Wikipedia’s Imminent Demise?
Sounding the death knell for social software applications (and classes of application) is a sport for some prognosticators and bread and butter for the naysayers. Most of the time they are equally wrong. Nothing ventured, nothing gained, right? But this major upcoming change to the Wikipedia editing system has me tempted to join in.
Technorati reveals no links to the page on Flagged Revisions, but given that Wikipedia’s success (not to mention any number of purported failings) is generally attributed to its open editing system, implementing multiple layers of bureaucratic approvals sounds like a very big deal indeed.
After multiple readings I am beginning to think that the documentation of this change is being purposefully obfuscated. The very terminology of “surveyors” who have the right to “flag” a particular revision as “sighted” (meaning: administrators who have the right to promote a page as being correct) is a bit mystifying. But the bottom line appears to be this: there will be a new class of administrators with a rather broad power to vet pages. The mechanism of this power comes in the display: the sighted (approved) version of the page will be the one that users who have not logged in will see by default. The fact that later revisions are available is touted as a major reason why this isn’t a bad change, but being available and being obvious and utilized are very different things. A closed system is a closed system regardless of whether the system is actually locked or only apparently so.
Users who are logged in (an exceedingly small minority of users to whom this change is not really directed) will see the actual, current version of the page. This makes sense: such users are typically those who are editing Wikipedia, they will understand these changes which will be opaque to more than 99% of the Wikipedia users. Far worse and more restricting is the fact that all users will, when they edit the page, will see the current version! This will not just be disconcerting when there are newer, unapproved revisions… it will reduce and dissuade contributions from the general user population which is a significant part of the group of Wikipedia editors. Registered users contribute the greatest number of edits, but I suspect that the majority of original and significant content– as opposed to stylistic, structural and essentially clerical changes– come from the huge unregistered population. It is, after all, what wikis and Wikipedia are all about! And let’s not forget that future registered users come from this vast core of anonymous editors. You know those users, they are the ones who go on to be dedicated, enthusiastic Wikipedians.
Is corporatization an inevitable attribute of long-term, sustained success and growth? More importantly, are all the familiar power structures that have been partially subverted (and re-created in different forms) bound to come back with enough time and popularity? I understand the motivation behind this change, but it seems like a poor– and potentially tragic– implementation. Is this an example of the kind of mediocrity by consensus that some of the negative voices claim? And will this new, officially licensed group do what all special groups of this type tend to do and, consciously or not, assume the role of power-seeker and empire builder?
August 27th, 2007 at 11:38 am
Hi Chris —
Not sure if you also found out about this through the quickly-withdrawn notice on the watchlists, but I think you are correct that there is a great deal of insiderness and obfuscation going on. I think you and I are (right now — 27 August 2007) the only bloggers to remark on what is going to be an absolutely massive change in the way wikipedia functions. (You can see my remarks here.)
Yours,
Simon
August 29th, 2007 at 1:19 pm
You seem to have misunderstood. This is not a massive change at all, but an incremental change designed to open Wikipedia up further, not to close things down.
For a long time we just had the ability to “protect” articles. This was not good because only admins were able to edit.
So then we created “semi-protection” which is more open and allows more people to edit even high risk articles… basically anyone who has had an account for at least 4 days.
“Flagged revisions” even further lowers the barrier to participation on those articles, by allowing *anyone* to edit, while the edits are held in a queue.
This is a very minor change, but in any event, it is not a closing of wikipedia, but an opening. For example, I would anticipate that once we go live with this, we will be able to allow public editing of the front page of English Wikipedia for the first time in more than 5 years.
August 29th, 2007 at 2:13 pm
I appreciate hearing from you… but I don’t think I have misunderstood. I think we differ philosophically about the magnitude of the change. It looks small to you, but not to me (and some others, clearly).
Having people editing a different version of the page than the one they are seeing is a bad move– I suspect it will hurt the most with the population that is most needed: the unregistered users.
Creating a formalized bureaucracy (when the informal one is plenty) also seems like a potentially bad move.
I stand by my post and particularly the questions in the last paragraph. They are questions; nothing is certain. But I do think that Wikipedia– of all organizations– should be *particularly* wary of “small” changes. They so often aren’t as small as one thinks.
August 29th, 2007 at 4:31 pm
Where you go wrong on this:
1: “there will be a new class of administrators” – You could call it that… if ‘all users who have been on the site for more than 30 days and made more than 30 edits’ is to be imagined as a ‘class of administrators’. I’d call it rather, ‘everyone who has any significant history of verifiable contributions’. Granted, having the software automatically grant access after the 30days/30edits criteria above is just one of several proposals, but it is one of the most popular and they are all along similar lines… the access is to be given out to everyone who has been around long enough to be somewhat confident that they aren’t going to make a mess of things.
2: “Having people editing a different version of the page than the one they are seeing is a bad move” – Can’t be TOO bad given that it has been part of Wikipedia since day one. Frequently when you go to edit a high traffic page you find that it has been changed in the time between when you loaded it and when you hit the ‘edit’ link. Perhaps a cause of minor confusion the first couple of times it happens, but hardly a great and insurmountable barrier to editing. Flagged revisions will be no different. As noted above, most active users will have the access to flag their own updates (and indeed, will have them flagged by default). The minority of edits from users without this access will be quickly reviewed, just as ALL edits are today by ‘Recent Changes patrol’ and flagged or removed if they are detrimental. Thus, this issue of ‘different versions’ refers to a matter of minutes, no different from current cases of getting a slightly different version when you click the ‘edit’ button… and only for users who are not logged in, the vast majority of whom seldom or never edit.
In short, this change is meant to do one thing… introduce a BRIEF delay between the time when edits by relatively new and unknown users are MADE and when they are DISPLAYED to the general reading public. Right now Wikipedia gets hundreds of vandalism complaints every single day… but the vast majority of those are actually removed by dedicated recent changes reviewers in the time it takes the person to write the complaint e-mail. With flagged revisions that problem goes away… the quick review and reversal in the vast majority of cases still happens just as it does today, but the change is never displayed to the unlogged public.
The PURPOSE then is to introduce a better way of keeping vandalism out of public sight. Better than the current system of protection and ’semi-protection’ that is. All those pages which are protected against vandalism can be UNblocked when flagged revisions are implemented… allowing everyone to edit MORE freely. The vandalism is still kept out of public view, indeed it is suppressed much better than it has been to date, but useful contributions can now be freely made to all pages.
Far from a calamity, this is Wikipedia finally reaching the state where it can protect its articles from vandalism without blocking out useful contributors as well.
August 29th, 2007 at 4:58 pm
You paint a rosy picture for #1– if the change was that innocuous then why make it in the first place? That is the paradox of needing restriction while touting freedom– for a change to have significant effect it almost *has* to have significant risk. Also, there is no good reason to worry in the general case about “people making a mess of things” when so far people have, in the end, made good on things… kind of the success of Wikipedia in a nutshell.
Point #2 is immaterial. There are already locked pages, those are already a problem. The number of locked pages is very small, the number of people effected by this change dwarfs those pages and their edits. It feels like two steps forward, eight steps back.
I do not misunderstand what is being proposed. I even understand the motivation– an effort to maintain an old-school idea of legitimacy for a public that essentially has no understand of what Wikipedia is. So a choice is being made that tries to make fewer visible reversions and changes (that is, after all, what traditional “legitimacy” is partly predicated upon, stability). And this method avoids many of the more obvious problems of a true class system of approved users and editors. But being better than some alternatives is different from being good, and being subtle doesn’t mean it can’t be seriously flawed.
As I noted, I could be wrong. This might go swimmingly: the lost edits made up for by others, the class problem not become a problem, etc … but I suspect this is a bad decision. I also think it will take quite a while for the effects to become clear, since this will not matter to power users and– as someone else pointed out– there are plenty of interested new visitors coming to make up for any that are lost. And who knows– Wikipedia is a hugely growing concern– the long-term damage could be significant and still get lost in the general trend. Wouldn’t be the first time that a bad administrative decision has had the effect of diminishing without actually destroying! Many bad decisions are survivable.
August 29th, 2007 at 5:59 pm
You could not have gotten this much more wrong:
“administrators who have the right to promote a page as being correct”
It will not just be administrators but any editor who has over so many edits ~300 and has been editing for so many days ~30 or who is determined to be not a vandal.
“the sighted (approved) version of the page will be the one that users who have not logged in will see by default.”
That’s kind of the point, that way readers won’t see nearly as much vandalism.
“I suspect that the majority of original and significant content– as opposed to stylistic, structural and essentially clerical changes– come from the huge unregistered population.”
I would say that too but as far as I know, no study has been done to determine what percent of Wikipedia’s anonymous readers edit as well. And this is targeted toward THE READERS.
Also, you seem to misunderstand a lot. This will not be automatically applied to every page, that would be stupid. This will have to be applied to each individual page manually and it will not be applied to every page. It probably won’t even be applied to the majority of pages. Everything else will be edited like it is now.
August 29th, 2007 at 6:07 pm
Chris and I seem to be fighting misinformation on two fronts. If you see something here that says Chris is wrong, check over at rhubarb and it’s probably refuted. And, of course, vice-versa.
MrZ’s statements are all incorrect.
1. Surveyor status is not granted automatically to anybody, 300 edits, 30 edits, 30 days, etc. It is granted individually to trusted users, who must apply to an administrator who then approves the request.
2. After the inital rollout, pages will automatically go over to “sighted” status — i.e., will be incorporated into the system — as soon as a surveyor marks them as “sighted” with the goal of this becoming universal in the wiki.
August 29th, 2007 at 8:52 pm
Yes, Mr. Z is confused and this fight, as Simon notes, is spread out on two fronts where important comments are being made.
I would add to Mr. Z that even if public edits *are* a minority (I don’t think they are a minority of quality content edits, but neither do I want to get into a debate about what quality content edits are; I am referring essentially to edits that are original and significant rather than matters of style, punctuation, grammar, or admin), I doubt they are an insignificant one… and the proportion doesn’t really matter because a major part of what Wikipedia (and the wiki) is about is the immediacy of being able to effect change. It’s like enabling moderation on a blog or forum, etc… even if approval is quick and consistent, it seriously discourages participation. The action, as well-intended as it might be limits conversation *and* limits the point of entry where most people who become dedicated Wikipedians start.
For some reason, some people think that the editing in the Wikipedia is not a social conversation. But it *is* and that process is at the heart of what makes Wikipedia interesting and resilient in the face of criticism.
Wikipedia is not a typical wiki, granted, but neither is it exempt from riding on the Cluetrain.
August 30th, 2007 at 1:05 am
No, MrZ is entirely correct and it is the two of you who are clearly spreading “misinformation”. Absolutely everything in Simon’s most recent post above is simply false. Absolutely and completely false. There is virtually no support for requiring the access to be manually set rather than automatically granted by the software OR having all pages flagged immediately when the system is rolled out. That’s complete fiction. The German Wikipedia, where this update will first be tested, has long settled that the access will be granted (automatically by the software) after 30 days and 30 edits and each page must be flagged manually. If a page has no flagged revisions then it continues showing the current update to all users just as it does now.
As to the ’social conversation’. Right now, most edits are reviewed for obvious vandalism, copyright violations, and the like within minutes of being made. The same will continue to be true after flagged revisions are rolled out. The vandalism will be cleaned up just as quickly as it is now. The difference is that during that brief delay it won’t be seen by the general public. No, positive contributions won’t be seen during that delay period either, but that’s hardly a dis-incentive. The user MAKING the edit WILL still see it immediately upon hitting save. To them it will appear exactly as today. If they leave the page and then go back before the update has been flagged (again, this kind of review happens in minutes currently) then their change may seem to be ‘lost’, but that should be uncommon.
There is no ‘limit on conversation’. No ‘bureaucratic moderation’. Every edit which is not obviously detrimental will still appear for all users within a matter of minutes. ALL edits will be displayed immediately to the person making them and the logged in users who make the majority of updates. Only the unlogged users who do most of the READING of Wikipedia won’t get them for a few minutes (or never if they are vandalism).
August 30th, 2007 at 1:19 am
CBDunkerson –
Your claim that the proposal suggests automatic granting of surveyor rights is false. The exact opposite is in fact demanded. Your claim that the proposal says surveyor rights will be granted with 30days/30edits is false. All of these simple facts can be verified at the proposal itself.
You and MrZ, despite both being administrators on wikipedia (1, 2), have done nothing but make things up and obfuscate the truth regarding flagged revisions. It is getting tiresome; this is the last time I will respond to your comments here or on my own blog.
August 30th, 2007 at 1:29 am
You’re just going in circles now.
1) An unregistered user who chooses to edit a page will, unless the page has been sighted, receive a different page (the actual, most current, whether sighted or not page). This is a problem.
2) An unregistered users revisions are NOT immediately available to other unregistered users. That is a problem.
3) Creating a new class of administrators is potentially a problem.
4) There is NO guarantee that it will be 30 days and 30 edits– it might be 10 days and 5 edits, it might be 120 days and 120 edits– that has not been decided. You can charitably assume it will all work out– some of us have been around a while and seen what can happen with good intentions.
Every barrier that is put between legitimate users and their immediate contributions is a potential problem. It has, to use a popular term, a chilling effect on the process of participation. It doesn’t matter how reasonable the barrier might be.
There are two central problems here and you can justify the wikipedia decision until you are blue and it doesn’t change things. First is the mechanics of editing and the effect it might have.
Second, and even more troubling, is the central conceit driving all this: a desire to achieve a kind of legitimacy that doesn’t really make a lot of sense of Wikipedia. A central and important positive characteristic of Wikipedia is that it is NOT Encyclopedia Britannica. It is a living, breathing resource. A successful wikipedia is a different proposition from a gatekeeper-controlled publication, and trying to make it appear to be the latter but operate as the former is an untenable and schizophrenic position. Not to mention pointless… we already have plenty of reference resources that can’t change in real-time.
The very minor positive effect of stopping some spam at the expense of delaying good contributions looks like a fool’s bargain in service of a fool’s errand. As I’ve said before I could be wrong. I hope I am.
The repetition of guesses presented as settled facts and the inability to see a bigger picture of a new, emerging kind of information resource that doesn’t need to be framed like existing traditional sources of information is tiring.
August 30th, 2007 at 5:18 am
I don’t suppose you read German, but if so you can see the plans for the initial rollout of this feature at;
http://de.wikipedia.org/wiki/Wikipedia:Gesichtete_Versionen
Hopefully, you can at least guess what “30 Tage und 30 Edits” means and thus see that you are off the mark on that being some fictitious figure. Those are the criteria which will be used in the first tests. Yes, other figures and methods have been debated on the English Wikipedia but that’s just people talking. There are always all kinds of different ideas and cherry-picking the worst of these, rather than the actual implementation plans cited by Jimbo, the Foundation, the developers of the feature, and the initial test rollout of it on the German Wikipedia IS mis-information and fear mongering. You are painting a false picture of what is going to happen by taking scattered ideas of individual users and constructing a scenario you describe as obviously destructive… yet somehow ignore that everyone else can ALSO see how obviously destructive such a plan would be. That isn’t the plan. That’s… well a stupid thing to do. Why are you assuming that the plan is to do the stupid thing? Rather than the thing that the people driving it (Jimbo, Foundation, developers, initial test on the German Wikipedia) are saying?
The same kinds of pronouncements of doom accompanied the introduction of ’semi protection’. Exactly the same arguments about it ‘creating a new class of users’ and ‘putting up barriers’. Some people suggested all kinds of restrictive standards for that too. What did they eventually settle on? Accounts had to be five days old. No massive loss of users. No ‘elitism’ since nearly everyone you might be talking with already had the access or would get it in short order. No ‘barriers to editing’… indeed, replacement of protection on many pages with semi-protection that most users could bypass.
Flagged revisions has always been envisioned and developed in the same vein. The plan (as Jimbo said above) is that it will have minimal impact… pages will be unprotected so everyone can edit them and the minority of contributions by relatively unknown users who don’t have the access yet will be delayed from viewing by unlogged users for the brief time (currently a few minutes on average) it takes to strip out obvious vandalism. That’s it. Nothing else changes. If the implementation doesn’t meet those goals then it will get tweaked until it does.
I never understand the myopia of conspiracies in overlooking that ‘everyone else’ can see the obvious too. Yes, making it hard for people to get their changes quickly displayed on Wikipedia would be bad. Everyone knows that. It’s obvious. So what possible reason could there be for assuming that to be ‘the plan’? Especially when they are telling you it isn’t? They’re lying? They WANT to do the stupid thing which will destroy their site? What kind of sense does any of this make?
You’ve taken random comments and ideas from various people out of context and put them together to form a bad plan which you then criticize… but it was never the plan to begin with. Seriously, go read the German page if you can. If not then at least try to have a little faith that people aren’t lying about not being morons. :]
September 3rd, 2007 at 7:55 am
Chris,
Sorry, have only scanned this, as it seems you were swamped by true believers pretty quick and the matter got framed accordingly. Here’s a fresh question: So fucking what? The second wikipedia gets to be too much like britannica some enterprising lass will take all that creative commons or public domain content and repurpose it. Hell, I’m tempted to do it myself. Hell, I’m even more tempted to see what we’d get if we as a matter of course opened up a brand new blank slate to recreate the sum of knowledge every N years/months.
There’s some really interesting social dynamics stuff going on here. Wikipedia has come to be a more valuable and credible resource than some of us ever imagined. Makes perfect sense that there would be a little ossification. The difference that makes the difference, I think, is that it’s all free. Don’t like what they’re doing? Open up your own. Which brings us right back to the core issues of authority and authentication and freedom and value. Damned exciting, ’cause in the end if wikipedia freezes up too much you can count on something to fill the void.
September 3rd, 2007 at 5:55 pm
One “so what” is that Wikipedia is an incredibly valuable resource and it would take years to cultivate something similar with a very small chance of success. I don’t need blank slates or startups– I need information I can use. Yes, if Wikipedia screws the pooch we can hope for something else, but that isn’t a good reason not to try to avoid the pooch screwing in the first place.
Opening up your own is easy, getting anything worthwhile is, as you know, much more difficult. I don’t have time or inclination even to try.
September 4th, 2007 at 5:49 am
I’m not sure I’ve been clear. If we stipulate that the wikipedia snapshot taken today is a valuable resource, it’s a creativecommons/public domain reasource. Slurp down a copy. Sure, that’s easier said than done. But maybe it’s time for a couple of spiders to do exactly that, then post the tgz to the pirate bay. Information wants to be free, and wikipedia _is_.
Say the powers that be freeze (or substantially chill) wikipedia; then count on exactly what I’ve said, with a yeastlike explosion. We’ll get “The Yale LUG Wikipedia” and “The Fulsom Prison Inmate’s Wikipedia” and the originators will get the kind of centralizing they like and the splintering forces will do their thing. Which, from an evolution of data and social projects standpoint, is kindof cool. This move to chill makes a strong argument that the criteria for explosive growth are not the same as the criteria for maintenance (and you agree with the chillers that there’s a valuable resource here to be maintained, the disagreement lies more in defining the boundaries of that resource.) But this time the move to switch to maintenance criteria can be predicted to actually accelerate growth by fission, don’t you think?
September 4th, 2007 at 8:46 am
No, you’ve been clear, I just think you are misguided in your suggestions. You can already download all the Wikipedia data– there’s a link somewhere. But continued growth is what keeps it vital– I have plenty of dead, print reference resources and digitized versions of the same.
And having a thousand Wikipedias flourish is a nice dream, but what you will end up with is a thousand dead sites with little or no updates. No thank you, I’d rather agitate to prevent bad decisions now that pack up my marbles and go play alone.
September 4th, 2007 at 8:48 am
One other note. If my fears are realized, then the problems these changes will cause with Wikipedia will take quite a while to become noticeable. Some people may never really pay attention because WP has so much momentum– it will take a long time to slow down, and the resulting path may be significantly less than it could have been while still being enough for many people. This is not a condition that is going to stimulate a lot of competition (as if such a thing has to happen, which I don’t think is true either).