My original blog The Emperor's Chess Board was republished on the ActKM forum and its been interesting to observe some of the exchanges, although I am staying out of them for the moment as part of a self-imposed break from listservs which may become permanent. In general they don't seem to have grasped the idea that I was challenging the concept of centralisation of a government knowledge function (I will qualify this a bit in a future post), arguing that it would manifestly lead to failure to achieve the key goals of making a nation more secure. So far the following themes seem to have emerged:
- The idea that this time we will get it right. The argument here is that knowledge management has learnt from its mistakes and a centralised function would not focus on information management solutions alone. This has also introduced the concept of of speaking truth to power, which seeks to directly address my earlier point about about the difficulty of joining up the dots. The argument here is that in the cases I quoted (the shuttle disasters for example) people did know what was happening but were overruled by political considerations.
- The scary idea that any approach needs clear definitions of knowledge management, and criterial by which various claims can be validated. This of course means adopting a particular philosophical approach (in this case a variation of Popper) and requires (i) a degree of intellectual vigor in government that is not likely any time soon and (ii) fails to appreciate the messy bottom up approach implied by taking a complexity science approach to the problem.
Now I don't intend to devote anytime to the second theme, I think it defeats itself by its nature. However the first this time we will get it right approach is interesting if only for the illustrations it gives of linear thinking which so prevail in this space. So in this third post in the series I want to address that, in effect an extension of my earlier two posts before moving to some hopefully pragmatic suggestions on the way forward.
Now it is undoubtably true that frequently people who need to be heard are not heard. With the benefit of hindsight those who got it right suddenly become seen as prescient martyrs whose advise should have been taken. On the face of it this seems to make sense, and if there advise was rejected that its easy to find a source to blame, politicians or whatever. We like to believe that there is a reason for things, its called fundamental attribution error but following it is rather like the Emperor who was conned out his rice crop, following what may seem to be common sense can cost you a lot, in this case probably detecting the pre-conditions of the next terrorist attack.
One of the many problems is there sheer issue of numbers again. In a very large and distributed workforce the number of people making educated guesses or hunches is very high. We tend to underestimate probabilities. Once the event has happened we pay attention to those whose predictions proved correct, we don't pay attention to the number that were proved incorrect. One famous con trick involved someone sending out a thousand emails and predicting that a stock would go up to half, and down to the other half. When it went up a second email went out to 500 to say "I got it right" then a prediction on another stock to go up went to 250, down to the other 250. The process continued until a small group of people thought the con-man was infallible at which point they were stung big time. In any event, after the event some will have been proved right, statistics tells us so. It does not mean either that there were right for the right reasons, or that they will be right again, or for that matter that they should have been paid attention to.
We did an experiment some years ago, working with the 911 report to get people to understand complexity concepts of attractors. the morning was a complete failure. With the benefit of hindsight they could all see what should have been done, they could not describe it in terms of uncertainty. Over lunch I was depressed. But then a conversation started about a current situation where the outcome was uncertain and what signals should have attention paid to them unclear. All of a sudden in the context of uncertainty, methods designed for uncertainty worked.
What you have got here, and it is VERY VERY dangerous, is using hindsight as a substitute for foresight. We need methods and tools designed for that and they do not include centralised KM functions for government, but rather something more complex. More on that when I have thought up the next variation of Empire for the heading.
Comments (6)
Since years I argue against too detailed planning when it comes to software requirements specifications, but this is ignored or considered "sloppy" or "poor planning". But yesterday, someone from "the other side" said: "... and we will not put too much emphasis on mapping business processes, because these will be different when we are ready to roll it out". Finally! SOunded like an acho of my words coming out of someone else's mouth! I can live with the fact that it sounded like it had been their idea and not what I was preaching for years. I know that they went through some painful experiences recently, and this was their conclusion. The good thing is that their learning (from failure) was on a higher level. Not: now we know that A was wrong so we do B in the future (which seems like what you mention above), but: now we know that we can't forecast A nor B, so we have to be able to cope with both (by: some few sound rigid basics plus appriopriate flexibility plus iteration ... but that's another topic).
Posted by christianhauck![[TypeKey Profile Page]](http://www.cognitive-edge.com/blogs/dave/nav-commenters.gif)
|
August 6, 2008 1:26 PM
Posted on August 6, 2008 13:26
Dave,
A good example of the idea that this time we will get it right, in the context of counterterrorism, is the recent paper by David Tucker Terrorism, Networks, and Strategy: Why the Conventional Wisdom is Wrong in the Homeland Security Affairs Journal. Tucker contends,
http://www.hsaj.org/?article=4.2.5
Posted by Larry Irons | August 6, 2008 3:28 PM
Posted on August 6, 2008 15:28
As someone who has undertaken postgrad foresight studies, I totally agree with your statement about using hindsight as a substitute for foresight. There are a host of methodologies and tools that work across input analyses, worldview exploration and prospective futures. However, I believe that centralised KM functions may be useful, if only to help distribute knowledge between the decentralised functions that may not be clearly visible; catalysing and clarifying rather than controlling.
Posted by Luke Naismith | August 6, 2008 6:19 PM
Posted on August 6, 2008 18:19
Larry.
Thank you. Yes, a lovely example. "I'm going to put my fingers in my ears, sing la-la-la and pretend it's all ordered really".
The scale of military defeat that the West is looking at just scares the hell out of me. Brave New War is just not getting listened to.
Posted by Brian Sherwood Jones | August 6, 2008 11:28 PM
Posted on August 6, 2008 23:28
The idea that this time we will get it right.
In my opinion, this notion has been a fundamental message from large consulting firms to organizations for a long time now (since the start or the large-consulting-firm industry ?) as in "our approach / methodology will help you get it (strategy, structure and tactics) right this time".
While it is acknowledged that change is constant, the use of scientific management principles has implied that said principles are inviolable and do not need deep examination or re-invention, such that it is only necessary to come up with that next better-thought-through strategy or structure. But then, when context, conditions, markets or customers change (which is increasingly evident, I believe, in the increasingly interactive environment of the Web) .. oops, gotta re-think that strategy so that "this time we'll really get it right". And so it goes.
The complexity, and the clues available from watching and participating in bottom-up interaction are ignored in favour of using theories and frameworks then used to categorise after which this next right strategy is inflicted, as you have often noted in your writings and courses.
A very tenacious mindset indeed.
Posted by Jon Husband | August 7, 2008 11:31 PM
Posted on August 7, 2008 23:31
I am not sure whether my understanding is close to being useful, but here goes:
IMHO it seems that the errors we make are a result of our assumptions about how cause and effect work across time and place.
For certain physical phenomena cause and effect are fairly consistent and knowledge can be treated as if it were a 'thing' from which our actions and arrangement can reasonably follow. With care prediction is reasonable.
For social phenomena, cause and effect are unlikely to be consistent over time and place and thus prediction is problematic, so we need to continually construct and reconstruct our knowledge, actions and arrangements.
Ivan
Posted by Ivan Webb | August 19, 2008 10:13 PM
Posted on August 19, 2008 22:13