The death of Wikipedia

This is part one of a two-part post. Part two, “Now let’s bury the myth,” is here.

Wikipedia, the encyclopedia that “anyone can edit,” was a nice experiment in the “democratization” of publishing, but it didn’t quite work out. Wikipedia is dead. It died the way the pure products of idealism always do, slowly and quietly and largely in secret, through the corrosive process of compromise.

There was a time when, indeed, pretty much anyone could edit pretty much anything on Wikipedia. But, as eWeek’s Steven Vaughan-Nichols recently observed, “Wikipedia hasn’t been a real ‘wiki’ where anyone can write and edit for quite a while now.” A few months ago, in the wake of controversies about the quality and reliability of the free encyclopedia’s content, the Wikipedian powers-that-be – its “administrators” – abandoned the work’s founding ideal of being the “ULTIMATE ‘open’ format” and tightened the restrictions on editing. In addition to banning some contributors from the site, the administrators adopted an “official policy” of what they called, in good Orwellian fashion, “semi-protection” to prevent “vandals” (also known as people) from messing with their open encyclopedia. Here’s how they explained the policy:

Semi-protection of a page prevents unregistered editors and editors with very new accounts from editing that page. “Very new” is currently defined as four days. A page can be temporarily semi-protected by an administrator in response to vandalism, or to stop banned users with dynamic IPs from editing pages.

Semi-protection should normally not be used as a purely pre-emptive measure against the threat or probability of vandalism before any such vandalism occurs, such as when certain pages suddenly become high profile due to current events or being linked from a high-traffic website. In the case of one or two static IP vandals hitting a page, blocking the vandals may be a better option than semi-protection. It is also not an appropriate solution to regular content disputes since it may restrict some editors and not others. However, certain pages with a history of vandalism and other problems may be semi-protected on a pre-emptive, continuous basis.

Ideals always expire in clotted, bureaucratic prose. It distances the killer from the killing.

The end came last Friday. That’s when Wikipedia’s founder, Jimmy Wales, proposed “that we eliminate the requirement that semi-protected articles have to announce themselves as such to the general public.” The “general public,” you see, is now an entity separate and distinct from those who actually control the creation of Wikipedia. As Vaughan-Nichols says, “And the difference between Wikipedia and a conventionally edited publication is what exactly?”

Given that Wikipedia has been, and continues to be, the poster child for the brave new world of democratic, “citizen” media, where quality naturally “emerges” from the myriad contributions of a crowd, it’s worth quoting Wales’s epitaph for Wikipedia at length:

Semi-protection seems to be a great success in many cases. I think that it should be extended, but carefully, in a couple of key ways.

1. It seems that some very high profile articles like [[George W. Bush]] are destined to be semi-protected all the time or nearly all the time. I support continued occassional experimention by anyone who wants to take the responsibility of guarding it, but it seems likely to me that we will keep such articles semi-protected almost continuously. If that is true, then the template at the time is misleading and scary and distracting to readers. I propose that we eliminate the requirement that semi-protected articles have to announce themselves as such to the general public. They can be categorized as necessary, of course, so that editors who take an interest in making sure things are not excessively semi-protected can do so, but there seems to me to be little benefit in announcing it to the entire world in such a confusing fashion.

2. A great many minor bios of slightly well known but controversial individuals are subject to POV [point-of-view] pushing trolling, including vandalism, and it seems likely that in such cases, not enough people have these on their personal watchlists to police them as well as we would like. Semi-protection would at least eliminate the drive-by nonsense that we see so often.

The basic concept here is that semi-protection has proven to be a valuable tool, with very broad community support, which gives good editors more time to deal with serious issues because there is less random vandalism. Because the threshold to editing is still quite low for anyone who seriously wants to join the dialogue in an adult, NPOV [neutral point of view], responsible manner, I do not find any reason to hold back on some extended use of it.

Where once we had a commitment to open democracy, we now have a commitment to “making sure things are not excessively semi-protected.” Where once we had a commune, we now have a gated community, “policed” by “good editors.” So let’s pause and shed a tear for the old Wikipedia, the true Wikipedia. Rest in peace, dear child. You are now beyond the reach of vandals.

CORRECTION: Jimmy Wales informs me that in fact there was never a time when “anyone could edit anything on Wikipedia,” as I originally wrote. “There have always been restrictions on editing,” he says. I guess I made the mistake, as others may have as well, of taking literally Wikipedia’s slogan that it is “the free encyclopedia that anyone can edit.” I apologize for my error. I have revised two sentences in the second paragraph to correct it.

95 thoughts on “The death of Wikipedia

  1. pwb

    The fact remains that virtually anyone may edit a Wikipedia article. It or its ideal is hardly dead.

    The eBay comparison is also stupid as everyone knows that there are a tremendous number of rules imposed on its users.

  2. Liam @ Web 2.5 Blog

    Reaching for the socratic method…

    Suppose that Wikipedia had always required editors to register, anonymously. Would we be having this discussion now? Is registration-free editing an invitation to vandalism? Would the frequent eviction of registered vandals be newsworthy?

    And would neutral-point-of-view be better served by the creation of pairs, or even sets, of articles, each with a distinct POV? Would that approach eliminate some of what is now termed vandalism?

  3. Seth Finkelstein

    I think it’s important to distinguish between the “silk purse out of sow’s ear” argument, and “free labor” argument. The hype around Wikipedia is basically, bluntly, that it’s magic. Throw together a bunch of sausage fragments, cover with a mystic curtain, incant the spell “Modsiw Fo Sdworc”, and poof – out will come a silky article.

    When it’s found out there’s really a man behind the curtain, some people seem resentful that you’re ruining the trick.

    If all that remains is an example of how a heavily hyped project with very elaborate ways of escaping accountability, can produce material on the level of a term paper, without paying the writers – well, you have to wonder at exactly who finds that so exciting, and why.

    It’s not a revolution in knowledge, it’s an innovation in deskilling. It’s taking the graduate-student model – get people to work for no money, to enrich and aggrandize the project-head – and applying it to middlebrow work instead of academic work.

  4. Nick Carr

    Seth,

    You’re the man.

    It is necessary first to demystify Wikipedia. Only then can we look hard and unsentimentally at its economic model and the consequences it portends.

    Nick

  5. Redvers

    A bit of accuracy would be useful, Nick. You don’t *quite* appear to know what you’re talking about at the moment.

    Administrators aren’t the “powers-that-be” – they’re just committed Wikipedians with access to extra functions that others elected for them to have.

    The powers-that-be are the owners of Wikipedia, the WikiMedia Foundation. Administrators didn’t create the semi-protection policy, the Foundation did.

    Restrictions on editing – specifically the ability to prevent an article from being edited – have always existed and have always been used. Semi-protection actually *increased* the number of people able to edit, by doing away with the need to completely lock many articles.

    Jimbo’s proposal is just that – a proposal. Administrators, editors and ordinary users are welcome to discuss the proposal. The reasons for the proposal are that the current message on the 154 (out of 1,151,768) semi-protected articles might be discouraging people from editing at all. A replacement message at the “edit this page” level would aim to direct those people to articles they can edit, without distracting people who just want to read the article itself.

    You say “once we had a commitment to open democracy” – Wikipedia has for years been very clear: Wikipedia is *not* a democracy. There’s even a page about why Wikipedia is not a democracy. New editors are usually welcomed with a message pointing them to that policy.

    You say “once we had a commune, we now have a gated community” – the number of people able to edit articles *increased* when semi-protection was introduced; Jimbo’s proposal is aiming to collect up those it still affects and lead them to places they can edit; and given a few weeks, they’ll then be able to edit the original article anyway. The “commune” has grown, not shrunk.

    You say Wikipedia is now “policed” by “good editors.” Wikipedia has *always* been policed by good editors. Wikipedia is not a democracy; it is also not anarchy. If it was, it would be useful to no-one.

    You say “You are now beyond the reach of vandals.” Wikipedia editors, administrators and readers everywhere wish this were true. It isn’t. If you’d like to come and vandalise Wikipedia, you can (but it’d be better if you didn’t). You just can’t vandalise 0.0000001% or something of the articles until you’ve been around a fair bit. That just saves everybody the trouble of having to read the same negative opinions about George W Bush being added to the article every 30 seconds, but doesn’t save it from being vandalised now and again (and reverted in under 30 seconds as a rule).

  6. Bob Aman

    I wonder if perhaps the misunderstanding that people clearly have on Wikipedia’s “democraticness” isn’t what’s also enabling the “free labor”? Or put differently, will people be as eager to contribute if they’re constantly reminded (mostly by blog entries such as this one) that Wikipedia’s not really a democracy? I wonder how many of Wikipedia’s major contributors are among the idealist group.

    (Also, apologies for not reading all the way to the bottom of the thread before posting the earlier reply, seems I just misunderstood your ultimate point; we’re basically in agreement I think.)

  7. Kingsley Joseph

    I agree with the necessity of both, demystifying the Wikipedia and understanding it’s economic model. I agree with Seth that the Wikipedia is primarily an innovation in deskilling – one that I think more strategic thinkers need to pay attention to.

    What I am disappointed by, Nick, is the false statements that you have made and refuse to redact:

    “There was a time when, indeed, anyone could edit anything on Wikipedia.” – False, as Jimmy mentions, it used to have more pages that you couldn’t edit.

    “Wikipedia hasn’t been a real ‘wiki’ where anyone can write and edit for quite a while now.” – False again, no ‘real’ wiki has ever allowed anyone to edit anything. And if that’s your definition of a ‘real’ wiki, Wikipedia never was that.

    “And the difference between Wikipedia and a conventionally edited publication is what exactly?” – How about anyone can edit most of the articles in it? How about, you can see who made what edits (so you understand their biases)? How about how anyone can post an opinion on an article, even if it is locked? How about the fact that all of this content is available for free?

  8. Simon

    Great article Nick.

    i think those who are “blating” about the harshness forget – history and fact are mearly perception. And the winners write the historical record – irrespective of its “factual” accuracy. This happens all the time, with all manner of “trusted” sources. This is why phases like “the accepted wisdom” are in our language.

    The interesting thing is that the Wikipedia articles that are of most value (and least contentious) are the definitions of technical concepts – they are invaluable.

    In a nutshell – a great resource with the same problems of all other information sources – “The Web, just like the middle ages – just fatser and with more colours”

    Simon

  9. Zephram Stark

    Whether Mr. Wales meant to or not, he enabled a Darwinism of information. Wikipedia has all the elements needed for natural selection to take place on a massive and accelerated scale: a huge diversity of competing ideas, an environment that can’t be controlled, and endless avenues of mutation in promoting one’s view of reality. Semi-protection is not the death of Wikipedia. It is simply the catalyst for a deeper level of mutation that will more effectively serve the purpose of natural selection. As Dr. Malcom (Jeff Goldblum) noted in Jurassic Park, “life finds a way.” Darwinism has entered the information age and nobody, not even the man who opened the door, can stop it.

  10. Dawn Foster

    This is not the death of Wikipedia; it is the natural evolution of the online encyclopedia. In a sense, evolution is like death. The original species becomes extinct and is replaced by one that is better adapted to the current environment. In human evolution, previous species like Homo habilis and Homo erectus became extinct leading up to the evolution of modern Homo sapiens.

    The open source community can be used as an example of how to make online communities function smoothly to produce a high quality product. We can certainly argue that Linux and other open source applications are high quality products that are created by a collection of people online, similar to Wikipedia. However, open source projects rarely (if ever) give access to the source code to anyone who wants to contribute. A smaller group of people have access to commit changes, while newer and less experienced members must submit code to others who review it and make the changes (or not) based on the merits of the contribution. These are commonly accepted practices that have been proven to work over time within open source communities.

    As Wikipedia evolves, it is adopting practices that are similar to those used by open source communities. Unregistered users and very new users are not given full access to edit any article; however, after a few days they can earn the right to make changes. Those that abuse the privilege to edit articles by vandalizing pages will no longer be allowed to make changes. This seems like common sense, especially when compared to the commonly accepted practices of open source communities. These practices help to prevent controversial entries from being edited with incorrect or incomplete information in order to protect the integrity of the information in Wikipedia and to preserve the notion that Wikipedia is a reliable and credible source of information.

    Most of us would never have an opportunity to contribute to a traditional encyclopedia, so Wikipedia is still very open when compared to other alternatives. In an ideal, utopian world where people always do the right thing, maybe we could have complete openness without restriction. These changes do not mean that Wikipedia is no longer “open”. Wikipedia has simply evolved as an online community in order to maintain its survival.

  11. Zanshin

    I have to agree with this article. I never edited a Wikipedia page before, but I wanted to give it a go, so after downlaoding the latest Simpsons episode, I saw Marge mentioned the blogosphere as a source of gossip, I thought that was worthy of a sentence on the blogosphere page, seeming that would be the first time hundreds of millions of people would have heard the word “blogosphere”. Well looks like someone wasn’t happy with it and it was removed the next day.

    If I did something wrong, or disobeyed some obscure wikipedia rules, then I apologise, but I thought I had the right to add information that I believe is relevant. Do I want to bother editing a wikipedia page ever again? Hardly.

  12. Ben Yates

    I can imagine a Wikipedia with “frozen” versions of articles — the versions that achieve the highest reputation (due to voting and/or due to the reputations of their contributors, weighted so that most recent contributors give the greatest contribution to the article’s reputation)– being made available first, or alongside the current version.

    You can do more than imagine it: you can build it yourself if you want to. Wikipedia content is available for download — all of it, if you want, all at once — and you could set that content up in a database, tie each static article version in your database to its live wikipedia counterpart (using a Greasemonkey script), and get people to use your app — to fill it with data, that is. I’m still surprised something like this isn’t going on. (I’m so surprised that I’d shop the idea around to VCs if I knew any; somebody could make a killing.)

    As for Nick: There are very few times I’m comfortable using the word “Troll”, and this is one of them.

  13. yozef

    I think Wikipedia is a great improvement tot eh internet. Sure you find alot of junk out there, bust st least this one is being supervised. It kind of comes to show you that History is Written by the winners. Anybody can argue his opinion on any subject, and so you can see a tracked history versions of any search you do.

  14. Buyo

    I found it a little ironic, though not contradictory, that a post focusing largely on criticism of “‘semi-protection’ to prevent ‘vandals’ (also known as people) from messing” should feature the following notice atop its comment box: “(If you haven’t left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won’t appear on the entry. Thanks for waiting.)”

  15. James Turick

    Many of the above people have said that semiprotection allows more people to edit an article, but that is patently false. The only time that would be the case is when the article would be fully protected anyways, but semiprotection is applied more liberally.

    As people love to use [[George W. Bush]] as an example, I shall continue with it. The administrator logs show that before the advent of semiprotection the article was protected a decent number of times, but always for short periods. Since its invention though, the article has almost always been semiprotected. Most of the instances of it being unsemiprotected were tests to see if the semiprotection was needed (and even then they were controversial).

    How can one say that more people are allowed to edit when the difference is between “no one can edit but only for short peroids of time” (really admins can technically edit the page, but doing so is generally against policy) and “almost all of the time only those that are not the 1% of the newest accounts can edit” (which is roughly accounts more than 4 days old)? This spurious assertion obviously does not hold up under scrutiny.

    Semiprotected pages currently on my watchlist:

    [[Japanese spider crab]] – Anons keep adding in things like “if you can attack them in their weak point to do massive damage.” At E3 there was a demo of a game that mentioned historically accurate battles–but also featured fighting giant crabs. The demo gave the advice that people wish to add to the article along with mentioning “real-time weapon switching” among other things. I would say there is enough activity that full protection would be warranted were semiprotection not availiable (see the page history).

    [[Elizabeth Morgan]] – Semiprotected to stop one particular user from editing (Amorrow or Andrew Morrow is someone that has been completely banned from Wikipedia as Jimbo has decreed that he should be blocked and reverted–or his edits removed–on sight. Nevertheless, he makes some valid contributions that people will still just revert. It is quite silly really, because if he actually tried to hide his edits most of them would go unnoticed. He has a big pool of IPs to use so blocking him does not do much good.). Not enough activity to justify full protection, thus less people can edit this article.

    [[Neuro-linguistic programming]] – Has been semiprotected for over a month. The article is a complete mess and has had an Arbitration case over it. History only shows a handful of anon/new user edits. Sprotection isn’t justified, much less full protection.

    [[Bogdanov Affair]] – Semiprotected since January. Also the subject of an arbcom case or the above post by Laurence. Highly unlikely that the page would be fully protected for so long.

    [[Christopher Ruddy]] and [[NewsMax.com]] – Semiprotected due to OFFICE for a month. There has been a case where admins were allowed to edit a fully OFFICE protected page before, so it is quite possible that it would be fully protected instead.

    I must agree with the chorus that Wikipedia was never anyone can edit anything, but the point at hand is an ideological paradigm shift not a technical one. Pages are being semiprotected liberally and for long peroids of time, just as those opposed to it in the first place predicted.

    Others mention that really Wikipedia is nothing revolutionary, that it is not democratic, that it does not represent the “wisdom of the crowds”, that it is not the “sum of all human knowledge”, etc. These are all true. Wikipedia is vastly misunderstood by those that do not edit it (and many that do from my experiences).

    It is something that anyone can edit! Nothing really that new. Wikipedia did not invent wikis, collaborative editing, or anything of that nature.

    It certainly does not represent wisdom of the crowds (or maybe it does but that it doesn’t exist) as many articles are just plain terrible and there are myriads of ones that are controlled by those that push their views over others (including some admins).

    The process of Wikipedia is not pretty either. People love to quote that vandalism is reverted 99.9% of the time, but that .1% is still horrible. Which is more damaging, someone putting “George Bush is gay!!!!!1111oneoneone” in [[George W. Bush]] or someone inserting inaccurate information (intentionally or not) that goes undetected? The first is reverted within minutes and the second (using an example of an anon changing the formula in a physical chemistry article) hasn’t been corrected since early February even though it has been mentioned on the talk page! There are also the massive edit wars, wheel wars (wars between administrative actions), rfcs, rfars, etc. Making a wikipedia article is like making sausage and laws, you probably don’t want to know.

    If you still wish to know, the best way is to hit the trenches and start editing. Wikipedians will attack you for the slightest inaccuracies while saying falsehoods themselves, along with the death threats and hate mail that comes along with zealotry.

    Charles Matthews says “By the way, we like the offbeat articles: no Wikipedian sneers at them. But there is a ragged fringe mostly because trying to draw the exact line is also a timesink.” I know ArbCom is busy declaring things that are clearly not vandalism are vandalism and such, but have you looked at AFD (the process through which articles are deleted)? It isn’t that surprising to see lots of articles attacked as being listcruft, fancruft, electioncruft, simpsonscruft, etc.

    Richard Schwartz, what you suggest is already partially done. See Wikipedia:Stable versions

  16. Danny Ayers

    I personally find Wikipedia an invaluable reference and use it on a daily basis. 5 minutes ago I made a minor change to a page, and I’m no administrator or anything. The quality of the material does appear to improve over time.

    Ok, so there are a number of articles which are under additional editorial control. I disagree with Jimbo’s suggestion that the fact that they are under such control should be hidden. In fact I suspect it would be better to leave these fully open, it’s usually clear which articles are controversial and hence usually less reliable. The reader is also a participant in this system, a little common-sense critical judgement is needed (as with all information sourceS).

    A process tweak like this hardly seems like “death”. I’ll reserve judgement on whether your post is a troll. But if it is, perhaps someone may like to edit the Wikipedia entry to update it to cover this aspect of the blogosphere.

  17. gidgreen

    I’m not a “Wikipedian” by any means, but I am a heavy user of Wikipedia. I think your argument (and much of the debate that follows) misses out the perspective of me, along with the tens of millions of other Wikipedia users.

    We couldn’t care if a few articles are semi-protected, fully-protected, locked, whatever. We use Wikipedia because it is a fantastically comprehensive and up-to-date *initial* source for learning about something, especially for current events, popular culture or off-the-beat subjects.

    I don’t see how preventing a few high profile articles from being edited will make any difference to that. I read Wikipedia to learn about obscure wars from the middle ages and the full plot summary of Desperate Housewives – not to read someone’s opinion on George Bush.

    For as long as Wikipedia contains so much compelling content for free, it will not die. I’m more than happy for the administrators to introduce minor policy changes in order to keep it running smoothly.

    The proof, as they say, will be in the eating.

    Gideon

  18. Sencer

    You are mixing up your definitions of anarchy and democracy. Even then you cannot apply democracy “one-to-one” to the net, as people with some net-experience quickly find out: In the real world there are limits as to who who can become a citizen of a democratic country – no such restrictions online; in the real world there is a physical limitation that prevents you from becoming a member of the same democracy more than once (like, say, opening 100 accounts).

    Finally declaring anthing that is changing/adapting over time to be dead is not only very childish, but is backwards: It is usually those site that do not adapt which die.

    Do yourself a favour and read up on community on the net. Try “LambdaMOO Takes a New Direction” for historical context. Look at the Meatball-Wiki. Or at least Shirky’s ideas in “Communities, Audiences, and Scale.”

  19. ryan joyce

    Wikipedia hasn’t been quite as ‘free’ as it’d suggest for a long, long timen now.

    It’s self-appointed administrators have always taken it upon themselves to shape Wikipedia in their own image, and whilst anybody could contribute, there were (and still are) individuals indulging in some industrial scale moderation.

    It’s veracity simply boils down to who has the time/energy to keep modifying any given entry, rather than a ‘valid opinion’ if such a thing even exists on the internet.

    Wikipedia was a lovely idea, but it was inevitably spoiled becasue people are bloody rubbish.

  20. ryan joyce

    Wikipedia hasn’t been quite as ‘free’ as it’d suggest for a long, long timen now.

    It’s self-appointed administrators have always taken it upon themselves to shape Wikipedia in their own image, and whilst anybody could contribute, there were (and still are) individuals indulging in some industrial scale moderation.

    It’s veracity simply boils down to who has the time/energy to keep modifying any given entry, rather than a ‘valid opinion’ if such a thing even exists on the internet.

    Wikipedia was a lovely idea, but it was inevitably spoiled because people are bloody rubbish.

  21. Nick Carr

    Jimmy Wales: In your latest comment you claim that the imposition of the “semi-protection” policy was a “bold step toward openness”:

    the truth is that semi-protection was a bold step toward openness, replacing our older practice of full protection in the vast majority of cases. The introduction of semi-protection, which you trumpet as a closing of Wikipedia, was actually a successful attempt to be more open, to reduce the need for full protection.

    But James Turick, in his comment, paints a very different picture of the effect of “semi-protection”:

    Many of the above people have said that semiprotection allows more people to edit an article, but that is patently false. The only time that would be the case is when the article would be fully protected anyways, but semiprotection is applied more liberally.

    As people love to use [[George W. Bush]] as an example, I shall continue with it. The administrator logs show that before the advent of semiprotection the article was protected a decent number of times, but always for short periods. Since its invention though, the article has almost always been semiprotected. Most of the instances of it being unsemiprotected were tests to see if the semiprotection was needed (and even then they were controversial).

    Is it true, as Turick writes, that (a) semi-protection is used more frequently than full protection was and (b) semi-protection is generally imposed for longer periods of time than full protection is? If so, then it sounds like your “bold step” claim may have been a bit disingenuous. Or is Turick wrong?

    Nick

  22. Andy

    There is no offense intended here, but it kinda feels like you developed a rant and began reaching for any excuse to support it. When you paralleled the wikipedia to opensource and then complained about editors “semi-protecting” articles, you seemed to make the assumption that opensource software is not subjected to the same editing and filtering process. As a matter of fact, it actually goes through an even more rigorous process generally – debugging, testing, etc.

    I got nothing out of your article.

Comments are closed.