Sometimes you can get a group of smart people together. Why limit what they can talk about ? Let them rant on about whatever is on their mind. Let them discuss things with each other. Then later categorize and organize what they said. BigBucketsFirst.

If you can’t get these people together in the same room, perhaps you can convince them to use a wiki. But which wiki ?

No idea what they are going to talk about

If you don’t know ahead of time what they are going to talk about, perhaps use a “ReserveWiki”.

(moved from TooFewWiki)

We can set up reserve wiki, - wiki with names like “wiki1, wiki2, wiki3,…” that people can use if they want to have a wiki. They can personalize it on the interior. We can leave a page behind that reads: “E-mail admin at wherever dot com to have the domain name changed over, and the wiki customized.” Yah! – LionKimbro

Only a vague idea about what they are going to talk about

Crazy idea: set up such “reserve wiki”, but give them Dewey Decimal numbers – since every subject that can be written about can be wedged into that categorization somewhere (unfortunately, often in multiple places) … Then everything that can be written about would be on-topic in at least one of the 10 top-level Dewey Decimal wikis, unless of course that one specifically delegates it to some other discussion group elsewhere where it is even more on-topic. – DavidCary

Librarians are abandoning cataloguing systems in favour of client-oriented indexing, such as Google and recommender systems (a la Amazon, which is a pretty amazing system if you think about it). Cataloguing systems suffer from the perfect ontology myth, the politics of choice and no-choice, an imperfect match, and a cynical view of the client. Dewey is particularly problematic, as it reflects a really old school perspective of How The World Is, like the separation of Science from Literature. The Library of Congress system has its own problems as well. – SunirShah, studying library science (in part) at the University of Toronto

Well that stinks! Can’t we have both? Hierarchial classification can be really useful (I can’t wait until one of these open classification projects for the web takes off, no thanks to “open” directory, grr..). I guess I see that point that there is no need to assign a number to each subject classification anymore. I’m gonna be annoyed if someday I go to the library and there is no way to ask for a list of material on a subject matter aside from keyword search or a recommender system.

“Metacrap” by Cory Doctorow comments on the futility of getting everyone to agree to a classification scheme.

good read, thanks. Of course, the thing to do is to live with multiple hierarchial classification schemes, and unreliable metadata, rather than throw out the concept altogether.

pre-classifying all possible wiki

(moved from DavidCary)

Here’s an analogy that occurred to me. It’s still a little fuzzy, but maybe just jotting it down here will help me understand. I wouldn’t be surprised to find out that someone else had said this ever so much better than this.

Everything that humans talk about can be described in words. If there isn’t really a good word for some thing / action, we typically start out describing it in some long, complicated phrase. (I suppose we could just not talk about She Who Must Not Be Named, but where’s the fun in that ?). Eventually, if we talk about it a lot, we abbreviate it into a single word – either a completely new word ( neologism ) or we stretch the meaning of some old word to cover this new thing / action. (Ever looked up “computer” in a pre-1920 dictionary ?)

All the text and images on every page of every book and every magazine that has ever been printed could (technically – let’s ignore CopyRight issues for the moment) be copied into various wiki. Some pages aren’t really on-topic for any of the wiki that existed in 2003. We have the same choices as before – not talk about it, or make a new wiki to stick it into, or stretch (CommunityWiki:EnlargeSpace) some wiki to make room for it.

Imagine a future where all that text and those images have already been wikified. Someone could simply read the Dewey Decimal Number off the spine of those books and create a Wiki:RoadMap. Imagine checking out a book from the library, then when you’re done thinking “Hey, that was really cool – I’d like to talk to other people who also read this book, so we can put into practice those ideas”. (Or thinking “This author seems intelligent, but obviously knows nothing about X or any of the books by Y. How can I tell the other people who read this book ?”). One way of contacting those people (in this utopian future) would be to read the DDN off the spine of the book, check the DDN roadmap, and jump directly the the relevant wiki.

The amazing thing about URIs and WikiNames is that the thing they link to doesn’t have to already exist (unlike the things listed in the Bibliography of a book or report). I could build this “Dewey Decimal Number roadmap” today. It should “cover” everything that humans can write about, right ? “cover” not in the sense that a single wiki contains all human knowledge, but that it points to some of the appropriate places, and it helps find a place for any particular bit of knowledge anyone wants to stick in.

Having a single person responsible for all human knowledge – maybe that’s not such a good idea. But perhaps creating 10 wikis (for 000, 100, 200, 300 … 900 series DDNs), we could find 10 people willing to be responsible for these wikis ? When too many people are posting in some area, then make children wiki and point to them.

I see that CommunityWiki:BigBucketsFirst has a related idea. Start with everything, then divide. I seem to see similar ideas in a few other places – Buckmister Fuller, and B-trees.

I did a quick web search to get the actual list and … what ? OCLC claims that “All copyright rights in the Dewey Decimal Classification system are owned by OCLC. Dewey, Dewey Decimal Classification, DDC, OCLC and WebDewey? are registered trademarks of OCLC.”

Also see “May I use the DDC to organize information on my Web site?”

I’m surprised; I thought this was public-domain information. has a nice copy of the system, but seems to claim that it’s a redundant copy that will eventually be deleted – I can’t seem to find the other copy. ?


I love the idea of categorizing wikis according to some well-known scheme used in libraries. The question is: How many libraries actually use the DDC? I can’t imagine every community library in Switzerland having a license agreement with OCLC. I googled around a bit and found the Library of Congress call numbers [1] [2], or the “Bielefelder Systematik” [3]. I think that in Germany there is no common system, because I found other libraries using a different system. It seems that various systems are in use. [4] [5]

What does that mean for us? At the moment I think the Library of Congress classification looks like the most likely candidate. I also remember faintly that many or all of these classification systems are inherently culturally biased. Do we want to use them nevertheless?

I think it’s only the current, up-to-date editions of the DDC which are copyrighted. Using an older, public-domain version should be fine. OCLC also might have a trademark on the name Dewey Decimal System, but that is easily avoided.

Still, why not go with the LOC, given the problems with DeweyDecimal?.

Inherent cultural bias is probably unavoidable, so it shouldn’t stop us.

This project should be part of SwitchWiki.


Everyone can make his own list of wikis, classified in his own culturally-biased way.

Yes, I’m not at all hung up on DeweyDecimal? in particular. Thanks for the links to the Library of Congress ( ) call number system. Maybe I’ll start with that. (Is it public domain ?)

What are the “problems” with DeweyDecimal? ?

DavidCary 2004-02-03 19:18 UTC

I meant the potential copyright problems with using a recent version of it.

I assumed the LOC was free, but maybe not:

I know that US government works can’t be copyrighted under US law. But, this page asserts that the LOC headings are copyrighted, “except in the US”. So, maybe the US has copyright treaties with other countries that make it copyrighted there even though it’s free here.

Anyways, it seems that in order to access the info, you have to agree not to distribute it. Maybe one could walk into the LOC itself and xerox some pages to get the data out without agreeing to anything like that. I have no idea what the legal situation would be if a US citizen like myself would get the data, produce a derived work, and then give it to someone outside the US. Would my derived work be free of copyright restrictions (because I derived it from something that was public domain), or hindered (because the source work was copyrighted outside the US?)?

Sounds too confusing for me. Someone would need to consult lots of lawyers and make lots of calls to the LOC, it seems. Although, perhaps the LOC would voluntarily give up the data for use in an international project, once the copyright problems were explained to them. Who knows.

Probably simpler and more reliable (in case someone at the LOC granted permission but there was a later dispute over whether they had the authority) to create a new system from scratch with a ShareAlike or GFDL license. Big project, though, and it should probably have some experts working for it.


I see that (also see Wikipedia:Dewey_Decimal_System )

claims that “the Dewey Decimal System is copyrighted; the one created by the Library of Congress is not.” but then claims that any version of the Dewey Decimal Classification published earlier than 1923 (like everything published earlier than 1923) has now passed into the public domain. (“the DDC has been around since 1876”).

I know precisely what topic we are going to talk about

Use WorldWideWiki:SwitchWiki or the WikiNode network to look for wiki on that topic. If one exists, great – use it. If not, use the closest match, or use ReserveWiki. Later if/when a wiki on precisely that subject is set up, MobileContent the group and their pages over there.

OK, my group has talked for a long time, and I think I finally know exactly what we're about.

Great. If there already exists an adequate wiki on a close-enough topic, use MobileContent to move your group there.

Or rename the reserve wiki, creating an entirely new wiki focused on precisely the topics your group is interested in.

Then leave a reserve wiki behind for the next group.

Should I rename this something like “UnnamedWiki?”? “PeopleFirstThenName?”? WikiWithoutNames?(Yet)?

I like ReserveWiki and UnnamedWiki?. Both have drawbacks: reserve implying that you create them ahead of time, which is not a prerequisite, and unnamed implying that it doesn’t have a name, which isn’t necessarily true, since you can have a name and still lack a purpose/goal/topic.

This is cool, and yes I did have this in mind for SwitchWiki. I like the idea behind ReserveWiki but mostly I like the idea of naming there wiki - say NumismatistsWiki? on a page and tag it with ReserveWiki to denote that it is looking to move to a wiki of its own. :-) Best, MarkDilley

Did I not say this was a crazy idea ? I didn’t expect someone to go ahead and do it.


See PlatonicCategories


CategoryIntroductionToWiki ?


Define external redirect: WebDewey PeopleFirstThenName NumismatistsWiki WikiWithoutNames UnnamedWiki DeweyDecimal

EditNearLinks: EnlargeSpace CopyRight SunirShah WikiNames ShareAlike SwitchWiki