We *are* a grand social experiment - Jimbo Wales, 26 January 2005
The goal of the investigation was to check on the progress of the social experiment commonly known as Wikipedia, the free online encyclopedia, and report any conclusions for future reference. For that purpose, six volunteers created target accounts on three different chapters each, with specific instructions. After less than two years, my account at the English chapter is the last one standing.
Wikipedia is a social experiment to test the behaviour of human beings in a new, open-door environment with rudimentary government. It is one step in a number of experiments in preparation of several scenario's of resettlement. Wikipedia is specifically set up to test how human beings function as participants in a single project when all are considered equal.
The six volunteers, all published researchers, were asked to work on the project in areas of their expertise. Three were to declare their expertise openly, three to keep silent. Instructions for attitude ranged from timid and cooperative to firm and leading, where I was asked to vary between episodes of editing, as well as to indicate vulnerability. We were all furthermore asked to test how a range of policies were functioning in practice, without (purposely) breaking them. Preliminary investigation produced the conclusion that in larger Wikipedia chapters, policies and guidelines were generally reasonable and fair, in middle-sized chapters chaotic, and in smaller chapters absent or unworkable.
The results of my investigation support that of the earlier reports by the other volunteers. The main findings are as follows:
As a whole, the experiment is failing. Medium-sized chapters generally reached their peak in 2007, larger chapters in the first half of 2008, to since decline both in quality of work and in human relations. Most smaller chapters never got off the ground or were quickly hijacked (illustrative is that my account at the smaller chapter of my choice was closed before I could make a single edit; the chapter is now dead). So far the failure has been noticed by only a few participants.
The quality of the product grew to variable, with pockets of encyclopedic information on par with conventional sources, and other areas never better than a random advertisement website. In the declining phase, articles on controversial topics typically get changed into misinformation, with heavy editing producing no net improvement, while other areas are abandoned in their entirety.
In larger chapters, participants invariably form pockets by topic, that create their own rules and mindset deviant from existing policies.
Expertise is unwanted in all chapters. Any participant showing expertise will immediately be mistrusted and opposed by the uninformed. It makes no difference whether the expertise is declared, nor does the attitude of the expert has any influence. Without exception, experts get attacked, their participation eventually sabotaged. Foremost, Wikipedia participants insist in their belief in a society of equals, and therefore anyone above their level must be dragged down as a matter of course.
Detrimental to the experiment is the way it is governed, where administrators, rather than experts, have the final say. All attempts to supervise the administrators have failed hopelessly. Once an inner circle has formed, it can easily perpetuate itself and gradually increase power.
Participants, other than experts, can largely be characterized as follows.
Those that consider the project an extension of their existing life or hobby. This is the most productive group, working together and engaging in fruitful discussion. Their activity is now declining, mostly because they see the articles they are working on as more or less complete, and generally have no idea that there is room to improve the quality.
Those that consider the project their new life. This is the most active group. They will stubbornly re-enter the same content 24/7 everywhere in the encyclopedia, or change lowercase into uppercase a million times while another participant is doing the opposite in the other half of the chapter, until they inevitably clash, after which they will find themselves caught in procedures and fruitless arguments until they break.
Those that see Wikipedia as an opportunity to gain status and wield power that they did not have in life. They will either end up as administrators, usually without any clue to what they are doing, or as rogue elements, whose sole intention is to force other users to do their bidding or to trap them into making mistakes in order to get them removed, where changing content is a means to find and destroy victims.
Existing policies, regardless of quality, are not consistently upheld, if at all. They are often abused, or purposely misinterpreted, to aid one side in a conflict, or to assist rogue elements in their endeavours.
All chapters, if not dying, are heading towards an end situation where the third group routinely targets and removes participants of the second group, and where the first group is virtually empty and no experts will consider participation. At this point in time, they are already best characterized as battle zones.
Wikipedia, as a phenomenon, must be considered a health hazard, in view of the addictiveness to the second group and the behaviour of the third, and as an education hazard because of the misinformation it provides to the readers. All volunteers in this investigation have regularly received hatemail and threats, including death threats, those working under their own name also by telephone and on other websites. In more than one instance, local authorities had to be informed.
My recommendations, based on the investigation, are these:
Resettlements should be set up hierarchically, with experts having the final say.
Participants should not be considered equal, but they should have equal rights secondary only to the first recommendation.
A basic set of rules should be provided as unalterable.
If the above are satisfied, projects can and should be divided into smaller entities with some freedom regarding their management, but not to deviate from the basic set of rules.
Den Broeder G (2008), "Wikipedia, the Social Experiment", report to the United Nations, DOSADI 2(6) Dec
This essay was written by Guido den Broeder, founder of Wikisage, on December 3, 2008