Hi, Max. What you're describing is a cooperative tool based on formatting tags. While this does work for small groups, it fails the requirements for an effective tool in four areas.
First, its basis on formatting tags requires a learning curve. Although the wiki-savy have become accustomed to formatting tags, most people cringe at the thought of sorting them out. They may seem simple, and truly they are simple, but they still take a little time to get used to. This creates a barrier for entry for users, both in reading and entering new ideas.
Second, the nature of a cooperative tool is that everyone puts in a bit of effort in order to get things moved to where they're supposed to be, prioritizes things, deletes garbage, that kind of thing. This essentially requires a moderator - someone with enough time to oversee things. Moderators are usually pretty unbiased, but even obvious bias can be hidden simply because something has to come first, so it might as well be the moderator's stuff. No matter how dedicated a moderator is, he's only human.
Also, as has been demonstrated in discussion after discussion on Wikipedia, a large volume of comments can readily swamp a human moderator, or even a group of them.
Third, forums suffer from the recency effect, and wikis suffer from the inverse. Nobody wants to slog through a thousand lines of argument in order to find the good stuff, so whatever's on top generally gets priority for how people think about it. My tool avoids this by segmenting all of the information and having it reorder itself through heuristics, usage statics, and advanced voting algorythms. In addition, it's designed to display pro/con(/neg) arguments in separate columns so the users don't get any one side's arguments ahead of the other's. (neg comments are those that state that a supposition is unprovable).
Fourth, the tool lacks any kind of rating system for concept validity. It's nice to believe that all concepts should be considered freshly by each individual, but that leaves a page where "everybody should worship my god" stands toe-to-toe with "there are 12 inches in a foot". If we're actually going to provide information to people it's important to also provide them with some measure of that information's validity.
So for small fixed groups arguing limited topics your system can work just fine, but it isn't scalable for population size or concept complexity, it doesn't allow for a dynamically changing population, and it provides limited facilities for the tracking of the reasoning behind the group's conclusions.
Steve, there was an implementation of Max's tool, minus his tagging structure. It was effectively how this Wikipedia page evolved:
http://en.wikipedia.org/wiki/Arguments_for_and_against_drug_prohibition
This was one of the things that I was looking at while coming up with some of my tool's more advanced features. It started out as a tedious argument on the original War on Drugs page, then moved to the arguments for and against page, then was moved to the discussion of that page, then was pared down through the intervention of some enlightened moderation. In the final revision, the most senseless argument stands toe-to-toe with the most supported by fact, and it takes intervention by an intelligent agent to examine everything and sort it all out.
On 4/3/06, Max Völkel voelkel@fzi.de wrote:
Hello Steve,
I have an IBIS-based "tool" for debates in Wikis.
Hi, do you have any publicly visible implementations?
Hmm, so you don't got my joke. The description how to structure the page, as shown in the email, *is* the tool. That's all one needs. I used it myself, found a solution to a problem. We also used it in a discussion between 6 people to structure the results of a face2face meeting and then continued the discussion in a wiki.
;-) Max
Wikitech-l mailing list Wikitech-l@wikimedia.org http://mail.wikipedia.org/mailman/listinfo/wikitech-l