H

Our Blogs  

 

Gravity is not the same as gravitas

By Dave Snowden  ·  June 22, 2014  ·  Reflections

Ye pushed them raw to the battle as ye picked them raw from the street.
And what did ye look they should compass? Warcraft learned in a breath,
Knowledge unto occasion at the first far view of Death?
So? And ye train your horses and the dogs ye feed and prize?
How are the beasts more worthy than the souls, your sacrifice?
But ye said, “Their valour shall show them”; but ye said, “The end is close.”

Kipling, from  The Islanders

It was interesting to find a picture in which reality caught up with fiction (but don't worry the elephant survived) and its probably time I made the connection between my P&W triad from yesterday's post and the story itself.  At its simplest the point was that learning solely through practice is probably a mistake.  The most basic knowledge of crocodiles and their nature would have rendered the experiment unnecessary, neither does the learning of itself justify the stupidity.  It worked then, it might not work in other contexts.    I'm coupling that with another Kipling poem, one that has imperial overtones but is also a passionate cry against the idea that simply playing cricket and rugger in the private educational system was an inadequate preparation for war.  Kipling was arguing for professionalisation of the army and against the idea of the gifted amateur; a good strategy for muddling though but ....

The lesson for those who focus simply on it worked here or good people will pick what works is that there is a clear role for  engaged theory.  By understanding the nature of crocodiles we can understand their propensities and postulate as to the disposition of a situation involving the banks of the Limpopo River and a lunch arriving driven only by satiable curiosity.  I'm deliberately using some specialist language there.  In a complex adaptive system, especially a human one, we don't have linear material cause but we can talk about the propensities of aspects of the system and the system's dispositions at a point in time.

I have argued for some years that in dealing with human systems the observe-correlate-prescribe model that dominates management science is flawed.  For this we can blame sociology (but then anyone of my generation with a Philosophy degree tends to do this).  The term itself was first used by Auguste Comte (1798-1857) although its nature is as much that of Saint Simon (1760-1825).  The attempt to create a science of society, taken up by John Stuart Mill a generation later can be seen as an attempt to mimic the science of Newton et al, and through observation derive general rules that would allow prediction.  It was a popular science fiction theme a few decades ago with Asimov's Hari Seldon and a short story (I can't remember its name or author) in which a sociologist asked to prove his value to the university sets of a tea party type movement from scratch.  The issue here is all about granularity and mutability.  Gravity is very different from gravitas.  The nature of change and interaction in human systems makes them a unique type of complex adaptive system of itself, and certainly not an ordered one.  Intelligence, intentionality & identity alone distinguish, and make more complex, anything involving humans and human interaction from other natural systems.

So in working with human systems it's not enough to simply observe what happens and try to generalise it.  Instead we need to draw in insight from related fields of knowledge where the scientific method works.  That means the nature of complex systems and human cognition form the starting point for our interaction with reality.   Now on the question of scale, you can't repeat the outcome of an evolutionary process, but you can replicate the starting conditions.  That means that an observed (or more likely reported) success does not of itself justify an approach.  It maybe a one time hit, a Hawthorne effect or you may simply not know of a particular set of circumstances that enabled the result (or prevented more common failure); context is everything. 

So if someone makes a claim'  for a method or approach based on cases (an important qualification), offers training and especially if they offer certification three questions need to be asked.  The 'Ws' of yesterday's triad:

  • What are you claiming actually worked? How many cases, how often repeated in what way?
  • Who can do this, what skills are needed, what experience?  Is it an apprentice model of learning (likely with anything serious) in which case how long and with who by what?
  • Why did this work? Does the reported result match what we know about systems or people?  Can we understand what is stable and can scale?

All of those need to be answered before the How, When and Where of yesterday's Kipling poem.

Now I qualified the question to those claims based on cases.  Under conditions of high contextual sensitivity or uncertainty anything even remotely case based is contra-indicated as a universal.  Here the question changes somewhat to:

  • Postulate: what do we know about the nature of systems or cognition that apply to this situation, what would that lead us to conclude about the propensities and dispositions of the situation? What are the stabilities, the replicable starting points that we can use?
  • Have we got People with sufficient breadth of experience and trans-disciplinary knowledge to handle unexcited eventualities?  If not how do we acquire that, or more ideally activate a network created for that purpose before the situation arises.
  • How can we create safe-to-fail experimental engagement with Practice to allow a sustainable and resilient solution to emerge?  What monitoring and feedback systems can we put into place?  How can we extend the boundary conditions of our situation to maximise exaptive opportunity and prevent strategic shock/surprise?

Now none of that is possible through a one week training course with a certificate at the end of it.  That needs a more substantial investment in network capability.  The certification and training mania, based on limited cases is plain bloody dangerous; the crocodile awaits.