Systems

Structure

Actions

controls? inputs?

An action is just a special type of element. When we perform an action, we directly change an element. Normally, action elements aren't changed by other elements. However, action elements change other elements via the relationships between them.

Outcomes

effects? results? outputs? consequences? impact? outcomes? responses? products? behaviours?

An outcome is an element whose value is directly important to us. Although the system may contain many other elements, we don't directly care about them because they're just intermediate values. For example, the amount of carbon dioxide in the upper atmosphere is not, in itself, a problem to us: we can't see it or smell it or sense it in any way, The outcome we care about is global warming.

Elements

Examples of nodes: a person, all people, animals, weather temperature, happiness, distance, speed, humans vs. other living and non-living things.

Elements need not be of the same kind. A simple system might have elements of the same kind (eg, two bank account balances) but, in general, there will be elements of different kinds (eg,  bank account balance, quality of furniture, happiness, etc).

Note: not normally drawn with boxes around them. However, on this site, elements are drawn with boxes around them so colours can be used to point out particular features.

The positions of elements in a causal-loop diagram aren’t significant; it’s the relationships that dictate the system’s behaviour. [relate to train pic and CLDs in relationship-scope qv?]

Relationships

Relationships = links, interactions, effects, influences, flows (stock-and-flow diagram).

+ and - on CLD arrows.

Delayed effect (double-bar across arrow).

The direction of the the relationship arrows indicates direction of influence, not the direction of motion. 

Delays

Feedback

An example of a feedback loop is the interaction between polar ice caps and global warming. Warming causes some of the ice to melt, which reduces the size of the ice caps. Unfortunately, ice is much better than water or land at reflecting the sun's energy back into space, so the result is that more of the sun's energy remains on earth, which causes increased temperatures. An additional effect is that melting permafrost releases large amounts of trapped methane, which is a greenhouse gas much more damaging than carbon dioxide. The increased temperatures resulting from those effects cause the ice caps to melt faster, so the cycle continues. DIA

As is often the case with feedback loops in practice, this behaviour doesn't continue forever: it reaches a limit and a new equilibrium state may arise. If all of the ice caps melt, the feedback loop can't continue so there would be no tendency for temperatures to continue to rise as a result. However, this doesn't mean that temperatures would revert to what they are now; instead, they would remain at levels that would make human life very difficult. (Of course, there's much more to climate change than just ice caps, so this effect can't be treated in isolation.)

Boundary

Behaviour

Rationalise with systems-thinking qv; transfer generic content from there to here.

Linearity

Non-linearity (non-proportionality).

Emergent Behaviour

Tipping Points

Can a tipping point result in a system having more than one possible emergent behaviour? A sand pile doesn't have a tipping point (other than little avalanches) and has only one emergent behaviour (cone). A bucket of water has a tipping point and has two emergent behaviours: full and empty.

Picture? tipping-point (bucket/glass toppling)? beam overloading? balloon overinflating?

Rejected