System Dynamics Applications / 1 / 1.3

1.3 THINKING ABOUT AND MODELLING NATURALLY IN THE TIME DOMAIN

From an early age we grow to appreciate that things take time. Although we may not have conscious memory of it, we learnt that our mothers did not feed us the instant we felt the first pangs of hunger and began to cry. We may be frustrated by it, but we know that some things take time. The passing of time is an inescapable fact of life. This can work both for us and against us. We know from experience that one day follows the next and day follows night. As we grow, we appreciate that it takes a long time to save money and a short time to spend it. A seed planted in early spring will germinate in a matter of days and grow in a few short weeks. We grow through childhood to become taller by indiscernible amounts each day.

Life’s experiences tell us that in so many instances one thing must happen, or must be done, before some desired event can occur. When one thing happens shortly after another, we associate those things. When this is so, we can expect that in time one thing will lead to another. Often our appreciation of cause-and-effect only develops when we can ‘see’ events play out over time. Conversely, when one thing happens but it is not clear what the precursor event was (or events were) we do not associate seemingly unrelated events as cause-and-effect.

However, we might appreciate events as being linked as cause-and-effect if we deliberately create, or have demonstrated to us, a dynamic model that shows how events are linked in causal terms. When we are able to do this, we face the highly desirable and inevitable consequence of an unsettling of our mental models of cause-and-effect. When this happens we enter into the regime of double-loop learning. Double-loop learning forces the creation of new dynamic hypotheses: we re-frame our understanding of the behaviour of the world around us through a better appreciation of cause-and-effect and its implications.

Dynamic hypothesis (Sterman, 2000: 86) is a term used to describe current theories of observed problematic behaviour, explaining dynamics as endogenous (internally generated) consequences of systemic feedback structure. Sometimes, and hopefully with increasing frequency, we appreciate that the world is complex but through modelling and simulation we are able to unravel some of that complexity by formulating and testing dynamic hypotheses about cause-and-effect relationships.

The analysis of physical problems, those which abide by physical laws such as the Laws of Thermodynamics or Newton’s Laws of Motion, is enabled by the precise nature of physical laws and our clear understanding of them. Physical laws are unambiguous, having been tested repeatedly.

In stark contrast, human activity systems (including those classes of systemic problems we categorise as social, economic, socio-economic, and socio-technical) are neither easily nor unambiguously defined. Unlike physical problems where unchanging physical laws define behaviour, governing business rules of human activity systems and transactions within them are frequently ambiguous. Further, they can and do change. We often change governing business rules as we learn, deliberately operate differently by changing the governing business rules, or invoke policies and strategies designed to remedy a perceived or real problem. How we understand these human activity systems changes. Our dynamic hypotheses also change over time.