InformationArchitectureOfThisWiki

ThoughtStorms Wiki

New summary of IA features

In a live version of Cardigan Bay there's full text search : CardiganBaySearching

Originally ThoughtStorms used CategoryCategory convention. Easy to search in a live version. Now you can go to the Category page and see all the backlinks.

ThoughtStorms also has a lot of LocalIndexes

Many pages have Compare, Contrast and See Also sections at the bottom.

Compare tends to be for similar things, Contrast for rivals / opposites. See Also for unspecifiedly related things.

There are WarpLinks which are like See also, but where I think I'm making a more "creative" connection.

There are BonusFreakyConnectionPoints for really ... er ... freaky or strange parallels / coincidences.

As for pages, there are ThoughtStormsPuns and CreativeMistypings which have spawned strange new ideas.

Finally, there is still RecentChanges to see what was edited recently

Previously

Some thoughts on the evolving InformationArchitecture and the moves from ConcreteToAbstract here.

Early categorizarion reflected the areas I was interested in : InternetCultureStuff including wikis, weblogs, free software etc. PhilosophicalStuff mainly about Popper. Some PoliticalStuff. Some ProgrammingStuff. Some stuff about OrganicArchitecture. And some pages about my MusicalStuff.

However, from very early on I was interested in making comparisons between these different areas, and noting the apparent similarities of process, forms and principles. My first attempt to talk about this was on the page TopicsDiscussedHere, which is still a good overview of some connections. But I soon realized I wanted to pull out some of the principles into pages of their own. These became the ThemesOfThisWiki. Themes, like FinishednessThinking and OnPlatforms, cut across categories such as music, architecture, software design and economics.

Another activity is to abstract certain formal properties of the discussion. Some pages were marked DevilIsInTheDetails, which was meant to warn the reader that while I believe I can see something going on, hey! I'm not stupid, I know that there are all kinds of countering forces, and all kinds of complexity which mean this might be a totally bogus prediction. It's not really a useful navigation aid, but if you wanted to see some of the things I felt pretty insecure about you could always do a reverse search on the name.

A more recent example of this, I noticed two contributers who'd written that I was setting up a false dichotomy. Why not start an index page called FalseDichotomies to link to all the "is this a false dichotomy" arguments? This is abstracting yet another quality which is neither category nor theme. Something more like ArgumentStructure. But maybe this itself is an example of something more general. Perhaps CounterThinking is also like this. And I've started collecting an index of TradeOffs which may another example.

In fact, now there are several of these higher-level or structural elements among the ThingsToSearchForOnThoughtStorms.

With all these different kinds of abstractions which could be done, CategoryCategory tagging of the pages looks cumbersome. New indexes are like new queries on the data. (Although created manually.) You can't expect to prejudge these questions in the data of the page.

Another thought. Wiki's network structure makes it clear that abstract categories are really nothing but usefully connected nodes. (So maybe I get BonusFreakyConnectionPoints for linking to NetworkEpistemology).

In fact BonusFreakyConnectionPoints are another kind of thing going on here. They started as a joke, reminiscent of MorningtonCrescent, but now I use them as a way of flagging (and justifying) long, imaginative analogical links. I also created WarpLink to flag less inspired, but possibly interesting, long links (and weak pun-based links.) I also started CategoryThoughtStormsMachinery to document these innovations.

The point is, that ThoughtStorms is partly about the continuous negotiation and refactoring between concrete and abstract. And that negotiation requires fractional, intermediate degrees of abstraction : not just "types and tokens" or "classes and instances" but something messier : wild conjectural pulling out of anything that might be a possible commonality between pages and might grow into a useful generalization.

See also my discussion with Zbigniew on TheArchitectureOfComplexity

(This last para copied to FractionalAbstraction)