Stigmergic Web

Back in mid August I gave a brief presentation on DIY eLearning systems, and amongst the many helpful and provocative comments Leigh Blackall got down to business: Tell us how it went then? Where to from here?

OK, Leigh – its a bit late, but here’s my report. The first part is easier to answer. Most of the other presenters were from the Computer Science field, and talked a lot about ontologies. My background is in education and, long long ago, in genetics – I know jack about ontologies. Luckily, a quick referral to Wikipedia cleared things up. An ontology, within computer science, is part of the product of an attempt to formulate an exhaustive and rigorous conceptual schema about a domain. Ontologies are useful, I suppose, when attempting to create a model of a phenomenon, but I suppose there is a danger of mistaking the map for the territory. And ontologies don’t seem, from my reading of Christopher Alexander (A City is Not a Tree) and Clay Shirky (Ontology is Overrated: Categories, Links and Tags), to scale well to large complex systems. Shirky, in particular, shows that ontological, tree-like hierarchies work very well with specific restricted domains, such as the periodic table in chemistry, but falls short when trying to inform us about the classification of knowledge.

So amidst all of these people talking about ontologies, I felt kind of out of place with my nice little eLearning Processes Using Small Technologies Loosely Joined presentation. Despite my misgivings, I received some really nice feedback, not only from the educational technology types but also from the computer science guys. What impressed me most was the way that both perspectives complement each other to look at some of the problems of eLearning. I found myself challenged by trying to understand the other discipline’s perspective.

Now the tough part – what next? Stephen Downes recently presented on Principles of Distributed Representation (powerpoint and mp3 available for download), and he points out that some of our traditional beliefs about knowledge are inadequate. Speaking of knowledge, he specifially mentions five characteristics of knowledge that are downright counterintuitive:

• Knowledge is subsymbolic
• Knowledge is distributed
• Knowledge is interconnected
• Knowledge is personal
• Knowledge is emergent

Read Stephen’s transcript for the full details, or listen to the audio. The bottom line is that our tools for teaching and learning need to match the characteristics of knowledge. I happen to think that a lot of the small technologies I spoke of at the symposium do just that.