Highlights: STAR series Dialogue March 21-April 4, 2005
Complexity: The Next Big Thing After KM

Please observe copyright statement at the bottom of this document

STAR Moderator Dave Snowden
Director, Cynefin Centre for Organisational Complexity
Marlborough, U.K.

Highlights by Jerry Ash
AOK found, consulting editor./writer
Inside Knowledge (formerly Knowledge Management) magazine

With the passing of the Industrial Age with its command and control, and the coming of the Knowledge Age with its dispersed initiatives and decision-making, management faces complex layers of different types of organization (ontologies) that include -- or amount to -- 'unorder' and demand multi-ontology sense-making (something Jack Ring dubbed MOSM).

Dave Snowden used a story about organizing a birthday party for a group of young children to translate an otherwise academic concept to practical terms.

Rather than create a project plan with clear milestones, a motivational video, a PowerPoint and an after-action review, Dave imagined parents would create barriers to prevent certain types of behavior, use attractors (such as party games, a football, a videotape), disrupt negative patterns early. At the end you would instinctively know whether the party was a success, but you could not have defined what success would look like in advance.

The the story was a metaphor to introduce a new simplicity in decision making under construction at Cynefin Centre, even though the intent was ironic given the use of unfamiliar terms like "ontology" (both simple and complex) and "sense making."

Later Dave made the statement that an unorder system (like the party) arises as a result of the interaction of identities (the children) around attractors (the games) within barriers (set limits). That's the essence of an MOSM style of management -- different rules for different types of domains:

Simple Order:		Sense-Categorize-Respond
Complicated Order:		Sense-Analyze-Respond
Complex Unorder:		Probe-Sense-Respond
Chaotic Unorder:		Act-Sense-Respond

He was helped in his quest to simplify the concept by several AOK members who are also students of this developing new approach involving seemingly contradictory styles of management -- as one popular song goes, "a little bit of country, a little bit of rock and roll" -- a blend of old and new style management.

Members cited numerous examples of MOSM at work.

Piers Young, KM researcher, London, cited an example of a U.S. Marine. On his first tour of duty in Vietnam, as a young lieutenant, whenever he heard gunfire, he got on the radio to ask what was happening -- his assumption being that there was an easily assessable situation and by his being told it, he could make the requisite changes in orders etc.

Where he stood out was in relatively quickly realizing that there was an "ontology confusion (wrong approach for wrong situation)". He quickly spotted the confusion: from an ordered position he was asking people in a complex/confused situation to explain what was going on. In his words: "The danger in calling is that they'll tell you anything to get you off their backs, and if you act on that and take it at face value you could make a mistake. Plus you are diverting them --they are looking upward instead of downward. You're preventing them from resolving the situation."

His resolution to the problem was, on hearing gunfire, to look at his watch and wait five minutes. If they need help, or to "look upwards," they would shout.

Dr. Larry Irons, managing partner, Osprey LLC, St. Louis, Missouri, U.S, had read an application of the Cynefin framework to emergency response management. The article, by Simon French and Carmen Niculae "Believe in the Model: Mishandle the Emergency," appeared in the Journal of Homeland Security and Emergency Management, vol. 2, issue 1 (2005).

The authors briefly apply Cynefin to Three Mile Island, Chernobyl, and mad cow disease. In the context of emergency response Cynefin appears very useful in managing the risk communication process with stakeholders, including the public.

Jack Ring, proprietor, Innovation Management, Scottsdale, Arizona, U.S., wrote: "One of the more dramatic cases was the incident in Somalia where bodies of dead U.S. Marines were dragged through the streets. In retrospect, this resulted from a clear difference between (a) the ontology of commanders on the scene vs. (b) the ontology of Bill Clinton and Kofi Annan (allegedly on behalf of his friend, the [then] reigning warlord).

That reminded Dave Snowden of another memory:

"Over the last three years we have worked a lot with the case of the USS Hood working through a lot of original material and concluded that the impact of the attack could have been minimized if the problem had been treated as complex not complicated. In the case of the Hood, a pattern had set in between 'going in' (to port) or 'not going in.' One part of the US Government said go in, it will be safe; the other part said it's not. When the 'go in group' won -- and because of the arguments -- it was assumed that it was safe and you got crazy things like people on the decks not authorized to use guns when a terrorist boat is steaming towards them."

Barb Krell, KM Consultant, Enterprising HR, Calgary, Canada, had a good example: "I recall a situation that was discussed as a problem that could benefit from using the Cynefin domains. The situation was one of funding for research which involved a university, government grant funding and researchers as key stakeholders in the process. The process of granting of the monies to actually conduct the research was completely stalled over conflict in creating acceptable contract language. The lawyers were involved trying to create contracts about ownership for intellectual property that might emerge as a result of the research. So they were trying to create very specific language about outcomes (cause and effect) in a complicated way that couldn't really be predicted in advance, but needed to emerge (complex ontology).

Dave Pollard, consultant, Meeting of Minds, former STAR moderator, Canada: "Yesterday I had a long conversation with a consultant in BC who works closely primarily with aboriginal communities on some very vexing social problems. He has found that 'problem-solving methodologies' simply don't work in complex (as opposed to complicated) systems. So he has fused three approaches, drawing to some extent on the ways in which aboriginal communities have always dealt with complex systems.

"The three approaches are: (1) Appreciative Inquiry, (2) Open Space meeting protocols, and (3) what he calls the Four Practices -- a set of skills/attitudes that bring out the best in the members of the team & community. You won't be surprised that responsibility, respect, conversations and stories all play key roles in this approach."

Late in the Dialogue, Dave Snowden introduced members to a model being developed at the Cynefin Center called Abide (Attractors, Barriers, Identity, Dissent/diversity and Environment).

The idea of ABIDE is:

• to create a way for managers to talk about unordered systems in such a way as to prevent their focusing on output conditions, or to impute linear causality.
• to allow managers to use ordered systems approaches in an unordered space, its what they have been trained to do so you won't stop it, but you can direct it at the ordered aspects of an un-ordered space and ABIDE is intend to provide that.

This way you can keep it very simple: Is it ordered or unordered? If ordered, engineer an outcome based solution; if unordered, what are the ABIDE elements, which can we change, which should we change, how do we monitor for and respond to emergence?

Snowden says that among Cynefin's more advanced customers and research partners, ABIDE is exciting as much if not more attention than multi-ontology sense making as it seems to resonate with a natural way of thinking for them, and enables them to translate ideas into action. The ABI elements are well developed; the DE elements less so.

Attractors: To date the developers have found it useful to use two separate ways of describing these and the two can form a matrix:

1. Separated into single point (a basin) and multi-point (the ice cream sellers) attractors. Single point is more controllable, but if the attractor goes then we have chaos; multi-point has more resilience. (This picks up on a major theme of our work namely the balance between stability and resilience that in nature are generally opposed).
2. A separation between those we can control (like an incentive scheme) those that just happen, to which we respond (like competitor activity) and those which have evolved over time as a result of multiple interactions.

Barriers: Rather than boundaries as it seemed (a) more accurate and (b) prevented confusion with boundaries in the Cynefin framework. Again there are two ways of describing them:

1. A distinction between rigid/brittle (any thing rigid is brittle because when it breaks you can't recreate it except at enormous cost), elastic (more resilient but can break) and permeable/semi-permeable (high resilience low control).
2. Repeating the control/happen/evolve one from attractors.

Identities: A human can be a part of many identities and the context will determine which dominates.

At the moment, Cynefin separates these as follows:

1. A powerful role or individual.
2. Context stimulated - if we are all in a building which catches fire, the context will create identities and associated behaviors that are only possible in that context.
3. Continuity of membership, the cohort group, people with a shared interest in sport or politics etc. Here the group has some common experience (we all did our MBA together) or belief (we hate capital punishment) which creates an identity that has cohesion.

Dissent/Diversity: This picks up on the exploration v. exploitation distinction that is frequently made in complexity writing.

Environment: At the moment, a catchall which picks up various aspects such as proximity and volatility.

Mark Ranford, managing director, Stratagility, Bali, deserves recognition for the most meaningful post among many quality posters. His in-depth sharing included the following:

"Complexity is ubiquitous, and when we start to recognize where it applies (or doesn't) using approaches such as MOSM, it is as if we have suddenly cleaned our lenses which have for so long been misted and obscured."

He went on to comment on the writings of other complexity theorists including Clay Shirky, Adam Bosworth, Steve Gilmour and Stuart Brand (author, How Buildings Learn).

Participants in this Dialogue, in order of appearance, included:

Jerry Ash, Valdis Krebs, Dave Snowden, Jack Ring, Verna Allee, Dermot Casey, Piers Young, Chris Macrae, Dr. Larry Irons, Shawn Callahan, Barb Krell, David Pollard, David Jones.

Thanks to all. Without these participants it would not have been a Dialogue.

Links

• Complete background and archive of the Conversation with Dave Snowden
• To follow and participate in future STAR Series Dialogues, join AOK

Back to top