Wikis and knowledge structures

And now, a mostly new riff. It doesn't quite capture what I want to say right yet, but I'm hitting Send now and hope it grows up (perhaps on the conversations wiki! :-))

(Jordan asked, "Do both wikis and science require ontologies to structure knowledge? Why?")

We've been grappling with the need -- or not -- for a taxonomy in a Massive Wiki. This got in our face early on because Obsidian, our primary MW client right now, has a very nice file-oriented page navigator built into it, which lets you do hierarchical taxonomy the way you've been doing it since you learned how to do computers, however long ago that was.

Turned out that hierarchical taxonomy is sort of a blind alley -- or a crutch, that you need if you're storing physical things like paper and books in physical things like file cabinets or libraries.

What I've been saying to MW folks lately is that relationship (context) is how you find, remember, and communicate with your past self / future self, and with other people. And, that relationship, for humans, is also contextualized in conversations and stories.

An ontology, or better, ontologies, is/are needed to structure knowledge. Ontologies work for knowledge the way grammar works for languages. Without a grammar, the way words tumble out of your mouth may or may not make sense even to yourself; they probably won't make sense to someone else.

And surprise, I think if you and someone else stuck on a desert island chattered to each other long enough, you would start to make sense to each other; but, in doing so, you would have instantiated a grammar that both of you can understand. Similarly, if you and your desert island friend started keeping knowledge in a wiki long enough, you'll start to make sense; and in doing so, you will have instantiated an ontology.

Neither of you might be consciously aware of the grammar or the ontology you've elicited out of the chattering, but I assert that even if you can't articulate the shape of it, you have a grammar or an ontology.

A taxonomy is less important, unless you're dealing with massive amounts of things that need categorization and hierarchy, like species of living things.

If and when you start a taxonomy, you've started to create limits and boundaries to the way you can perceive and reason about the things you're working with. So, it's good not to over-taxonomize. It may help to use more than one taxonomy. (In a wiki, it's easy to fit things into multiple taxonomies at the same time, unlike pieces of paper or books.)

But in general, think about the way things are connected, and how to tell stories and make conversations around that. That's how humans really work under the hood.

And more: how do you frame up the knowledge you have and want to grow?

In conversation, in real time, and creating artifacts (wiki pages, diagrams, terminology, ontology) as you converse.

It seems that you want at least 3 people in a conversation.

The act of explaining to each other what you mean, as you create or morph the artifacts, helps you know what to build. And you just fractally scale up from there.

Humans are pretty good at doing this natively, without much training or experience. But, like anything else, it helps if you've got some participants who do have experience in organizing and sensemaking; you can save a lot of simple mistakes and rework that way. As long as you don't turn over the whole process to experts; it's important that many people, with many backgrounds, participate.

Related posts: - Ontology (back to basics)