Understanding System Dynamics

Self-Regulation and The Hidden Power of Feedback Loops

“You cannot stop the waves, but you can learn to surf”
Jon Kabat-Zinn

OLYMPUS DIGITAL CAMERA

At the age of 16 I spent a hormone laden high-school year in La Jolla, California. It turned out to be one of the more dramatic rollercoaster rides in my personal development, in large due to the stark contrast between the old order of the sheltered Hamburg liberalism and the flashy cocktail cherry life of the nouveau riche in Southern California. It was a good year with many strange fruits to taste. Digesting the experience, I spent hours in the Cove, simply gazing at the Ocean with its mighty waves rolling to shore and the surfers on their boards waiting to catch the next big one. I never had the aspiration to become a skilful rider in my limited time there, but the scene certainly left an impression on me – the immediate relationship of surf and surfer has a fascination that still lingers. It represents a strong symbol of the Sense and Respond mode and – within some boundaries – a fine metaphor for process work.

If we want to ride the dynamics of an organisation, it is imperative to get a feel for its underlying currents and patterns of development. As these patterns tend to be complex, at times even chaotic, a purely analytic approach will not get us far here. High focus attention, the sharp observing eye and categorizing mind that is classically associated with intelligence, can easily fail to catch the bigger picture. Rather, we need to be gazing – observing with open awareness, a state that David Gelernter calls low focus consciousness in his work “The Tides of Mind”. It is not just the eye that judges the size and angle of the oncoming wave; it is not just the feet in the water that feel the swell; not just the sense of balance that detects the backdraft of the board; and not just the frontal lobe that counts the rhythm of small and large waves and determines the timing for the next move. It is all of the above, melted in a trance like state of sensing – open awareness. We experience this state very naturally a couple of times a day as we move up and down the high focus-low focus continuum. And we can train the zoom in-zoom out motion between these states like we can train a muscle. Meditation helps. Sitting quietly observant in a turbulent environment helps. Regarding a piece of art or a familiar face with a shifting mode from examining detail to experiencing the whole will train the muscle. 

When I go into an organisation to accompany a process I don’t start with a firework of smart and stirring proposals or the big bang of a process kick off. I start with some very quiet talks and observations. I gaze. I put my feet in the water, sense the drafts from underneath, take in the rhythm of communication and sense the oscillating swing of competing values underneath everyday business. Of course I play my role in setting up the process, but I try not to disturb the organisation in its existing patterns quite so quickly. I don’t believe in an initial pattern breaking intervention, a big shake up to unfreeze the system. I don’t think it is helpful or sustainable for a profound development process. Sometimes this makes people nervous and anxious. They want to be unstuck and get on with it. They want revolution or divine intervention to open a blocked road or a new strong hand to lead the way. That’s not my style. I believe that every pattern and every dynamic holds the key to its own development – feeling the pattern and sensing the nature of this key comes first.

Of course, there is an analytic part in all of this as well. In order to make sense of the patterns we observe, we need to be fluent in the universal language of system dynamics. This part is about understanding the functioning of flows and stocks, feedback loops, buffers and delays. It is universal in the sense that all open systems run on these mechanisms. Using the system dynamics language, we can try to describe the individual make-up of the system we intend to ride. Let’s take a look.

The Basic Language of System Dynamics

One of the great works to advance my understanding of organisational dynamics is Donella Meadows’ “Thinking in Systems”. Meadows was a system dynamics geek at MIT. In her lectures she would draw out wild models with multiple nested feedback loops to explain such diverse things as perfectly tempered bathtubs, jealousy in relationships or the world trade in petrol. A scientist at heart, she made sure the models were lean enough to provide an X-ray view of very complex issues and comprehensive enough to avoid oversimplification. Paired with that scientific rigor was a glowing passion for a sustainable world (actually, as the lead author of the Club of Rome initiated report “Limits of Growth” in 1972 she may be one of the most influential but least known figures in the history of sustainability). Unfortunately, Meadows never saw the publication of “Thinking in Systems” as she died unexpectedly in 2001. Tankfully, Diana Wright of the Sustainability Institute put things together in the years to come.

The foundational idea of system dynamics is beautifully simple and yet may lead us to stunningly complex considerations. Let’s start with the basics: In Meadows’ words, “a system is an interconnected set of elements that are coherently organised in a way that achieves something” (Meadows 2015, 11). This something is what we call the “function” of the system, in case of human systems it may be called “purpose”. Random elements in proximity to each other (e.g. sand scattered on a road or a crowd waiting on a train platform) don’t match this definition – even if some parts may be interconnected, the elements don’t have a common function or purpose. Functions, in Meadows’ terms, are not defined by a stated goal, but by the actual operation of a system. If a company’s mission statement proclaims sustainability but all deals are made without regard to environmental impact, the system’s true function is most likely maximizing profit rather than sustainable business. In the Planetary Model, it would thus be the layer of Organisational Behaviour that gives away the true purpose of an organisation, not its Codex. Nonetheless, it is the Core that holds the purpose, and of course behaviour is attributed as much to the make-up and pull of the inner layers as to environmental factors. But let’s leave this thought aside for a moment and follow Donella Meadows.

System Dynamics assumes that all systems self-regulate to maintain their integrity. They effect both, a healthy target level of key parameters (relative stability or dynamic equilibrium) while continuously adapting to their environment (constant change). It is the interconnections, that play a key role in this process. We are a lot more used to seeing elements rather than interconnections. Interconnections are marked by flows – physical flows, flows of information or communication (in social systems). The mechanisms regulating these flows are called feedback loops. When looking at self-regulation in systems, we can distinguish two kinds of loops: balancing and reinforcing feedbacks.

The balancing loop will decrease a key parameter once it exceeds a certain target level (and vice versa: increase it, when it has fallen below that level). Let’s take the body as an example of a self-regulating system: Sweat is the body’s reaction to heat in order to keep its temperature at a healthy level around 37 degrees Celsius. Shivering in turn is the body’s intervention to regulate upwards if its temperature drops too low. Another classic example is the self-regulating nature of populations in ecosystems: Once there are too many predators in an ecosystem, the prey population will be decreased, in effect decreasing the predator population as it runs out of food. The balancing feedback loop – without any single member of either population consciously acting on it – is keeping the ecosystem in dynamic equilibrium (i.e. the sum of inflows equals the sum of outflows). This is how sustainability is built into the system (until it gets messed with). 

In organisations we find this kind of balancing feedbacks around parameters that need to be maintained within a certain range for strategic or cultural reasons: for example, if the organisational culture is built on harmony, any looming conflict will be subject to a balancing feedback loop that quickly defuses heat and resolves tension. A more complex version of this balancing mechanism comes into play where competing values make up an organisation’s core (take the classic freedom vs. commitment dilemma as an example). In this case most likely, we will find a balancing feedback loop to either side of the equation: if one side grows to become dominant, the other one will rise up and reinstate a “healthy” balance. In many cases this will lead to a pattern of oscillation, a pendulum move between the two value states (obviously: this oscillation is not always the best and most sustainable way forward). Systems can get stuck in the pendulum swing. In that case, the balancing feedback loop puts a break on transformation.  We will come back on how to deal with this.

The other kind of feedback loop is the reinforcing loop. In this mechanism the increase or decrease of a core parameter is progressively enhanced in the same direction and can escalate to effect dramatic change or break down of the system. Classic case: The rich get richer, the poor get poorer. Erosion is another example of a (negatively) reinforcing loop: plants are taken away, soil erodes, even less plants can grow. In organisations we find a number of such reinforcing mechanisms: Most prominently, power has a reinforcing nature: The one in power can access more resources in effect furthering his/her power. This, by the way, is what fuels the “Iron Law of Oligarchy” described by the sociologist Robert Michels. Equally, organisational cultures and frames are built around strong reinforcing feedbacks: Where there is a dominant interpretation, there tends to be a filtered perception, thus stabilizing the dominant mental model. We know what we see – but we only see what we know. A dominant mind-set, just like a strong organisational culture can quickly become a trap.

So how do we work with these loops?

  1. Don’t fight a balancing loop: One of the classic experiences in organisation development is the push back of a system that gets pushed. The stronger you navigate to one side, the more forceful the system resists and veers to the other. The mechanism underneath this puzzling effect is the balancing feedback loop protecting one side of the Core Value Equation. If you are pushing for commitment, the freedom side will push back. If you are pushing towards decentralisation, the centralisation impulse will gain strength. And rather than pushing on with more force, which in turn will produce more counter-pressure, it may be wise to step back and address the underlying equation. Is it still valid? Is it applied correctly here or projected with distortions? And are there ways beyond the either-or continuum? (Syst’s Tetralemma solutions seem a way forward here). As long as you fight a balancing loop on the surface, you are fighting yourself.
  2. Use the swing of the pendulum: Earlier I talked about the oscillating nature of development swinging back and forth between competing values. Oscillation can be annoying. As soon as you think you are on track, things suddenly swing backwards. One answer I find to this in my process work is the hibernation strategy. If I know the backswing will come and go, I can simply sit it out. Every time the pendulum swings in the direction I want to move, I use the momentum of that swing. While things swing the other way, I simply take a breath and step back to observe. I know, the pendulum will come back around. This is different from not taking things seriously. It is simply a way to work with them. 
  3. Keep an eye on the reinforcing feedback loops: Reinforcing loops can be your friend in effecting change and getting things unstuck. The Arab spring could not have happened without the reinforcing feedback built in the architecture of social media. Online, attention breeds attention in an exponential way. From a revolution standpoint, this is great. From a systems operations standpoint this is deadly. Reinforcing loops may destroy a system, if an accelerating dynamic gets out of hand. If your agenda is reform and not revolution, you may want to put some padding on the reinforcing loops. 
  4. Factor in delays: A development process takes time. Between deciding on a strategic move and its actual implementation, several months may pass, in some large and bureaucratic structures even years. This delay time impacts the self-regulation of the system and may lead to further oscillation. Here is a progression I see a lot: A team recognizes an internal collaboration issue and jointly agrees to tackle it. An agreement on a new way of doing things is made with everyone on board. Of course, unlearning old behaviour takes time and requires practice. The new conduct does not magically happen overnight – things move slowly. Some people get impatient, they call for even stricter rules and sanctions as they feel the first agreement was too lax. So new, tighter rules are made. By this time, all movement towards the initial agreement is depreciated. Those that were sceptic but willing to go along, are frustrated. Commitment is revoked, the agreement collapses and things are worse than before. From here, the story starts over again. This oscillation, caused by delays and subsequent overshooting, is quite similar to the classic shower experience (the hot water faucet is turned on but doesn’t deliver hot water immediately, leading to a further opening, which, as the hot water arrives, turns out to be an overshoot, i.e. the water is too hot, which is then compensated by turning the hot water off, thus effecting an undershoot, i.e. the water gets too cold etc.). Sometimes all it takes is a patient, steady hand.
  5. Read the true purpose: As mentioned above, a system’s function is revealed not by what is written on the package but by its actual behaviour. Most of the puzzling patterns in organisations have to do with a discrepancy of formal and informal system layers. Discrepancy analysis is a powerful tool to reveal these rifts: Balancing loops that make no sense in the light of the stated values; reinforcing loops that counter the official compass needle; once we decode the true purpose of a system, these weird behavioural patterns suddenly make sense and in turn the codified formal layers seem off – we realise, it is the Codex that needs attention. In my experience these realisations are break through moments in a process.
  6. Strengthen the system’s ability to self-regulate: As consultants and facilitators (not to mention leaders) we often take on the responsibility to regulate a system, especially, when its ability to self-regulate seems to be impaired. This can become a trap. The more we substitute internal with external regulation, the less likely the system will build up the ability to self-regulate. 

Schreibe einen Kommentar

Deine E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert.