Deliberate Practice - Cognitive Edge

In the 30s most people who went to on to become vets were from farming or veterinary back grounds. Alf Wright was an exception to this. His father was a ship plater which meant that unlike his many fellow students he did not grow up with animals. I know about the harsh reality by the way; the lot of a vets son is to help out with blood samples, sheath washings and the like during the holidays. You learn through bitter experience as two examples that are still vivid 30 years on will illustrate. Lifting the rail of a Friesian so that a sample can be taken involves standing to the side to avoid the deliberate and malicious use of streams of excreta by that species; you only make that mistake once although the laughter of the famers and your own father make it stick for ever. Sheath washings involve entering a pen containing a +1000Kg Hereford bull, grabbing its penis and well, washing it. I could tell you about the slaughter house inspections as well but I assume that many a reader may still be consuming Christmas leftovers and have a sensitive stomach. A few years of this and I decided that the mathematics of theoretical physics and the purity of Philosophy were my future.

Now the point I am making here is that most people entering a veterinary college already had a body of experience of animals acquired over the years. The college provided an academic top up to that body of experience. Familiarity was important. I remember during the many outbreaks of foot and mouth my fathers generation would make a rapid diagnosis, slap a Form A (movement restriction) if the decease was evident and then send the samples off to the Weybridge Labs for confirmation. They rarely got it wrong and the trust of the farmers was therefore won. Now with the romanticization that came with All Creatures Great and Small applications to Veterinary Schools went up, the qualifications went up to the point where you had to get three As at A Level to get in and the experience level of graduates went down considerably. At the time of the last foot and mouth outbreak, I was told by several senior civil servants that the percentage of test confirmed diagnosis had gone down over previous outbreaks. Lack of experience, compounded by the long time interval between outbreaks meant that a more academic generation played safe, trusted their judgement less and in consequence produced a lot of false positives. Net result reduction in trust of farmers, more reluctance to report potential case and a key 24-48 hour delay in reporting which increased the spread of the decease.

That brings us to the 10,000 hours popularised by Gladwell. Now Gladwell is a good journalist, he is adept at picking up the soundbite and summarising supporting evidence. Wilson in his posting makes the important point that it is the nature of the experience that counts. He argues for experimentation as a means of achieving this and sites research which shows that experimental approaches are more effective. I'm reading the autobiography of Steve Jobs at the moment and that contains myriad examples of this simple fact.

However I don't think Wilson goes far enough. The original data on the 10k hours comes from things such as musical education which is an apprenticeship model. That means that much of the experimentation results to failure and embarrassment (think Friesian & Hereford) in a social context. 10K is also too broad brush a number and it needs to be qualified by things such as natural talent, the quality of the learning environment etc. Overall however the point remains; you need an extended period of experimental learning with a significant proportion of that learning being tolerated failure.

The same applies for an organisation as for an individual. Its not enough to run experimental programmes, the key word in deliberative practice is deliberative. There are some key additional requirements:

1. Experiments need to be constructed to run in parallel
2. All experiments must be designed as safe-to-fail
3. Amplification and dampening strategies need to be in place before the experiments are run
4. Managers need to be targeted on the basis that at least half of their experiments should fail
5. Research and monitoring needs to provide real time feedback

Now those are all features of interventions in the complex domain of Cynefin. However it may not be enough if we ignore the social processes of learning that create collective capability as much as individual competence. Matrons (something I blogged about back in 2006) had individual competence, but it was developed within a highly ritualised social situation. We need to start thinking about the social context of knowledge development as much as, if not more than, individual competence.