49316521497_8665749e10_o.jpg
Dave Snowden

Cynefin St David’s Day (4 of 5)

RSS Feed

The ian S5grJkDb9w unsplashIn this penultimate post in the current series, I want to look at what you can manage in the complex domain of Cynefin and return to the which domain is X in debate.  It is worth remembering the end of the Children’s Party Story: We manage the emergence of beneficial coherence, within attractors, within boundaries.   These days I further develop that by pointing out that we can only manage the constraints, the catalysts (attractors can only be catalysed) and the allocation of energy;  everything else is a waste of time. Now that is a hard lesson for traditionally trained managers and the consultants who advise them who have come to believe that its all a matter of setting clear objectives, having a plan and suitable reward mechanisms.  We know they are wrong, but it is also true to say that those who use the idea of emergence to argue for no management, no budgets, no estimates etc. are equally wrong in swinging the pendulum too far the other way.  Although we cannot completely control what happens we can influence it, and we can do a lot to make some behaviours and outcomes very unlikely.   Good management has always been about dealing with uncertainty and doing so in such a way as to leave options open as long as possible.   To be clear I think there is a lot in the Beyond Budgeting movement as well as the No Estimates advocates.  They have a key set of intuitions and experiences of dealing with reality in the complex domain; but an understanding of constraints moves the debate from an either/or to a both/and.  So I am arguing instead for something more nuanced within both those and other such debates. To know that something has been wrong in one context, does not mean that its opposite is right in all contexts; think dialectic not dichotomy.

The other major error here is to assume that because the overall system outcome cannot be known in advance that all management disciplines should be abandoned.  Its emergent is too often an excuse for a lack of appropriate discipline.  Remember the etymology of the word manage: in the early 18th Century the french for household management, manege merged with the Italian maneggiare meaning to handle and train horses. With our knowledge of complexity a de-merger is called for menege for order and maneggiare for complex.  it is surprising that English (considered by the Celts a language far too good for the English) with its propensity for many words with subtle differences for different things hasn’t maintained some difference.  But as Napoleon accurately observed the English are a nation of shop keepers which might explain things.  Management in a complex domain requires greater discipline that order as you have to carefully monitor for emergence as you make changes to the elements of the system which van in various ways be managed.

Just over 16 years ago I put together the ABIDE perspective question for complexity to match the ASHEN one I had created for knowledge audits several years before.   The idea of both is to create a mnemonic for knowable elements of a system. I was staying at my parent’s house at the time and playing with combinations of letters and my father, then in the last months of his life and a crossword addict came up with it.  The net result is that I have been reluctant to abandon it and many people use it to good effect.  It stands for Attractors, Boundaries, Identity, Diversity (or Dissent) and Environment (or everything else).  The problem is that attractors are catalysed and can not be managed directly, boundaries are one type of constraint, identity is an emergent property of the interaction between roles, often heuristically triggered and the environment is just too much of a catchall for what is better described as a dispositional state.  So to be honest with myself it really isn’t sustainable and it reflects a 16-year-old understanding of complexity.  So of late, I have been playing with creating an alternative, along with associated mapping and intervention techniques and while the work isn’t complete its time to share some thinking.  The importance for this series on Cynefin is that it represents aspects of the complex system that have varying degrees of order and can be managed using a process.  There may be a need to create two mnemonics here: one for what can be managed, the other for what should be monitored and I will reflect that a little in what should be seen as an initial and provision list.

Another reason for publishing this now is the current Covid-19 pandemic.  Knowing what you can manage and what you should minter is vital here and it is no coincidence that epidemiologists have nearly all understood Cynefin on first encounter while Behavioural Economists don’t.  The fact that the UK is currently privileging the latter over the former with its herd-immunity policy is deeply worrying, although we are now told this has been adandoned.

So my current and very provisional thinking is outlined below and with links – but don’t take those as definitive.  Also there are not enough vowels in the things we can manage to create a good mnemonic!  It builds on previous lists such as this one, but I haven’t really had time to do a full search on all posts to check I have it all covered

We can manage:

  1. CONSTRAINTS which can contain or connect and which can be resilient or robust and for which we have developed a typology which to date has worked well.  Mapping constraints then determining safe-to-fail means to change those constraints to see what happens is a key technique here and remember in complexity constraints are not negative per se, to be removed to allow flow. Often then enable better flow and the distinction between Enabling and Governing constraints (per Juarrero) has been a part of Cynefin for many years now with the five domains representing rigid, governing, enabling, no effective and unknown constraints as a part of their definition and consequent practice.
  2. CONNECTIONS or LINKAGES are key in that a complex system is defined by the changing and volatile nature of the way identities with agency link and unlink and the nature of those linkages.  Now I am thinking of creating a separate typology of linkages (well more than thinking I almost certainly will) so that constraints are seen as containers. In organisations informal networks matter more than formal systems as the linkages there are trust-based rather than role or function-based.  But changing connectivity and also breaking connectivity is something managers can do – and it is more ethical than trying to engineer a change in so-called mental models.
  3. ROLES can have varying degrees of formality and can have longevity and semi permanence or be temporary and contextual in nature.   Identity is an emergent property that arises from the multiple roles that we all occupy in our daily lives, a crew is the emergent property of deep training and experience in highly ritualised environments and so on.   Our stories have archetypal characters that also represent more abstract roles that we see played out in society.  All job descriptions contain roles, but not all roles are job descriptions.
  4. RITUALS I have referenced from time to time but not yet fully written up a lot of the work.  The key to the use of ritual is that it engrains practices without the need for conscious thought or assessment.  it is also key to cognitive activation.  In an operating theatre, the act of scrubbing up is not only about hygiene it is also an identity shift, changing the response pattern and allowing radically different and professional behaviour for a period of time.  How you join a company, and how you leave are critical to creating identity and engagement and are often neglected.  I remember when I joined Datasolve I was plunged directly into a client as an expert in something I was only just starting to be trained in, and I had to survive with little backup.  When I subsequently left IBM (after one management buyout and a take over I had many years of service) by abiding memory was being chased by a security guard who demanded I park my company car in a different area of the car park.  Both summarised the culture of the company at that time, neither was properly ritualised or thought through,
  5. CATALYSTS are key to attractor stimulation and are often (and more resiliently) oblique in nature.  An attractor can be amplified or dampened but it cannot be created per se.  The essence of managing a complex system and the basis of parallel safe-to-fail probes is to see what can be stimulated and sustained.  You don’t know until you do it, and anything you do will have unintended consequences.  Big interventions are always problematic but may be necessary but even then (in a crisis) creating parallel probes in conjunction with a decisive action makes a lot of sense and reduces risk downstream.  Any decisive action is time-limited in its impact and ability to gain compliance.
  6. GRANULARITY is one of the key elements of understanding a complex system.  The more you chunk things into highly structured and constrained form the less resilient, the less able to adapt.   I normally use information and organisations units as an example.   Highly structured documents take time to create and read, the fragmented nature of information on the internet combines and recombines in novel ways much faster.  This is, of course, good and bad.  Organisational units of 5, 15 and 150 have different characteristics and organising via smaller groups with defined interactions allows more flexibility. This is not an either/or question.  You don’t abandon books and formal documents in favour of blog posts and tweets; you use both in the appropriate context.  Hierarchies can co-exist with the multi-team approach implied by the 5-15-150 construct.  Higher levels of uncertainty to you need more finely grained response, less uncertainty more structure.  Narrative enhanced doctrine is a key capability of SenseMaker® in this domain.
  7. DIVERSITY/DISSENT is also critical.  I have frequently complained about the homogenising effect of common value statements, the whole purpose movement and so on.  The key concept here is the shift from homogeneity to what I have termed coherent heterogeneity.  All systems need dissent which is why I frequently make the point that it is the cynics who care about an organisation; they are prepared to be unpopular, as not been seen as playing team, to say what needs to be said.  Ritual dissent as a method has been one of the most popular things I have ever created and for good reason.  Diversity is complex as diversity simply for the sake of being politically correct can in practice mean two homogeneous opposing forces.  So I tend to talk more about dissent than diversity and, if I do talk about diversity, I focus on cognitive and experiential diversity rather than the more conventional categories of sex, the colour of skin, ethnic backgrounds etc.
  8. CADENCE relates too the speed with which you change things, the cycle time of change and review and with that the nature of the review process.   Taking a one size fits all approach on budget cycles, organisational design etc. etc. is a mistake.  Ironically company handbooks with a fixed review cycle can be better than a constantly updated web site as the cognitive load is less and recognition of change more ritualised.   This is all about flow and kinetic energy.  Too much change and nothing really changes, not enough change and, well, nothing really changes.  The issue is to get the balance right and to know when and how too do things.  There are key links here to the other factors listed above.

In contrast with what we can change, we also have a list of what we can and  must monitor:

  1. DISPOSITIONAL states
  2. PROPENSITIES
  3. IDENTITIES
  4. ATTRACTORS
  5. ATTITUDES
  6. WEAK SIGNALS
  7. COHERENCE

That is for starters, but this post is already too long and I will return and extend that list in a future post

But …

The key missing element here is the question raised by the opening picture of the knot – what is (and should be) in and what is (and shouldn’t be) out of focus at any particular point in time?  With the current crisis on COVID-19 it is a critical question.  I plan to return to this in a wider post on crisis management later this week.  But before then I need to complete the fifth and final post in this St David’s series.

 

The banner picture looks to the summit of Snowden from the col between Mynydd Perfedd and Foel-goch on the descent from Elidir Fawr taken on the 4th of January walking with Euan Semple. It has that wonderful mysterious light that you get in winter as the day draws to a close in the mountains.  The in text picture is by The Ian on Unsplash and uses selective focus to great effect.  I am using it here to demonstrate the ideas of entanglement, coherence and focus all of which are key to understanding complexity.

Top