The Four Hops: why knowledge capture and transfer isn’t easy
One of the challenges with knowledge capture and transfer (drawing the knowledge out of one person’s head and transferring it to another’s) is the inevitable mismatch, misrepresentation, and miscommunication that occurs along the way. The figure below illustrates “The Four Hops” over which these disconnects can occur.
On each end of the spectrum, you have a knowledge source and a knowledge recipient. Each has their own, highly personalized way of thinking about the world: their mental model. An accountant thinks about and views the world differently from a medical doctor or an airline pilot. Another variation could be where the knowledge source or knowledge recipient might not be an individual, but rather an automated system or an organizational agent, such as a standards committee, a selection committee, or other decision-making body. This would imply the need for an equivalent of a mental model which would be applicable to all agent types – h/t Bo Newman.
And you shouldn’t only be looking at the differences. There are similarities as well. Some different types of mental models are shown on the left- and right-hand sides of the figure.
What’s in a person’s head may not translate well into something on paper. We call these “mediating representations” (the second set of bullets from the left). These are attempts to capture the essence of an expert’s thinking and worldview. For domains or topics that are process-oriented, flow diagrams, workflows, etc., are the preferred choice. For domains that tend to be case-oriented or heavily situation-dependent, lessons-learned stories or case histories work well.
However, the representations used to draw knowledge out of the expert’s head may not be fully suitable for storing in a computer system so they can be easily accessed, either by search or navigation (see the middle column of the figure). At this stage, some sort of characterization and indexing scheme needs to be applied, along with business logic or other rules which spell out how the knowledge is best applied under certain conditions. It’s also extremely important to include a clear explanation of the reasoning behind the knowledge being represented – where it came from, how it was derived, what options were considered, etc.1
Once the knowledge is stored and accessed, it must be presented to the knowledge recipient in a way that’s clear and makes sense. The fourth column of the figure lists some sample modalities such as FAQs, simulations, games, etc. Note that any of the capture modalities listed in the second column from the left may also be used as a transfer modality, if appropriate.
Finally, on the fourth “hop,” the knowledge once again takes on the form of a mental model, as the knowledge recipient attaches their own views and perspectives to what’s been transmitted.
It pays to keep in mind that mental models may change over time. If so, the three intermediate steps (mediating representations, structured content, and on-screen dialog) need to change over time as well.
Finally, it’s important to note that every “hop” presents a chance for degradation. A slight misconception in the capture stage can grow into a complete disconnect by the time it enters the knowledge recipient’s mind.