How Wikipedia could save itself
Sat 19 Feb 2011 by mskala Tags used: psychology, socialnetI don't think Wikipedia wants to save itself. But if they really wanted to, I know how they could do it.
Ban non-anonymous editing.
The problem with Wikipedia isn't anonymous vandals. Sure, those exist, but they're easy to deal with, because they don't have much invested in the system. For every hundred people who think it's fun to insert "penis" into every third sentence of the entire project, there'll be one committed volunteer wielding a script and willing to revert all those edits - and since each such volunteer can hold off a lot more than a hundred of the anon vandals, the volunteers win that battle. Vandals don't have much invested; serious editors have a lot invested.
The real problem is those valuable volunteers! After you've made however many thousand edits, you have social status in the Wikipedia community. Furthermore, you won't put in the effort to attain social status within Wikipedia unless that status is what you want - it is much greater than any other reward for participating in Wikipedia, and anybody with the necessary skills who isn't a lonely teenage boy, will be inclined to use those skills elsewhere, to get money or other things that Wikipedia doesn't offer. So what happens? The core of committed volunteers ends up being the kind of people who very much want to acquire social status and exercise it over others. They aren't nice people.
Then the whole thing breaks down, with Wikipedia's elaborate rules being used as rationalization for destructive behaviour rather than as guide to constructive behaviour. Outsiders who complain about such things are deliberately misinterpreted as objecting to the idealized theoretical rules rather than to the corrupt distortions of the rules that are actually followed in practice. Insiders don't complain because they benefit from the corruption. As Graydon Saunders put it in a comment on Making Light:
Google and Wikipedia have the same fundamental problem -- distributed mechanisms intended to label information quality function as mechanisms of apportioning social status, at which point the incentive to hack them is functionally infinite.
The solution is to remove that incentive. If Wikipedia is to save itself, it must remove the reward of social status from Wikipedia participation. And that has to be taken all the way. I don't just mean forbidding people from linking their on-Wikipedia identities to their off-Wikipedia identities. Most people don't choose to do that now, when it is allowed; social status within Wikipedia alone is already a big enough carrot for the problems to develop. I mean forbidding people from *having* on-Wikipedia identities. No more user accounts; no more user talk pages; no more barnstars; no publicly-visible linkage of one edit to the next. No more social status for editing. Every edit must stand by itself on the merits of the edit alone, not the merits of the editor.
If there are actions that really must be limited to some kind of authenticated account? Well, first of all, maybe there aren't really very many of those. Hey, what if any IP could block any other IP for a day at the cost of itself being blocked for the same length of time? Without user accounts there's a whole lot of administration of user accounts that no longer needs to be done. And if there are still admin actions that really, really must exist - such as looking in the logs to find out whether a series of problem edits were all by the same IP address, because remember that would no longer be public information under my proposal - well then limit those to a small cadre of paid employees, whose actions are VERY public and whose job descriptions do not include being allowed to make regular edits.
They won't do it, but that's what it would take.
3 comments
The trouble with making IP addresses public is that then they become surrogates for user names. If you are the kind of person who would become a problem on Wikipedia, then it doesn't matter whether your series of edits is linked to your real name, to "def0", or to 192.168.12.34; what matters is that that identifiable entity (whatever it is named) will accumulate reputation, and then the person behind it will have a stake in the status of that entity's reputation. Then as Saunders said, there's a very big incentive to abuse the system.
"Accountability is not a bad thing. One can get reputation for good things as well as bad.": My suggestion is that maybe accountability actually IS a bad thing. The word "accountability" suggests people being punished for doing ill things, but the other side is that there seems to be a genuine and serious problem related to people being rewarded for genuinely good contributions in a way that leads to their later choosing a destructive path. Wikipedia's current model is designed around the assumption that one will only get reputation for good things, and that for people to have reputation will only have good consequences on their behaviour (encouraging them to do more good things). But it seems to be a fact of human behaviour that once reputation becomes a significant motivating factor, then people will seek it for its own sake - and they will use it once acquired to do bad things. What I'm asserting is very similar to the statement that power corrupts.
I don't have enough experience with other-language Wikipedias to say whether this kind of problem is as serious on them as it is on the main English-language Wikipedia. If it isn't - and again, I don't know that it isn't - then maybe one reason for the difference could be that those are much smaller projects, so that social status within them feels like a much smaller reward. Another, more speculative difference might be that it's pretty much only in the USA (and to a lesser extent other English-speaking cultures including Canada) that being smart equals social marginalization in the mainstream culture. Anywhere else, someone with the kind of aptitudes necessary to climb in the Wikipedia food chain wouldn't be so angry at the world. A third might be simple cultural differences in how acceptable it is to be rude to others.
Matt - 2011-02-19 09:03
I think pride in one's work is a powerful motivation and must not be taken away.
Forced anonymity does not counter this.
It would actually be preferable to see the IP address rather than any user names that obscure the origin of any contribution, be it vandalism, astro-turfing, or genuine content.
So those should be public.
Accountability is not a bad thing. One can get reputation for good things as well as bad. However, just because someone is a valuable contributor does not mean they are more correct about things in general or should have more of a say in how things are done.
There must not be a cabal.
Only the existence or perception of priviledge provides both motivation and opportunitiy to game the system.
This appears to be a problem with the English part of Wikipedia, and less so with other languages.
(As usual, I reserve the right to be wrong about anything and everything.)
def0 - 2011-02-19 02:50