Imparting knowledge through storytelling, Part 2

emergent properties of collections of stories that have been developed at IBM’s Institute for Knowledge Management.

In addition, the representation of stories must be much richer than simple text. We’ve already seen how multimedia could be used to represent story components as well as directly present the story itself. And the representation of the retrieval process of stories would also be much more powerful if it went beyond the simple text listing of story titles, dates and a brief description.

In fact, storytelling retrieval systems could be good candidates for new computer interfaces such as those envisioned by David Gelernter and others as they look beyond the standard desktop metaphor of files and folders.

Imagine a browser interface into a library of stories. It might consist of a graphical map of a world. The world could include countries (communities), cities (stories clustered around communities), natural resources (stories clustered around categories of various types), roads or connections between cities that could be customized by the user or the community to which the user belongs. In addition, there might be actors moving around the map performing actions--such as mining some resources, exploring the connections between stories or even representing a particular storyteller.

Double clicking on a city would open up another display that might show a structured set of resources, charts and spreadsheets, and a decision panel that is personified as the governor or librarian of the place, an avatar who could offer help, information, and act as an agent, launching searches out into other information spaces and digesting the results according to rules the user or community could set up.

Of course, one response to that image might well be, why on earth would we go to all that trouble, setting up complex graphics, etc., when simple interfaces can do the job? As we saw in earlier discussion, the answer lies in the way the brain works--we remember things better when more than one sense is involved. In fact, we do most sensory tasks better if more than one sense is involved. That is particularly true when it comes to stories rather than simple informational documents.

Another future direction might be to use stories or narrative schema to improve learning in a variety of situations even including ones that are not really stories.

As one researcher on the IBM Story Project put it:

“Narrative schema may be applied in many situations that are not story related . . . It seems that narrative schema is an option in processing data even when there are no human characters or the events are essentially nonsense . . . We may see stories in random material, information and data where stories do not otherwise exist.”

One example in which story elements are already in use is with the use of personas to organize information about certain key common elements. Rather than ask users to wade through a series of specific questions about what information they need (or typically need), simply ask them to find the closest matching persona. Personas are already in use in areas like financial services and computer sales sites.

There are many directions that storytelling in a knowledge management environment could take. However, underlying any new direction should be a rich and powerful knowledge architecture. Without that architecture, storytelling will likely continue to languish either in abstract academic research white papers or hidden in the undiscovered byways of personal interactions within corporate communities, and knowledge management will miss the opportunity to extend its scope and depth by incorporating one of the most heavily used knowledge transmission mechanisms in corporations today.