Harvesting knowledge - can we really do it?

Why harvesting is the wrong metaphor

Knowledge is not an object, it is ephemeral, emergent, very relationship & context dependent. To get someone to share their 'knowledge' you need to have their trust and they must be motivated to spill the beans. Knowledge is not something that has a separate existence - you cannot grab or contain knowledge - it requires dialog, community, time, testing & verification to develop and evolve.

Many companies are climbing on the continuity management bandwagon, riding the demographics as the boomers begin to retire and there are dire predictions of severe erosion of institutional memory and organizational brain drain. Let's revisit their arguments:

* identify your experts and asses your risk of knowledge loss - really hard to do for tacit knowledge
* have a plan to harvest ahead of retirement and loss from job changes - expensive, risky and gives uncertain results
* capture and package your knowledge assets before it is too late - there are deep issues here
* leverage your knowledge capital - difficult to do without a community
* speed new employee learning and induction via access to collected knowledge - a myth

The difficulties of capturing and depositing 'knowledge' are well documented. Acquiring understanding and meaning depend on context, exposure and continuous dialog. Knowledge is best 'preserved' or stewarded within a community in the form of stories than in a repository in the form of assertions, rules, examples or patterns. Knowledge is emergent, it flows via relationships and dialog, it requires interpretation to be surfaced from information, it needs a shared context to be explicated and validated.

Tools that claim to harvest knowledge are little more than templates with generic questions and following a prescribed process. Knowledge acquisition is hard, takes a long time and requires domain expertise (enculturation) for effective transfer:

* colleagues and staff are wary of giving up their tricks and tips. There must be a suitable culture and climate in place
* most of the really valuable heuristics are tacit in nature - the experts are not aware and then cannot explain how they do it
* true knowledge is not static, it is outdated at the moment of capture as the environment changes
* what adds value and makes the difference is the expert's ability to handle new problems and make novel distinctions - this cannot be captured
* experts rely on a large invisible personal network to keep them aware, to validate their ideas and to surface new solutions - when the expert departs, they take this network with them or at the very least the social capital that enables network cannot be easily regenerated
* something is always lost when you represent the knowledge - there are subtle trade-offs, language substitutions, issues with context specification and it is impossible to record all the connections that make emergent knowledge so rich.

Much of the early work in knowledge acquisition and elicitation was connected with expert systems, but this approach was found to be 'brittle', it required well-bounded static domains and was unable to incorporate learning. A far more promising avenue is to encourage communities of practice where informal and social learning happens and staff gather around a set of issues and capture their joint expertise through doing and being.

Here some knowledge harvesting tools - take a tour for yourself:

* Cerebyte
* Knowledge harvesting
* Knowledge elicitation
* Knowledge acquistion - AKT

Comments

50/50: 1/2 of the cited information is accurate; 1/2 is inaccurate.

[ian]

this site is madness