Why I Want Wikipedia to be Less Bureaucratic

Patrick Nielsen Hayden at Making Light, a place I hang out in a lot, started a bit of ruckus in the thread at the link above, by commenting on a row at Wikipedia over admin powers and responsibilities. In the course of the discussion, the subject got to the enforcement of style guides and how this both shapes and discloses the uses and abuses of power on the site. The style discussion gets kicked off by comment #70, though the whole thread is worth reading.

I'm not going to recap the discussion; most of the commenters are at least as intelligent and articulate as I am, if not more so, and they don't need me to paraphrase them. But I want to expand on one comment I made at #153:

What concerns me the most about the way Wikipedia is going is that it's a grand experiment that will very likely affect the way organizations are built on the web for many years to come. In particular one of its experiments is in automating the routine aspects of organizing and regulating a consensus-driven endeavor, and so far that part of the experiment has been pretty much a failure, as shown by how much energy goes into these "routine" decisions and their aftermath. That's a shame; if we could learn how to do that, it would make creating and operating dynamically-organized work groups much easier and faster, and therefore more common.

One of the reasons I've spent the last 25 years or so working on computer software is that I've had a vision of the computer as a tool that can be used to help people communicate and work with each other. There are a lot of necessary tasks in any social endeavor (and all groups of people working together are social in nature) that people are typically bad at, because they involve mind-numbing drudgery, or require highly-consistent results on routine decision-making. The result is either non-optimal decisions, delayed or ignored decisions, or both. It seems to me that some of that problem can be solved by automating some of routine work that has to get done.

It would also be helpful to automate routine organizational work: figuring out who needs to be notified about things, managing the form and content of routine messages, scheduling routine tasks, etc. Doing this would allow the building of organizational frameworks that can be parameterized and customized for a given organization. So starting up a new group to do something like what an existing group does would be relatively easy. Properly done, customization could be continual, so the organization could be modified to deal with new requirements and newly-discovered consequences of old requirements. Essentially what I want to do is remove the human bureaucrat from the organization; make routine the domain of the machine, and make sure that there are humans available to monitor the artificial bureaucrats and prevent them from becoming autocrats as human bureaucrats so often do.

In my wilder dreams, I hope that this idea can be applied to government as well. I'm somewhat of a minimalist in government; I've often said that I'd be an anarchist if I could figure out how to make it work in the real world. But I'm not a Libertarian, because I know how much grunt work has to be done to keep any social structure running, and I know how easy it is for that gruntwork to become bureaucratic makework.¹ So some sort of formal regulation structure is necessary, much as Libertarians and Conservatives may not like the idea.

Here's an example of how this idea might work. Let's say I want to create a special interest group on theatrical costuming.² The group should be a resource for discussion of the history, purposes, techniques, and styles of costuming, for the exchange of designs and tools for creating costumes, and for exchange of information about jobs in the field. And it will probably acquire other purposes as time goes on. An organization of this sort needs some structure around maintaining a membership list, if only for mailing out notices and newsletters. If it requires dues and provides services from them, it needs some financial structure and some oversight mechanism.³

As things are done today, my first step, assuming I expect this organization to be of any size and geographical distribution, is to find someone with experience in running such an organization. That may be a professional, whom I will have to hire and pay. It probably also means hiring a lawyer, at least to set up articles of incorporation and some sort of initial licensing and registration with governmental bodies, where appropriate.

In a better world, I would first go to the open source organization repository, and look for frameworks designed for similar kinds of organization. I'd chose one and download it, then run its setup program, filling in parameters and selecting customization options. The result would be (among other things) a description of the organization in some standard form that could have legal status once approved and registered. This could be passed among the people who are interested in the initial formation of the organization and discussed and modified until consensus on the form is reached. The final description could be sent to a lawyer to be vetted, in the same way one would get legal advice on a contract.

Once the description is agreed upon, the server program that implements the organization is started and connected to database, web, and email resources as needed. Then the initial set of officers, administrators, whatever you want to call them, the people who will monitor the operation of the organization and handle out-of-routine decisions, would be identified to the server and the organization could begin to function. As problems with the description or new requirements are discovered, the administrators would modify the customization to deal with them using tools built into the framework.

For some time, I've viewed Wikipedia as a (rather primitive, I admit) test bed and protoype for some of these concepts, and that, if it is successful, as an illustration and demonstration of the advantaqes of this sort of automation. But the amount of energy that's wasted in the wrangling over decisions that should be made and appeals handled routinely as well, has convinced me that it would be better not to consider Wikipedia as a test of these ideas, because they would surely be deemed failures based on the current outcome.



1 Classic description of this process in fiction: in Ursula LeGuin's "The Dispossessed", the department responsible for scheduling people's regular cleaning and maintenance assignments in an anarchical society ends up running the society.

2 Let's assume for a moment that organizations like the International Costumers' Guild, Inc. don't already exist.

3 You need to be able to select (perhaps elect) a Treasurer, and you never leave the Treasurer with non-transparent access to the money.