Dave Snowden

Safety zero?

RSS Feed

An interesting day at the Summer Institute which ended gloriously with a long evening of discussion and story telling with Vernon Cronen and John Shotter at the conference dinner.  All three of us products of a liberal arts education which meant the discussion ranged from Thurber's humour to comparing politics in the US to that of Trollope's Palliser novels.  It also included a comprehensive discussion of the Battle of Gettysburg, the Civil War in general and the failure of British and French observers to learn the lessons of the Siege of Richmond in WWI and so on. It was the sort of enriching conversation that is increasingly difficult to get at events and that worries me.  With the Economist's cover article last week talking about 'Creative destruction' of the University through neo-liberal consomerism I am starting to think that for too many people what I would consider dystopian is being heralded unthought and unthinking as a new utopia.  

The day started with my rushing over to do a camera interview on Cognitive Edge and Cynefin then I had a chance to hear Barbara Czarniawska talking about Negotiating Selves which was an interesting take on identity discussed mainly in the context of gender but with an interesting analytical framework.   I'm going to take that up in a future post but its been good to have a few fleeting conversations with Barbara although both of us are frustrated as we want to attend each others sessions but yesterday and tomorrow we both have two sessions, but we run in parallel so that won't be possible.   A pity the organisers didn't pair us in one session as we both have significant narrative work to talk about and the interaction would have been good.

From lunch time I left the Summer Festival to be a part of a workshop on safety run with customers and prospects of Attractor (the consultant group who are the main sponsor of the Festival).  Now this was interesting and not without controversy.   I opened with a basic introduction to complexity thinking and then talked about the various approaches to safety work that arise from that.  Those I summarise below and there is a lot of SenseMaker® in it.  We are partnering with Attractor at the moment in this and other fields and it looks to be a good partnership.   I was then followed by Erik Hollnagel who was presenting what he has called Safety-II which is based on an understanding of what work actually is, rather than what it is imagined to be.   Now when I read that I thought we would have a rushing clash of agreement but it turned out not to be the case which is going to make out joint session tomorrow interesting.

Now there was a lot to agree with.  He argued against the causality credo of zero failure, the assumption that you can create a system in which nothing goes wrong.  He validly pointed to the way (a great example here) of how when we move into a new hotel room we immediately rearrange the desk to match our personal work pattern.  He was arguing (rightly in my view) that people manage safety by multiple small adjustments or adaptation.  He used walking through a busy street and a roundabout with no markings (note that for later) in India (I think) where the traffic self manages.  Again good arguments here, but I will qualify the roundabout example later.

Where I started to disagree was when we got a dichotomy between Safety-I and Safety-II with the argument being made that we should focus on the day to day when everything goes right and ignore the events in the tail, the exceptions of accidents etc. where most safety work focuses.  In later discussions he regretted even using the safety word.  This is where we disagreed.  As far as I am concerned the tail is critical (and I will develop this tomorrow) as its the outliers where you find threat and opportunity.   I also think that anyone responsible for safety who neglects this is simply going to loose their job.  That said I don't think Erik was advocating that other than as a thought experiment to enable a different perspective, or at least I hope not!

What Eric, and another respondent from Siemens both said (and I agree) was that we have to accept a degree of non-compliance and failure is inevitable.  Where we disagree is what we should do about it.  So, in summary the points I made:

  1. I introduced exaptation as a concept, arguing that narrative approaches to safety allow rapid exaptation under unforeseeable future states, its just letting people get on with things but finding ways (which technology enables) to break down the barriers of time and space that normally inhibit knowledge and information flow.
  2. I moved onto co-evolution with the strong argument that you have to deal with where you are, not where you would like to be and made the brown/green field architecture point which has been a recent theme of several keynotes.  In that context I talked about one of our new SenseMaker® offerings namely to provide an attitudinal map of safety.  Attitudes are a lead indicator, compliance is a lag indicator.
  3. Thence to complexity and the children's party story and the importance of paying attention to weak signals, to small indications in the tail and the management of boundary conditions.   This i where (in the question session at the end) I went back to the roundabout.  The point I made is that the indian traffic management system is a unique product of a co-evolutionary process between an asian culture and a developing and changing mode of transport.  You can't transpose that without change.  In contrast the magic roundabout does similar things, but it is constrained by some white lines and compliance with a limited set of enabling constraints or rules that drivers in the main comply with.  The choice is not between rigid constraints or no constraints, but the level of constraint and its fluctuation over time.
  4. That gave me a base to talk about extracting heuristics from existing practice and instantiating them as a formal governance mechanism when the rules no longer apply.  In turn that allows simpler more memorable rules with rules about who, in what context, can break the rules.  But if you don't just accept that rules will be broken, you use the heuristics to govern behaviour.  Break the rules and follow the heuristics and is OK, ignore the heuristics and you will get punished.
  5. Build worst practice systems, people are better able to avoid failure than imitate success.  One person in the audience thought I was talking about punishment and encouragement which I was not, I was simply talking about ​small noticings (see yesterday's post) or anecdotes of failure and success that provide learning.   There I talked about conceptual blending and the need to allow action to be taken based on a mental rehearsal of response to a cluster of shared anecdotal material rather than an adherence to best practice.
  6. Ritual, in particular changing cloths, passing through obvious barriers are another way of ensuring that cognitive activation enables people to pay attention in a different way
  7. Field recording systems (something SenseMaker® can do well and in future releases will do even better) provide real time feedback and critically safe people time in the here and now.  Such systems allow immediate real time feedback which is key to managing a complex system.  This got attention in questions and I needed to emphasise that you can't ask someone to take on a new reporting requirement if you don't first take away a more burdensome one.  So its keep this field notebook uptodate and we won't require a end of day report to be written or similar is the way to do it.  Such systems can also measure attitudes and provide a new way of gather material for other purposes.

I did a lot more than that and I really need to write a wider article about safety.  As luck would have it one of the participants (who has been on our training) wants to co-author a paper so that will get underway soon.

Finally I decided (and I may have to apologise for this) that I was not interested in Safety-I or Safety-II, but rather Safety-0.  Sitting back and thinking again, synthesising the old and the new and at all costs avoiding dichotomies …..


​The sign that illustrates this post comes from a great collection, some rude but all fun!