Wikipedia and Other Wikis

Network effects & benefits of gritting one’s teeth & submitting to a Wikipedia’s rules, rather than using Wikia or one’s own site.
Wikipedia, technology, design
2009-01-272012-05-12 finished certainty: highly likely importance: 7

“You also raise an in­ter­est­ing point about spe­cial­ized wikis. Don’t they rep­re­sent a threat to the mass-mar­ket Wikipedia? In other words, as power Wikipedi­ans di­vert their at­ten­tion to the spe­cial­ized wik­is, this re­duces the at­ten­tion paid to the main site.”1

As Wikipedia takes a dele­tion­ist turn, mov­ing con­tent and con­trib­u­tors to spe­cial­ized wikis cer­tainly does seem like an at­trac­tive move. If Wikipedia (WP) does­n’t ap­pre­ci­ate you, take your efforts to a wiki which will! And if you start your own, you can be Ad­min­is­tra­tor, Bu­reau­crat, Stew­ard, and pol­icy writer all rolled up in one. But this is a rosy dream for most.

The fun­da­men­tal fact is: Spe­cial­ized wikis can com­pete effec­tively with WP, but only if the WP let them. It is WP’s game to lose. The other wiki can win if cer­tain con­di­tions con­trolled by WP are cleared.

The advantages

The ad­van­tages of do­ing all your work on WP should­n’t be un­der­es­ti­mat­ed: ac­cess to a large pool of com­pe­tent ed­i­tors, with low tran­si­tion costs - for ex­am­ple, it would be free for me to switch from work­ing on, say, me­dieval his­tory ar­ti­cles to work­ing on Star Wars ar­ti­cles, be­cause the meth­ods, or­ga­ni­za­tion, tem­plates, cus­toms and so on are mostly iden­ti­cal; be­cause my ac­count and rep­u­ta­tion fol­low me there; be­cause all my cus­tomiza­tions and per­sonal tools like cus­tom JavaScript or text ed­i­tors will be equally ap­plic­a­ble in the new area, and so on.

This is not true if the me­dieval ar­ti­cles and Star Wars ar­ti­cle are on en­tirely sep­a­rate wikis. A skilled WP ed­i­tor switched to an­other wiki will flail around: why is Me­di­aWiki act­ing this way, where did that op­tion go, who on earth ap­proved these pic­tures, and how do I get the ta­ble of con­tents back?

There are also net­work effects to con­sider - pretty much every ar­ti­cle sooner or later needs to link out of its spe­cial­ized area to more gen­er­al­ist ar­ti­cles. If the sep­a­rate wikis have in­ter­wiki link­ing set up, it is­n’t all that much harder to link to ar­ti­cles on other wikis than it is to link to an ar­ti­cle on the home wiki, but that’s an es­o­teric and not pop­u­larly well known fea­ture, and even at its best still adds fric­tion to the work. And there are many other sub­tler ad­van­tages hard to enu­mer­ate and ap­pre­ci­ate un­til one works on a small & non-Wikipedia wi­ki.2

Necessary conditions

So the only way a spe­cial­ist wiki can sur­vive or thrive is if it is on a topic Wikipedia re­fuses to cov­er.


I was in­volved in the Star Wars area of the Eng­lish WP for a long time, and it did well, un­til it be­gan to do too well: ed­i­tors from other ar­eas saw the de­tail and pro­fu­sion of ar­ti­cles in the SW area, and be­gan mov­ing to trim them down dras­ti­cally and raise stan­dards for the sur­viv­ing ar­ti­cles. In other words, WP in a very real & im­por­tant sense just de­cided to stop car­ry­ing SW ar­ti­cles.3

This re­jec­tion prompted most of the hard­core SW ed­i­tors to fork and be­gin build­ing (a SW wiki hosted on Wikia) us­ing the base of SW ar­ti­cles from WP. Now their re­source is so good that SW edit­ing is mostly dead on WP, with the ex­cep­tion of a few ed­i­tors like Deckiller who fo­cus on or­ga­niz­ing and re­mov­ing and im­prov­ing what’s left, and oc­ca­sion­ally bor­row­ing ar­ti­cles from Wook­ieepe­dia on the no­table new or miss­ing stuff. A sim­i­lar ex­am­ple hap­pened on the Pol­ish Wikipedia with rail­road ar­ti­cles4.

Sim­i­lar­ly, Wik­i­books works be­cause the Wikipedias refuse to carry de­tailed ‘how to’ man­u­als and ex­pos­i­tory works, but peo­ple want to work on text­books any­way. (They aren’t so good at fin­ish­ing them, though.)

In con­trast, Wiki­me­dia Com­mons does­n’t have the mind-bog­glingly vast col­lec­tion of me­dia and users that it ought to have, be­cause ed­i­tors can still up­load every­thing to their Wikipedia - they haven’t been forced out to Com­mons5.

Potential wikis

Fre­quently in com­mu­ni­ties, I see pro­pos­als which go ‘let’s cre­ate our own Wikipedia about our fa­vorite sub­ject, X!’ The pro­posers see clearly that there’s a body of knowl­edge which can be writ­ten down con­cisely and pol­ished as a stan­dard com­mu­ni­ty-wide ref­er­ence (like the old FAQs that Usenet groups would main­tain); but what they don’t see is that they are try­ing to com­pete di­rectly with WP and are doomed. As the Za­w­in­ski quote about reg­exps goes, now the pro­posers have two prob­lems.

I saw this re­cently on Im­, one of the largest com­mu­ni­ties de­voted to dis­cus­sion of sup­ple­ments and sub­stances which en­hance men­tal per­for­mance - since WP only cov­ers s to the ex­tent that in­for­ma­tion ap­pears in peer-re­viewed stud­ies, they should start their own wiki and write bet­ter ar­ti­cles. Such a pro­posal is pretty much doomed, al­though since nootrop­ics hob­by­ists (like my­self; see my own ) are heav­ily aca­d­e­mic and techie-ori­ent­ed, such doom may be very slow and drawn out and waste a lot of effort be­fore fi­nally ex­pir­ing.

So what should Im­minst fo­rum-go­ers do? As above, they should cover what WP won’t. In this case, any Im­minst wiki would at the be­gin­ning of each ar­ti­cle point to WP’s ar­ti­cle for all the im­por­tant back­ground, and then im­me­di­ately pro­ceed to non-RS or peer re­viewed ma­te­ri­al: what does the mass of anec­dotes say this nootropic does, what nasty effects have peo­ple re­port­ed, where can one buy it and for how much, what does it syn­er­gize with (‘stack with’), what novel con­clu­sions have we drawn from the re­search lit­er­a­ture and gen­eral bi­o­log­i­cal novel (strin­gently banned on WP as ‘orig­i­nal re­search’), and so on. The in­stant any­one adds RS or peer re­view ma­te­ri­al, if pos­si­ble, it should be re­moved and in­serted into the main WP ar­ti­cle. If an Im­minst wiki tried to write its own Wikipedia style, it dooms it­self to writ­ing stale half-assed ar­ti­cles which no one will read. (WP is so pop­u­lar that even highly niche ar­eas can kept up to date by ran­dom anony­mous ed­i­tors with­out any reg­u­lar ed­i­tors spe­cial­iz­ing in the sub­ject area.)

As time pass­es, a cleav­age will de­velop - if you want back­ground in­for­ma­tion, you go to Wikipedia; if you want in­-depth and prac­ti­cal in­for­ma­tion, you go to the Im­minst wi­ki. This is a sus­tain­able and use­ful sit­u­a­tion.


To sum­ma­rize, if Wikipedia al­lows some­thing, the ben­e­fits of do­ing that some­thing on that Wikipedia are so com­pelling that ri­val wikis find it hard to com­pete and likely will with­er. But if WP cracks down on some­thing, then that crack­down some­times pro­vide a seed of con­tent for a new wiki to grow around and also a em­bry­onic com­mu­nity of ed­i­tors who want to work on that topic and have given up on do­ing so in WP.

  1. “Wikipedia Will Fail in Four Years”, by Eric Gold­man↩︎

  2. Who will main­tain the spam fil­ters? Will there be enough ed­i­tors to up­date all the ar­ti­cles? Or enough to de­feat trolls? Con­ser­va­pe­dia is fa­mous for hav­ing true be­liev­ers edit­ing but also so many trolls and satirists that the for­mer might ought to con­sider it an ab­ject fail­ure rather than a suc­cess.↩︎

  3. To give an idea of how wide­spread the purge was: un­der my orig­i­nal ac­count, I had some­where around 6000 deleted ed­its; most of them were to Star Wars ar­ti­cles.↩︎

  4. Marcin Cies­lak, 2012-03-06, “[Foun­da­tion-l] Pre-wikis vs. ma­tur­ing Wikipedia: tak­ing away ded­i­cated ed­i­tors?”:

    I re­searched re­cently some ma­te­r­ial re­lated to a re­cent cat­a­strophic event in Pol­ish rail­way his­tory and I found out that vol­un­teers who tra­di­tion­ally dealt with rail­way mat­ters on Pol­ish Wikipedia have vir­tu­ally dis­ap­peared.

    I re­mem­ber that com­mu­nity be­ing strong few years ago, and now we found out that even some ba­sic in­for­ma­tion about in­fra­struc­ture is left un­changed.

    Few peo­ple who still main­tain that stuff on the Pol­ish Wikipedia showed me that at least two other Me­di­aWik­i-based projects have been started to fill the gap: [2][3] The lat­ter greets you even with a very nice shot of the rail­way junc­tion that was in­stru­men­tal in a re­cent rail­way crash.

    One of the projects got started by ex­pe­ri­enced Wikipedia ed­i­tors. They still copy some of their con­tent to the Pol­ish Wikipedia, but only after it ma­tures; I asked them about the rea­sons to go out­side of the Wikipedia and they said:

    • They have to do lots of orig­i­nal re­search; it is im­pos­si­ble to fol­low de­vel­op­ment of the rail­way in­fra­struc­ture and op­er­a­tions us­ing only high qual­ity pub­lished sources;

    • They got bit­ten a bit by the “no­ta­bil­ity” dis­cus­sions in their field; they want to doc­u­ment every track, every junc­tion and every lo­co­mo­tive and they are tired of dis­cussing how “no­table” a par­tic­u­lar piece of rail­way equip­ment re­ally is.

    I would have said it’s just a sin­gle case, but I’ve seen some suc­cess­ful web por­tals be­ing launched by peo­ple in­ter­ested in his­to­ry; what is differ­ent from many his­tory re­search and fan pages is that I’ve also seen some ac­tive mem­bers of Wikipedia com­mu­nity be­com­ing more and more ac­tive on those in­de­pen­dent sites.

    …In gen­er­al, I think this is noth­ing new. There are thou­sands of fan wikis on places like Wikia, where cer­tainly some con­trib­u­tors copy over some ma­ture con­tent to Wikipedia, should li­cens­ing al­low that.

    But maybe there is some trend that could prob­a­bly be bet­ter re­searched…

  5. In part this is be­cause dis­abling ‘lo­cal up­loads’ en­tails for­bid­ding any use of ‘fair use’ pic­tures. When you con­sider the fact that the im­por­tant pho­tos for 99.9% of in­ter­est­ing global events for ap­prox­i­mately the last cen­tury can only be used un­der the fair use doc­trine, you be­gin to com­pre­hend what a loss in ar­ti­cle qual­ity a dis­abling of lo­cal up­loads would en­tail. Now, in the­ory one could keep lo­cal up­loads en­abled for fair use and force all the Free im­ages over to Com­mons (prob­a­bly via an ag­gres­sive cam­paign of dele­tion-tag­ging and au­to­mated up­loads to Com­mon­s), but the Ca­bals fear that this would lead to in­dis­crim­i­nate and legally dan­ger­ous overuse of fair use. So there are few prospects that this stale­mate will change any time soon on the Eng­lish WP.↩︎