Wednesday, 21 December 2016

How KM deals with "We don't like to write things down"

A few times I have worked with organisations who say "We don't like to write things down in this organisation". So how do you address this in your KM program?

Image from "The writers advice" blog
This happened to me again a couple of weeks ago. "People here are not keen on writing" I heard - "documenting knowledge is the wrong approach" - "we need more conversations between people" - "the best place to store knowledge is in the heads of the experts". 

All these people I was speaking with had a good point. Documents are far less rich as a source of knowledge than experts, and reading is a far less efficient way of gaining knowledge than dialogue. I estimated in a previous blog post than transferring knowledge through writing is 14 times less effective than transferring through conversation.

So why is any knowledge documentation needed? Could I not say to this organisation "Yes, don't bother writing anything down - just build the conversations through which knowledge is transferred"?

The truth is that documentation and conversation (or content and conversation) are not an either/or choice; more of a both/and, and there are some cases where documentation is safer in the long run than storing knowledge in memories. Memories have the following flaws;
  • Memories fade over time, even though we think they don't. The "illusion of memory" is a real thing, which needs to be taken into account in KM - you cannot trust a "fact" that has been stored in human memory alone for more than a few months.
  • There are false memories. People remember things clearly, which never happened; for example the story of the journalist who publicly accused George Bush of making a statement which in fact he had not done.  This is "misremembering".
  • There are other human cognitive biases that distort memory. The confirmation bias is very strong, for example, leading people to select examples which support the opinions they already have. All of us are subject to this bias, even our best experts. 
  • Groups of people can form knowledge bubbles; a strong form of confirmation bias where facts and knowledge counter to views held by the group are just filtered out. Witness Brexit, for example.
  • Memories are held in heads, heads are attached to legs, and legs will eventually walk out of the building, never to return.

So on the one hand there is human memory which - as rich and nuanced as it is - contains much that is opinion, bias, prejudice, or just plain false. 

On the other hand we have documented knowledge - sparse, lacking in detail, cumbersome to create and digest, but immutable. 

How do you choose which route to take? And how do you get an organisation that doesn't like writing, to start recording knowledge?

I answer the first question in this blog post, where I suggest that the times to leave knowledge in the heads of people are

  • if there is limited demand for it;
  • if the expert is guaranteed not to forget (ie the knowledge is in constant use)
  • if the expert is guaranteed to stick around 
  • if the practitioners - expert and novice - are connected
  • if there is enough openness and challenge in the organisation that the knowledge bubbles will not form.
In all other cases, the knowledge should be written down. But if the organisation really does not like to write, how do you capture the knowledge? 

You have to find out why they don't like to write, and then find some way to address this barrier. 

  • If there is a cultural barrier, such as fear of exposure, perhaps you can keep the written knowledge confidential among the practitioners)
  • If there is a lack of writing skills, perhaps you can use knowledge champions or lessons facilitators to do the documenting
  • If there is a lack of language skills, maybe you ask people to capture knowledge in their native language, and translate it later
  • If they don't feel they have the time to write down knowledge, then it is a question of asking management to raise the priority of knowledge work.

If you are going to build a robust KM program, then you need to get people writing, as well as just remembering.

No comments:

Blog archive