User:The Unknown Rebel/SpiritWorldWiki

From Meta, a Wikimedia project coordination wiki
Jump to: navigation, search
Note: The creator of this essay recognizes that it basically needs to be rewritten from scratch and streamlined for easier reading...Also, any help you can provide with the flowcharts is welcomed...I can give you the Visio file.

I have been considering some possible technological methods of tying different wikis together in a way that will render somewhat irrelevant the inclusionist/deletionist debate on Wikipedia. Currently, Wikipedia's dominance in the realm of MediaWiki-based encyclopedias lends its judgments on what subjects are and aren't notable undue importance. I believe this one monolothic entity is no longer sufficient to meet the world's needs, and it is time to switch to a new paradigm. Specifically, I propose a wikisphere of independent but interconnected wikis that together can meet the public's need for information, including the more obscure subjects that Wikipedia is not willing to cover. Some creative software and web development will be needed to make this successful, however.

Why Wikipedia alone is not enough[edit]

The word "encyclopedia" means "comprehensive" or "broad in scope or content." Wikipedia Founder Jimbo Wales once said, "Imagine a world in which every single person on the planet is given free access to the sum of all human knowledge. That's what we're doing." Yet, it is clear that Wikipedia and its related projects (WikiNews, WikiBooks, Wiktionary, etc.) do not aim to be encyclopedic in the sense of encompassing all knowledge. Even verifiable information must meet notability thresholds established by the community, or it will be removed. The Economist recently ran an article on this matter.

Yet, it is clear that there are editors able and willing to write about subjects deemed non-notable by Wikipedia, and the spread of wikis covering niche subjects (e.g. Homestar Runner wiki, Memory Alpha, Electorama, etc.) suggests that there is a demand for this coverage. The disadvantage to having articles spread over all these separate wikis is that some of the benefits of a unified system are lost. One of the great advantages to an online encyclopedia is the ability to easily cross-link articles. Although interwiki links are possible, it is not particularly common for a Wikipedia article, for instance, to link to an article on another wiki (and many editors believe it is unwise to do so). There is also some duplication of effort as editors on different wikis create parallel articles on the same subject. Further, as the largest wiki in existence, Wikipedia typically has higher visibility than specialized wikis, generally appearing higher in Google search results; and incoming wikilinks can further increase traffic to an article from people already on Wikipedia. All of these factors make contributors want to concentrate on Wikipedia; and those with a fondness for relatively obscure subjects often fight for their article's inclusion there, and are disinclined to settle for less.

Presently, Wikipedia is somewhat of a monopoly in its niche. While anyone can download Wikipedia's 2,000,000 articles, there are significant barriers to entry for anyone seeking to attract anything on the scale of Wikipedia's 6,000 editors. This is evident in the fact that, as this article notes, startups such as Citizendium, and various forks, have had trouble competing head-to-head. They couldn't "catch up" to Wikipedia's momentum, partly because of the problems of duplication of labor on articles on the same subjects being constructed in parallel. But as any free market economist will tell you, there are no true monopolies, as there are always competitors – in this case, other sources of information. There just is not something else of that nature that matches Wikipedia in both comprehensiveness and mass-collaboration. But if a way could be found to simultaneously duplicate those benefits through other means, then the monopoly would be broken, and Wikipedia's systemic and political problems would be rendered largely irrelevant.

Example illustrating shortcomings under the current system[edit]

The following example will show some shortcomings in the present system. Suppose that Bob, a Star Trek enthusiast, creates an article about a new game called Star Trek Quadrillion that was mentioned in Star Trek Magazine as being in the design stages. He writes a rather enthusiastically-worded paragraph about the game without citing sources, thinking to add references later. The article is immediately tagged for speedy deletion as spam, and before Bob has a chance to act, it is gone. He receives a message saying that Wikipedia does not allow advertising in articles.

Somewhat peeved, Bob writes an indignant note to the admin saying that it was not spam. The admin suggests he read more about about Wikipedia policy, which he does. Bearing in mind the neutral point of view and verifiability rules, Bob rewrites it from scratch in a more moderate tone and cites his source. It is immediately nominated for Articles for Deletion.

In the AfD, several users immediately pile on, voting to delete as "cruft" and so on. Taken aback, Bob begins looking for more sources, becoming slightly upset at some of the responses, and concerned about finding a way to turn the tide before the 5 days are out, and the AfD closes. He makes a number of arguments that seem reasonable to him, such as "You have an article about Star Trek Trillion; why not this one?" He receives a flurry of responses referring him to WP:OTHERSTUFFEXISTS and a slew of other obscure abbreviations. As improvements are made to the article, a few "weak keep" votes begin coming in toward the end, but it is too late and the article is deleted. Under Wikipedia's Criteria for Speedy Deletion, if the article appears in the near future, it will be speedily deletable as recreation of deleted content.

At this point, Bob is not particularly happy with what happened, and considers his options. He can post it to Memory Alpha, a wiki specializing in Star Trek-related articles. However, because of the copious wikilinks that he has, by this time, added to the article, linking to software development subjects and game designer biographies that exist on Wikipedia but not on Memory Alpha due to the latter's more specialized focus, those will appear as redlinks pointing to articles that will never exist on that wiki. Yet, he is not aware of a wiki on software development, and if he were to post it there, then the wikilinks to many of the Star Trek-related subjects would turn red because of the limited focus of that wiki. Due to smaller article base of the specialized wiki, the article will also likely suffer from fewer potential incoming wikilinks; this, combined with the smaller editor base, may also lead to relative neglect from editors. Editors on Wikipedia, for instance, creating an article on Star Trek games in development, may have a redlink to Star Trek Quadrillion (perhaps anticipating its recreation at Wikipedia), but are less likely to have a link to a Memory Alpha article; instead, they are more likely to source their Wikipedia article to the primary source, e.g. the article in Star Trek Magazine. Moreover, because each wiki has its own Watchlist, Wikipedians would likely not be watching changes to the Memory Alpha page, and talk page discussions relating to it, unless they also had an active account on Memory Alpha, and were checking that wachlist regularly.

Technological solutions[edit]

Wikipedia appears unlikely to make radical changes in its notability thresholds; indeed, it is conceivable that they may become more restrictive, not less. But the need to collaborate more effectively on relatively obscure subjects outside Wikipedia's purview remains. I believe that creative technological enhancements are the solution. There are many possible approaches, but here are a few.

A partial solution might be to create a customizable interface for viewing wiki pages that would show wikilinks to other wikis according to the settings in the user's profile. For instance, if the profile is set up to include Memory Alpha, then as the user is browsing the list of Star Trek games in development on Wikipedia, certain items such as "Star Trek Quadrillion" might have a small icon next to them indicating the availability of an article on the subject at Memory Alpha. This could be implemented, for instance, through a Firefox plug-in, through another website slightly analogous to that would parse the sites' content, or through Mediawiki modifications. However, I prefer the path that can be done independently without relying on the cooperation of individual wikis, and therefore am more inclined toward the first two choices, especially at first. The interface might sense the length and quality of the article (perhaps by how its WikiProject had rated it), which might be reflected somehow by the icon. E.g., if an article on the same subject exists at two different wikis, the user might prefer to read the one that is four times longer than the other.

As an alternative to the above idea, there is also a potential rudimentary way to implement interwiki links. Within a particular Mediawiki, wikilinks turn red if the article exists and blue if it does not exist. However, external links to articles on other wikis (using urls) do not change colors. This could be remedied by getting a list of all the articles on that wiki as of one moment in time, and updating that list regularly using the New Pages log and the Deletion log. That list could then be used to turn external wikilinks blue or red depending on whether or not the article name is on the list.

It would be good to have an easy, fast way to search across wikis, and find out if an article is already in existence on one of them; and in addition, allow users to have an integrated, customizable watchlist that can detect changes made to watched articles on many wikis with one click. This might be integrated into the customizable interface mentioned above. The power of metasearching has been demonstrated in many other contexts. For instance, Dogpile is a metasearch engine fetching results from Google, Yahoo, Live Search, etc.; searches many encyclopedia and other sources; and searches,,, etc. for job listings. The difference is that this search would specialize in searching wikis. Wikia Search represents one attempt to have a cross-wiki search, but its development has progressed rather slowly, and it is still in the alpha stage. This tool is a watchlist that works across Wikimedia wikis.

In this revised wikitheology, articles may start in a shadowy world of pre-existence, go to Wikipedia, die (get deleted), go to the SpiritWorldWiki, and then from there go to any number of wikis, including Wikipedia; or if they are truly bad, they will permanently go to article hell (deletion even from SpiritWorldWiki, if they consist only of illegal content). Multiple cycles of article reincarnation are possible.

Lastly, I propose creating a wiki for articles to be placed when it's not clear what wiki they belong on. For lack of a better word, I will call this SpiritWorldWiki, after the place in Mormon religion where new souls reside before they are incarnated in a specific place on earth; and where they go upon death, before being sorted into the Celestial Kingdom, the Terrestrial Kingdom, or the Telestial Kingdom. New users not familiar with all the different specialized wikis out there can do a quick search and then start a new article on the SpiritWorldWiki. A more experienced user can then come along, and suggest a wiki to move it to; perhaps moving it himself if he's reasonably certain. If an article appears suitable for inclusion on Wikipedia, it can go there immediately; however, it is missing sources or seems spammy, it can be reworked on the SpiritWorldWiki until it is ready to transwiki. This could help prevent some speedy deletions, AfDs, etc. Some articles might remain on SpiritWorldWiki for rather long periods; for instance, Star Trek Quadrillion might remain there until it gets to a suitable point in development to where it is considered notable; or it could be moved to Memory Alpha. Meanwhile, collaboration on improving it could continue.

SpiritWorldWiki would also serve as a repository for articles rejected by other wikis. A dump of Wikipedia's entire database could be obtained, and all recent changes logged continuously. In the event of deletion of an article from Wikipedia, the most recent version before deletion would automatically be spawned at SpiritWorldWiki. It could then be moved by editors to a more appropriate wiki, or remain on SpiritWorldWiki for further collaboration. A site called Wikibin and another called Deletionpedia already make some deleted articles available for viewing, but they do not allow editing of those articles.

In accordance with the pure wiki deletion concept, nothing except copyvios, libel and other illegal content will ever be deleted from SpiritWorldWiki, although clearly useless articles (e.g. "Trevor is gay") can be reversibly blanked, with the page history remaining viewable. The idea is that articles not included anywhere else will go here by default; and that doesn't hurt anything, because we're not putting ourselves out there as being encyclopedic. There might even be Hoaxwiki and a Spamwiki (again, probably not by those names) for hoaxes and advertisements to be permanently placed, although the contributors would be encouraged to voluntarily categorize their contributions as such. Although this might seem silly, it could be useful to catalog this material to aid in identifying it if it appears elsewhere. We might have wikis for garage bands, ordinary people, etc. Whatever there is verifiable information about is fair game - and there could be outlets for the less verifiable stuff as well. And if the wikis can all be tied together, it might be worth the time to have such wikis. Certain wikis, rather than having an NPOV policy, might be dedicated to promoting causes. As wikis proliferate, communities sharing certain attributes might form, which would allow users to "subscribe" to whole categories of them, including them in searches. Together, all these systems would implement the concept of a "wikisphere," similar to the blogosphere in which different blogs are connected and cross-linked in a collaborative way.

The culture of SpiritWorldWiki would probably be laid back. My vision of it is that free rein will be allowed as to user activites, as it will also be a repository for meta content and userpage content that was not allowed on other wikis. Social organizations in particular would be allowed, and this might include those expelled from other wikis, such as Esperanza. Time will tell how good an idea this is, but I think it is good to start out being liberal. Experimentation with new ideas will be encouraged, as SpiritWorldWiki will not worry too much about its reputation, due to it being partly a dumping ground for the rest of the wiki universe (or perhaps it would be more accurate to call it a combination of a nursery and recycling center).

It is conceivable that different versions of articles existing in different wikis could be reconciled through merges or using Subversion branching.

Continuation of previous example, illustrating how SpiritWorldWiki would work[edit]

Let us pick up, then, from our previous example, in which Bob's article has just been deleted from Wikipedia. SpiritWorldWiki's system has already noticed its deletion from Wikipedia and automatically spawned it in the SpiritWikiWorld mainspace, with a template noting its past history. By the time Bob get there, other editors who noticed its arrival while on New Pages Patrol have already begun discussing on the talk page the best wiki for it to be moved to and what changes should be made to improve it. They join a few Wikipedians, whose attention was attracted to it during the AfD, in resuming work on improving it. Someone nominates it for the SpiritWorldWiki "Did You Know?" and it attracts yet more attention. A couple weeks later, SpiritWorldWiki's system notices that a new article by the name Star Trek Quadrillion has been created on SpaceGameDevWiki, one of the wikis that it continuously scans. Accordingly, a link to it is automatically placed on the Notifications page for the SpiritWorldWiki Star Trek Quadrillion article. A few editors who have it watchlisted pick up on that, and begin talking to the creator of that article; they end up merging the two articles.

A year later, the actual computer game called Star Trek Quadrillion hits the market and the article's editors, believing it now to satisfy Wikipedia's notability standards, transwiki it there, in a much improved form over how it looked when it was deleted from Wikipedia. Throughout this time, the process of article improvement has continued uninterrupted.


With SpiritWorldWiki in place, the focus is on learning from the AfD, working toward the future success of the article, and finding it a good home. This is a far cry from the current system, in which people often get highly emotional about AfDs, because the success of their article is at stake and where they can easily get discouraged if things don't go their way. Many Wikipedians have left the project for such reasons. But perhaps there are solutions that can make everyone happy. It won't matter so much if an article gets arbitrarily deleted from Wikipedia if its inclusion on another wiki would be rendered just as good, because of enhanced interwiki cross-linking; and if work on it could continue in preparation for it eventually going back there anyway.

It is worth remembering that Wikipedia itself represents a technological solution to the age-old problem of figuring out what topics are notable enough to include in a paper encyclopedia. Now the problem is at least partly political in nature - Wikipedians legitimately disagree over whether certain niche content should be included in what aims to be taken seriously as a reliable encyclopedia. Some want to stick to the basics, while others want the benefits of a more inclusive source of information. Once again, it appears that the way to make progress is through additional and/or enhanced technology.

I welcome any comments on these ideas. Also, if you are interested in collaborating with me in implementing it, feel free to contact me on my talk page. I hope to get started soon.


Notice that the process can move in a circular pattern.



See also[edit]