Sense Making Models

Kurtz and Snowden [2003] have proposed a sense-making framework, device or model called Cynefin also see here. Cynefin has been used to help cope with situations where the normal assumptions of order, rational choice and intentional capability break down. The framework is intended to help us assess whether a particular approach is appropriate for a particular situation or set of circumstances.

The four zones as shown in the diagram are intended to guide us in the way we approach our problems. So if we recognise that our situation is chaotic we may wish to react differently than if we think our problem is known .

Kurtz and Snowden emphasize that none of the zones are more ‘correct’, or to be preferred, than the others but rather are more or less appropriate in given problem situations. For example in the complex zone three generally held suppositions may not always apply. First, assuming a system has order implies that there is cause and effect and hence a right or ideal way to do things. As a consequence they define, rather misleadingly, emergence (the appearance of properties of a whole due to the interactions between its parts) as a kind of unorder. On the contrary emergence is about order arising from underlying interactions whereas un-order implies no order at all. Second, rational choice implies maximizing a utility such as pleasure. This is, we would argue, a very narrow view of rational choice. Third, behaviour is always based on capability and is never unintentional. Again we would argue that many practitioners understand that all decisions may have unintended consequences.

Whilst these three assumptions are neither necessary nor sufficient to explain past complex behaviours they are, we believe, necessary requirements on those responsible for dealing with future complex problems. In other words we would expect practitioners to understand cause and effect through theoretical or practical models. We would expect them to realise that because the models may be variously dependable there is not just one right or ideal way to do things but there is a need to assess the uncertainties and the risks and to be able to admit that they do not know when genuinely they do not know or are having to use judgement. We would expect them to use rational choice but not to rely on only single (or indeed a few) criteria but to use all of the required multiple values and criteria. Finally we would expect them to use their intentional capabilities – in other words to show a proper duty of care expected of all professionals.

We do agree with Kurtz and Snowden that humans are not limited to one identity, do not necessarily act according to predetermined rules, but can act at various levels from the detail to the ‘big-picture’. We explore here the dynamics of learning movement between the zones because they help us understand better possible learning paths. But we should remember two things. First, practitioners have a legal duty of care to use best practice in all situations – not just to the zone of known problems – so any labelling of best practice must apply over the whole domain. Second in order to improve best practice we need to consider how we can clear away some of the fog that clouds complexity.

Below is a modified Cynefin diagram. First we have inserted two axes as measures of the zones. The horizontal x-axis is the increasing repeatability of tests to establish the dependability (common sense, context dependant, truth). The vertical y-axis is a measure of the mix of FIR uncertainty (which we propose can be covered by an Italian Flag).

The Tame zone (bottom right) contains problems containing propositions which can readily be tested and have low uncertainty. Near the origin (bottom left) we have a zone of difficult to test propositions but also with low uncertainty. Issues in the complicated zone have high repeatability of testing but are highly uncertain. Kurtz and Snowden define chaos, like most dictionaries, as the opposite of order – a state of utter confusion and lack of organization. Deterministic non-linear chaos theory is not part of the Cynefin framework. In the figure above we interpret utter confusion as part of complexity. Complete disorder will be at the topmost left of this modified diagram. Deterministic chaos theory on the other hand will be lower down and approaching the contingent zone because there is detectable order which has some of the characteristics of randomness. However if we define randomness as ‘lack of a specific pattern in some information’ then deterministic chaos is not random because there is some underlying order. To understand this we must understand an attractor. An attractor is a set of parameter values to which a system tends to evolve for a variety of starting conditions. A system close to an attractor – even if disturbed – will remain close to it. Dynamic systems often have more than one attractor. Each attractor has a basin of attraction which is the set of initial conditions leading to that attractor. The qualitative behaviour of the long-time motion of a given system can be fundamentally different depending on which basin of attraction the initial condition lies in. Attractors can correspond to periodic, quasiperiodic or chaotic behaviours of different types. A useful way of conceiving a basin of attraction is as a terrain of hills and valleys with a ball moving around and settling at a peak or trough. A dynamic trajectory is the moving ball. A static state is when the ball is at rest in stable equilibrium in a valley or unstable equilibrium resting on the peak of a hill. Simple examples are the buckling of a strut or the dynamics of vibration where at a point of unstable equilibrium the behaviour bifurcates or ‘snaps-through’ from one state to another.

As we move up the vertical axis the uncertainties of the deterministic models underlying the chaos theory behaviours increase until we reach the upper left zone of complexity. Synonyms for complexity include knotty, tangled and labyrinthine. Much of the literature defines complexity as we have defined complicatedness. Complexity science has developed over the last 50 years from a number of different strands of research work. In 2015 it is typically defined as the multi-disciplinary study of systems with many parts that interact to produce global behaviour that cannot easily be explained in terms of interactions between the individual constituent elements. It brings together previously disparate topics such as non-linear dynamics (which includes chaos theory, stability and time series analysis), pattern recognition, game theory, evolutionary theory, artificial intelligence, graph theory and networks as well as systems theory. Key ideas are concepts such as emergence and adaptation. Emergent attributes apply at only one or more layers as a result of interactions between holons at lower levels that do not exhibit these attributes. Adaptive behaviours include evolution, learning journeys and self-organization. There is, as yet, no consensus or agreed definition as to what constitutes complexity science since the subject is undergoing a major paradigm shift.

We think that managing problems in the complex zone requires us to learn our way through the issues.