patterns. Before I show how this might be done, let me explain these concepts and their relation to patterns and complexity.

It’ll help if we stop thinking about things as things, as nouns, as objects. Because when we do so, we picture these things as if they are static and unchanging. Whereas in Chinese thinking, everything undergoes cycles of change. It thus helps to think of everything as processes: nouns as verbs, objects as on-going processes of change.

Yin and Yang are essentially labels that we use to describe processes in relation to each other. Yang refers to active processes, while Yin refers to passive processes. Within each process (be it Yin or Yang), we can further sub-divide the component processes (within the parent process) into Yin or Yang. And we can keep doing this sub-division of Yin and Yang ad infinitum, as far as it helps us to serve our purposes.

Yang, the active process; Yin the passive process. (Image source: KAdDigArt, )
Yang, the active process; Yin the passive process. (Image source: KAdDigArt, )

Now, Yin Yang theory states that processes undergo cycles of change: from birth, to climax, and then it reverses into decline. Once it reaches the bottom, it reverses again, and the cycle repeats itself. Because of this pattern of change, processes will not remain active or passive forever. Yang processes will eventually become Yin processes, and vice versa.

The ultimate good, so to speak, is to ensure the continuity of these cycles, the continuity of such processes. (Conversely, what is to be avoided, is excess into such extremes, that reversal is not possible. In which case, not only is this cycle doomed to fail, but all related processes will be affected in very dire ways.) For this reason, to sustain processes, Yang processes must be complemented by Yin processes (and vice versa) at every level, be it on the micro, meso or macro level. In doing so, cycles are preserved.

This is potentially useful for the study of complex systems, and to manage/manipulate such systems in constructive ways. If we can identify cyclical patterns (similar to that of birth, climax, and decline) in complex adapative systems, then Yin and Yang will be very useful as a kind of short-hand to identify and understand the state of processes within a complex system, and how to rectify it if problems ever occur. For example, if we identify that a process in a complex system shows patterns that it has gone too far in its Yang activity, we can then try to implement Yin patterns to complement and harmonise the excess and preserve the system’s integrity (i.e. preserving the cycles within the system that keeps it going).

Of course, right now, this picture seems too simplistic to be of any use. So I shall now bring up correlative cosmology. In Chinese correlative cosmology, we have the concept of the Five Phases (Water, Fire, Earth, Metal, Soil). These are merely names of PROCESSES (and not fixed states of being). In Chinese thought, these Five Phases are inter-related. A change in one phase will subsequently lead to changes in all the other phases across time.

Illustration of the relation between the Five Phases. (Image source: Wikipedia, )
Illustration of the relation between the Five Phases. (Image source: Wikipedia, )

Classical Chinese thinkers have tried to fit almost everything into these five phases: food, smells, sounds, colour, bodily fluids, bodily organs/processes, emotional states, mental states, socio-political actions, and even natural phenomena on the ground as well as in the heavens. There is this idea that things of the same phase will resonate with one another.

Unlike the Western way of thinking, classical Chinese thinking did not have the same kind of causal thinking that we have. Rather, if they said that A causes B, it’s not that A had a direct linear cause with B. Rather, A acts in a certain way that causes B to resonate and act, much like performers in a dance. When Dancer A performs, the series of patterns that is acted out resonates with Dancer B to perform his/her own series of patterns. To use classical Chinese terms, Heaven, Mankind, and Earth are involved in a ritual with each other. Everyone has their own ritual role to play. And when each performs their own ritual role well, then the patterns that one performs cues and resonates well with the other to perform just as well. If one were to slack and not perform the patterns well, then the other would be confused and may not know how to respond properly. Because the patterns of one were off, the patterns of the other would would be out of sync.

But what can we gain from this? Rather than to dismiss all these asmere superstition, I think there’s something valuable in the correlative cosmology that we can use and adapt into complexity science: it is the idea that things on every level are inter-related with things on other levels. Just as how there are different levels of emergent phenomena arising from a combination of factors at each level (be it on the micro, meso, or even the macro level), correlative-thinking is a kind of short-hand that we can use to describe the relation of one set of patterns with another set. While we may not yet know how to explain why the aggregate of several elements can produce a phenomenon completely unlike the constituent elements that generated it, we can nonetheless establish relations of between different levels of elements/agents and their respective phenomena, based on the idea of resonance from one level to another, or from one set of patterns to another set of patterns. Perhaps if we begin to adopt the thinking that “one set of patterns resonated agent(s) to generate another set of patterns,” we might be able to explore complex systems more deeply, and avoid the trap of wanting to say that A caused B (in a very linear, and reductionist manner).

Moreover, though we do not need to use the framework of the Five Phases exactly as they are, we can still adopt some of its ideas. Five Phases, after all, are descriptions given to five different types of processes (each of which can be divided (or even sub-divided) into Yin or Yang depending on its relation with the other processes). Nonetheless, we can, in our study of complex adaptive systems, identify a sets of regular patterns that are common to several layers of phenomena (e.g. micro, meso, macro), and then group each set of commonalities into the same process-category.

Perhaps, if we were to apply the concepts of the cycles (birth, climax, decline), cycle-preservation, Yin and Yang, resonance, and these inter-relation of different processes – all of which rely upon pattern identification, we might have a very versatile toolkit and possible a language for speaking about patterns unfolding within processes, and be able to discuss non-linear causality and the co-evolution of its agents/processes in their contexts, in ways that preserve the complexity of the systems we speak of.