Sail, not Rail
Originator of the Cynefin Framework Dave Snowden states:
“In an ideal approach you define how things should be and attempt to achieve it. In a naturalistic approach you introduce technologies and practices on a safe-to-fail basis and see what works. Amplification of good patterns, dampening of bad patterns allows something to emerge that is more resilient and a risk free, not to mention a lower cost, solution.”
An idealistic analogy would be a driving a train while a naturalistic one would be sailing a boat.
Knowledge Management 1.0
Systems Thinker Ikujiro Nonaka is credited with establishing Knowledge Management as a discipline circa 1990s. In his HBR article The Knowledge-creating Company, he helped popularize the notion of “tacit” knowledge, the valuable and highly subjective insights and intuitions that are difficult to capture and share because people carry them in their heads. The KM challenge was also to find ways to manage “explicit” knowledge which could be documented, stored, retrieved, and disseminated.
Early KM 1.0 initiatives adopted an idealistic approach. One would install an intranet/portal, create a taxonomy and motivate people to codify their knowledge. Another admirable plan would appoint a KM Officer or Coordinator to design a standard for building a community of practice (CoP), with roll out plans, templates, process, budget, etc. After a big launch, people would be encouraged to submit their documents. Some would, buoyed by the initial fanfare and excitement. But once the novelty had died down, KM sponsors noted a huge fall-off of submissions. Motivation tactics such as incentives, annual performance goals ensued to get people to re-engage. Consequently, KM in the Control of Information paradigm became a “I have to” vs. a “ I want to”. People still contributed but only just enough to get by and avoid punishment.
Interestingly, other KM initiatives were started but on an informal basis. They were mainly ad hoc with no structured process, budget, nor authority. People were not directed to attend; they came because they wanted to share knowledge. Under this naturalistic approach, informal work and social networks sprung to life and were useful in resolving organizational issues. Not surprisingly, when the people in charge of KM found out about them, their first inclination was to put them under their control! Clearly an unwise move since this type of activity is complex.
Knowledge Management 2.0
KM 1.0 is a techno-centric approach. Get users to submit knowledge to be stored on a central server. Often your submission is vetted by a KM coordinator, cleansed and indexed. For access, submit a formal request to a gatekeeper (human or machine) that searches and returns objects according to some hit ranking.
KM 2.0 is people-centric. Remove the gatekeepers to ease capture, storage, access, distribution, and removal. The advent of social media tools has immensely aided the KM evolution.
Often there is no “end-in-mind” vision. Because you don’t exactly know what people will use, you begin experimenting by blogging to post your thoughts and opinions. With linkages to Facebook, Twitter, LinkedIn and other free-to-join social network websites, you discover that you’ve attracted others who have a similar interest. You host a Skype video chat or Google+ Hangout and garner their feedback to further test value and commitment. You decide to open up a Wiki to begin documenting and sharing your content knowledge in a more structured manner. Through positive reinforcement (e.g., favourable comments, repins, retweets, +1 and “Like” icon clicks) the network you started goes viral and up the life cycle Growth stage you go. It’s naturalistic, not idealistic.
Narrative KM example:
Human Sensor Network
Here’s another way KM databases are being developed. In Healthcare, for years KM involved maintaining patient records as paper documents stored in a filing cabinent owned by the patient’s doctor. Sharing of information was achieved through the postal or a courier service. Advances in IT computerization led to digitizing of records housed in central servers. Records are updated by the doctor or assistant via keyboard entries. Today’s paradigm works but is limited since it only captures the physician’s side of the story. What about the patient’s?
Experiments are underway to equip patients with smart device apps that allow voice recording of their experiences. Their stories are then self-indexed and sent to a narrative KM database.
We’re just beginning to discover the immense benefits of a narrative database to augment the medical records. Now when a new drug or pill is prescribed by a physician on a trial basis, the patient can describe how she physically and emotionally felt after taking a dosage. This feedback loop provides early awareness and triggers the physician to stay with or change the dosage level. If the doctor and other physicians have prescribed the trial drug to other patients, we essentially have a “human sensor network” in place. You can see why this experiential database is deemed more valuable by doctors and nurses than the information provided by the pharmaceutical company.
Can patient stories be accepted as reliable data? Because they behave as agents do in a complex adaptive system, the answer is yes. Firstly, the storytellers have enough knowledge to not act foolishly and treat the reported experience as a serious matter. Secondly, since patients rarely know each other, they are act autonomously and therefore, their experiences do not influence each other. This characteristic of complexity is called Distributed Cognition, also commonly known as the Wisdom of Crowds.