Who Writes Wikipedia?

http://www.aaronsw.com/weblog/whowriteswikipedia

by hiteshiitk • 16 years ago

165 55 16 years ago

jdp23• 16 years ago

Fascinating analysis, using a different metric and coming to a completely different conclusion than Wales did. If he's right, it'll be interesting to go back and reread what various people have written about Wikipedia. Four years later, has anybody run the larger-scale analysis he suggests here?

guelo• 16 years ago

2006

xentronium• 16 years ago

As a moderator of a wiki-project, I can confirm Aaron Swartz's results. After we had gained enough popularity, most of the articles were written by outsiders and core community was there only to maintain -- categorize, wikify, create interlinks here and there, delete garbage and, last but not least, have flamewars in the discussion pages.

sliverstorm• 16 years ago

It's only logical- the people best suited to write articles are people who are well-versed in the subject, and the core community can't possibly contain experts on every subject.

ErrantX3• 16 years ago

I'm not sure that is always accurate; WP simply summarizes content available in reliable secondary sources.

For most topics it is possible to do so without expert knowledge. Of course; some areas have always needed experts to write. I'd never touch a medical article, or macro-biology, for example as it is all double dutch to me :) But, say, history is often easy to write if you have the source material.

thewordpainter• 16 years ago

100% agree. the core community needs to scale with the number of pages that are added every year.

erikpukinskis• 16 years ago

A key takeaway for me is that you can't just take crude metrics, make some inferences and run with them. Jimmy Wales has been running his organization under a wide reaching and completely wrong understanding. That's because he equates number of edits with importance of contribution.

It's a basic psychology 101 concept, but one thats easily missed: don't equate your operationalized variables with the phenomena you are trying to measure.

billswift• 16 years ago

That's a danger with any management - it is all too easy to overvalue what is easy to measure and more or less ignore what is hard. I have read about problems caused by that in everything from the military through software engineering in the present. It is also referred to as "you get more of whatever you measure".

mseebach2• 16 years ago

So, in the article it's claimed that Wales wanted to do a study using this method, and (implied) if appropriate, revise his conclusion.

This is four years ago, anyone knows if that happened?

Edited to add: One thing that has been annoying me a little, is when articles, regardless of their quality, are deleted because they lack notoriety. If these deletes adhere to the conclusions in this article (written by an outsider, then deleted by an insider, rather than both written and deleted by insiders), this strikes me as an example of a policy that should be changed in face of this evidence, since capturing the knowledge of these drive-by contributors seems more important than "saving space".

aaronsw• 16 years ago

Wales admitted he was wrong to me through a mutual friend. I don't think he ever publicly recanted, though.

mseebach2• 16 years ago

That is rather disturbing. Well, as PG says, there's still room to do to Wikipedia what Wikipedia did to Encyclopedia Britannica.

Alex3917• 16 years ago

For what it's worth, here is a thread I started on the Wikipedia mailing list in May 2005, which I believe was the first public debate about this:

http://marc.info/?l=wikipedia-l&m=111502457428025

There are a ton of interesting posts from Jimbo, SJ, et al. Here are a couple of key comments by Jimbo about where his beliefs came from:

http://marc.info/?l=wikipedia-l&m=111503845405275&w=...

http://marc.info/?l=wikipedia-l&m=111521774525655&w=...

gwern• 16 years ago

Of course he didn't.

Do you remember when anonymous page creation was disabled after Seigenthaler? The Foundation promised us a study of the effects.

Years later: http://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/20...

They never bothered.

riffraff3• 16 years ago

maybe _that_ is why they need to escalate to 25M of expenses next year!

Uhhrrr• 16 years ago

I found the followup at the bottom of the page equally interesting: http://www.aaronsw.com/weblog/writefp

jaysonelliot• 16 years ago

One of the primary reasons that Wikipedia is written by such a tiny percentage of its users is that the process of writing and editing is arcane and shrouded in one of the most insular online cultures in existence.

I don't know if greater access to its priestly class would mean a better or worse site, honestly. I'm sure there are thousands of people who are experts in their fields and could improve the content of the site greatly - but I can imagine that opening the floodgates and working to democratize authorship and editorship could also drag it further into the gutter.

carbocation• 16 years ago

The article that we are discussing is actually saying quite the opposite: that the bulk of Wikipedia is written by many, many users. The bulk of the edits on the other hand, are written by the few.

brianmckenzie• 16 years ago

Out of curiosity, how many HNers write on Wikipedia? I do it every so once in awhile, but only only on a few topics I'm intimately familiar with, or to fix obvious vandalism.

sidek• 16 years ago

When I was sick over a summer, I wrote a lot. Not so anymore.

thyrsus• 16 years ago

I'm an infrequent contributor.

gojomo• 16 years ago

I make additions and contributions occasionally, often anonymously. Thus I was disappointed recently when the threshold to vote on some Wikipedia issue was 200 edits, and my signed edits fell short.

_delirium• 16 years ago

I do fairly regularly. I stumbled across it earlyish, so I was pretty involved in meta-Wikipedia stuff (mailing lists, Arbitration Committee, etc.) circa 2003-06, when the barrier to entry and formality was low.

Now, researching & writing articles on moderately-important-to-obscure historical subjects is something I do as an odd form of relaxation. I occasionally edit articles in my actual area of expertise, but that feels more like "work". Collecting a few sources to write a decent first cut at an article on a historical figure or event feels like recreation. Plus, I learn some things.

There'a also very little controversy and conflict compared to writing about hot topics and recent events, so not much wikidrama. People will argue over all sorts of things, but strangely enough I've never had an edit war over an article like http://en.wikipedia.org/wiki/Georg_Gradnauer or http://en.wikipedia.org/wiki/Giuseppe_Pasquale_Ricci. =]

ubernostrum• 16 years ago

Once upon a time I did. Then I had the experiences which led me to write this comment:

http://www.tbray.org/ongoing/When/200x/2008/06/15/Deletionis...

hsmyers• 16 years ago

Every now and then--- mostly when I look up something and it is incomplete or wrong or if I spot a typo. Used to be an editor and I can't help myself...

flomo2• 16 years ago

I started doing so, until I realized the only way you could really contribute an article was to enlist for a lifetime of janitorial duty. Otherwise, whatever work you did quickly degenerates into run-on sentences and trivia.

The article's point resonated with me, because it seems that the whole process is based on the assumption that you're a no-lifer who spends all day patrolling Wikipedia.

They advertised themselves as the "open source encyclopedia", but they have not adopted any other concepts from the software development world, such as "quality assurance" and "stable releases".

binbasti2• 16 years ago

They advertised themselves as the "open source encyclopedia", but they have not adopted any other concepts from the software development world, such as "quality assurance" and "stable releases".

Adopting the concept of forking, distributing the whole thing, and making it easy to push and pull in changes from any source would be a great goal.

Someone already started hacking on a tool for converting the dumps to Git repos, which may or may not be a suitable base to build upon: https://github.com/scy/levitation

tokenadult3• 16 years ago

I started editing only this after I saw some other HNers mention that they are wikipedians. I plunged in gradually to make edits to articles on the subjects that I have been researching the longest for both work and recreation, including topics I've been producing bibliographies or working papers on since 1993. I got the nerve to do that, after seeing how appallingly bad the Wikipedia articles on those same topics have long been, by reading the books How Wikipedia Works and Wikipedia: The Missing Manual.

I'm very turned off by the amount of poorly sourced point-of-view pushing that goes on on Wikipedia. I've tried to light a candle in the darkness by compiling source lists in my user space (which I link to from article talk pages), and by mostly adding current reliable sources to articles, but even at that a lot of my edits get reverted by POV-pushers (or their sock puppets or meat puppets) on ideological grounds. This continues to happen even after one group of articles I work on went through an Arbitration Committee case in August of this year. The sanctioned editors learned how to cheat on their sanctions, and the conscientious editors are still badly outnumbered (at least as to visible accounts actively editing the articles at any one time). The administrators are beleaguered, and aren't using their mops actively to clean up the mess.

http://lesswrong.com/lw/c1/wellkept_gardens_die_by_pacifism/

I have severe doubts about the statement made by Jimbo Wales quoted in the submitted link here: "'I’m not a wiki person who happened to go into encyclopedias,' Wales told the crowd at Oxford. 'I’m an encyclopedia person who happened to use a wiki.'" While I give many of the high-edit-count old hands a lot of credit for trying to maintain encyclopedic standards on Wikipedia, I can't agree that their sound editorial judgment characterizes most of Wikipedia's editorial culture. On any topic that is the least bit controversial, the culture is all about ideological edit-warring, and many active wikipedians seem to be quite proud of their lack of acquaintance with libraries or the other resources used by genuine scholars. There doesn't seem to be anyone at the top of the leadership of Wikipedia backing up the wikipedians who are doing the best work for the project and adding the most reliably sourced content.

http://strategy.wikimedia.org/wiki/Strategic_Plan/Movement_P...

If pg has said that someone could do to Wikipedia what Wikipedia did to Britannica (where and when did pg say that?), I'd like to know how to join that effort. What have any of you heard about efforts to bring about a friendly competitor to Wikipedia? The best way to help Wikipedia might be to build a point of comparison that does better work, just as East Germany was best helped by the continually visible example of West Germany until the Stasi couldn't make the East Germans afraid anymore.

steveklabnik• 16 years ago

Only the Hackety Hack and Shoes pages.

jawee• 16 years ago

I used to dedicate a lot of time to it and be active in the Wikipedia/Wikimedia channels on IRC. I would mostly add articles on the needed articles pages... just start a few paragraph stubs; or in some cases I would extend stubs. I´ve felt it harder to edit lately, although I have enjoyed working with Wikimedia Commons more...

mfukar• 16 years ago

I used to write, until my edits were regularly reverted by people that had no clue about the subject(s) at hand. I then realized it wasn't worth the effort - I rarely refer to Wikipedia since.

ErrantX3• 16 years ago

On/off contributor for the last 4 years or so. Just worked my way back in during the last four months to edit extensively on the Digital Forensics topics :)

elblanco3• 16 years ago

I don't (anymore).

About two years ago I went through about a six month period where I spent pretty much every day in a revert war with various people who definitely didn't know anything about the field/area I was trying to provide some information on.

I happen to have a fair amount of expertise and experience in a certain area (Music Trackers) and at the time thought I would lend a hand to flesh out the still woefully inadequate content to that part of Wikipedia and started or fleshed out dozens of pages of content. I probably spent north of 100 hours on content creation even going so far as to setup emulated environments in other OSs so I could get screen captures of some of the software. I spent dozens more cleaning up missing or outright incorrect information.

Every time I did something, the change was reverted. To this day, not a single edit I did ever stayed on Wikipedia longer than a week. Eventually, tired of endless arguing with editors who just wholesale reverted entire sections of material rather than edit or augment what I put in there, I just gave up.

Particularly egregious were the endless arguments over notoriety of this or that. Responses to the editors demonstrating conformance to Wikipedia's notoriety guidelines were met with silence and further reverts. One in particular, http://en.wikipedia.org/wiki/Andrew_Sega, actually has a page now anyways -- but it's not mine. Fascinatingly, I think that two of the editors causing most of the problems didn't even know what the demoscene was! In a bizarre argument from one of the editors, I was called unqualified to write about the demoscene since I myself was a 'scener!

I think I'd still be contributing if the editors had...I don't know..."edited", instead of reverted. I would have happily participated in ensuring things like objectivity and article organization were well followed, but wholesale deleting information? Nah, I'm done with participating in Wikipedia as a contributor.

tokenadult3• 16 years ago

What sources (published sources by observers of the industry) did you have to refer to at the time you posted the articles you created?

vrruiz• 16 years ago

Felipe Ortega wrote a Ph.D. thesis devoted to this topic. "Wikipedia: A Quantiative Analysis", available at http://libresoft.es/Members/jfelipe/phd-thesis His conclusions were featured on a Washington Post article, "Volunteers Log Off as Wikipedia Ages", discussed here http://news.cnet.com/8301-1023_3-10403467-93.html and here http://news.slashdot.org/story/09/11/25/160236/Contributors-...

Ras_2• 16 years ago

Second opinion (with stats) by Wikimedia Foundation Data Analyst Erik Zachte:

http://infodisiac.com/blog/2009/12/new-editors-are-joining-e... (addendum: http://infodisiac.com/blog/2009/12/why-i-changed-the-title-o...)

alex_c• 16 years ago

What are some examples of policies that would be detrimental to casual contributors?

blahedo• 16 years ago

Anything that discourages them. For instance, immediately reverting their edits for whatever reason[0] rather than fixing them or suggesting a fix; nasty messages posted to their talk pages criticising their edits; drive-by tagging pages they created with lots of boxes indicating that their article is bad and/or inappropriate. I was a Wikipedia regular for several years, but these days it seems like if I do anything more than basic spelling fixes it becomes an unpleasant experience.

[0]Lack of notability and lack of source information are the two big ones there.

redthrowaway• 16 years ago

There's a huge problem with the current Wikipedia community being almost outright hostile to new editors. Edits that are either poorly formatted or poorly written are often reverted rather than improved, and the edits of unregistered editors are very likely to be reverted. Mostly, this is due to the desire to fight vandalism, and fairly strong xenophobia.

With tools like huggle, it becomes trivially easy to scan hundreds of edits an hour, and reverting is as easy as pressing 'r'. The problem is, vandal patrolling with huggle is really boring. You're on the lookout for things to revert, and can quite easily misinterpret someone changing a number ("512" to "568") as subtle vandalism. So, someone comes along, fixes a figure that was wrong, and within 1 minute, their change is reverted. It presents a really uninviting face to the uninitiated.

As for xenophobia, the wikipedia community is almost paranoid when it comes to outside influence. This distrust isn't unwarranted, as there have been many instances of various groups plotting to have wikipedia reflect their reality. What this means, however, is that external calls to edit a particular article will likely wind up at AN/I (Administrator's Noticeboard / Incidents, basically where you go to tell on people), and the first thing any new editor would be greeted with would be a notice saying a diplomatic version of "we're on to you".

Policy itself isn't the problem; it's the community. Policy both reflects and shapes community dynamics. However, Wikipedia is a "Jimbotocracy". Jimbo can, has, and will continue to, overrule the community and do whatever he wants, usually in the form of banning / unbanning / stripping or granting of privileges. In addition, ArbCom, the "Supreme Court", if you will, defers to Jimbo. While Jimbo doesn't set policy directly (although he reserves the right to), he does heavily influence it. His views on what policy should be are often taken as passed down from the mountaintop on stone tablet, so it matters what he thinks and why he thinks it.

zavulon• 16 years ago

Even though I disagree with his views on just about every political issue, Aaron is an inspiration for me. I would love to be in a position where I don't have to worry about money and invest significant amount of time in things that interest me, without worrying about money.

It seems like ever since I started working for myself, I don't have time for hobbies at all - everything I do has to go through a prism of how will this affect my bottom line...

woodall• 16 years ago

Here is a dump I made sometime ago but just posted. I am torn on what to think. These companies know more about their products than the general public, and thus are more qualified to talk about them, but there's that little cynic in me saying they shouldn't. Take it with a grain of salt because some of them are legit.

http://www.christopherwoodall.com/blog/?x=entry:entry101221-...

quant18• 16 years ago

The article's dismissal of an editor by complaining that "[his] edits were all deleting things and moving things around" is a perfect illustration of why "bytes added" is just as bad a contribution metric as "edit count". We're not in 2001 anymore. The Internet is not short of bytes about Alan Alda or Anacondas, and nor is Wikipedia short of people who add bytes about them.

Newly-added bytes may be true or false. They may be useful or not useful even if true --- readers do not want every byte about a given topic, they want a few tens of thousands of the most useful ones. (This is entirely orthogonal to the inclusionist-deletionist debate about what topics should be included. Some bytes may be useless in the context of a main topical article about Alan Alda himself, but they would be very relevant to a subtopical article about Alan Alda's dental health).

For a popular topic, you'll have dozens of people adding bytes of varying quality. Insiders subtract the false or useless bytes (an action easily captured in statistics and then maligned on the internet by pundits), but also look at the true and useful bytes, fact-check them, and then leave them in place. This contribution --- curation --- is not captured in any statistics, but it is an important part of the mechanism by which you can have 1 expert and two enthusiastic amateurs stop by every few weeks on their lunch break to expand an article with no centralised notice or approval, without having have the place turned into a mess by the 97 vandals and well-meaning incompetents who came by in the meantime.

The real problem Wikipedia faces is in the long tail of topics, where there is only one person adding the bytes, and that person is either grinding an axe, self-promoting, or afflicted with incurable "nerdview". The well-meaning, harried, underinformed Wikipedia insiders inevitably screw up when they try to distinguish useless vs. useful bytes on these topics, but I wouldn't call the outsiders who added the bytes in the first place "experts". Unfortunately both sides' conduct may be scaring away the people who are actual experts on those long-tail topics ...

tokenadult3• 16 years ago

Insiders subtract the false or useless bytes (an action easily captured in statistics and then maligned on the internet by pundits), but also look at the true and useful bytes, fact-check them, and then leave them in place. This contribution --- curation --- is not captured in any statistics. . . . Unfortunately both sides' conduct may be scaring away the people who are actual experts on those long-tail topics.

Correct. But not just long-tail topics, but any topic that can only be properly treated by deep immersion in the subject, and especially any topic that is controversial. A lot of point-of-view pushers on Wikipedia do a lot of their POV-pushing by "framing," just making sure that some wikilinks to articles they like persist, while others are deleted, or that navigation templates or article categories play up the articles that best represent their point of view. And that's before we even get to the issue of deleting reliable sources.

fanf2• 16 years ago

My experience is that insiders often delete new and useful bytes as soon as they are added without even leavingthe contributor time to add citations!

jrockway• 16 years ago

There is that girl from Indonesia, and Jimmy Wales. And some other people. According to the ads.

ErrantX3• 16 years ago

The problem is; neither method alone is a good measure of value.

Plenty of articles I come across have "drive by additions" of some length that are unreadable, repetitive and disturbingly worded. There are contributors whose sole purpose is to go round and copyedit that content, which they may do in a number of edits... but change (by comparison) very few words. It can still take hours of effort.

A Wiki article is the sum of all those edits.

known• 16 years ago

I came across many self appointed editors in Wikipedia.

Who Writes Wikipedia?

Related Stories