User:Martijn Hoekstra/There is no consensus

Most outcomes of processes in wikipedia are defined by community consensus. At the same time, processes that work on community consensus are getting complaints about their workings. Take for example WP:AFD, or WP:RFA. There are strong sentiments that these processes do not work how they should, or are 'broken'. AfD is often critisised for its random ganging up on discussions. RFA is critisised for the questions asked, and that it is more like a vote than a debate. The problems with these processes are caused by them being determined by community consensus, while there is no such thing as community consensus

There is no consensus

Community consensus can be described as a rough agreement of the community. This is determined by an individual, be it an administrator, a bureaucrat, or an editor. The difficulty in this, is determining the consensus amongst the people who joined the debate, and even harder, determining how the discussion represents the entire community. So how does one determine the consensus in a debate. For RFA for example, there are fairy clear rules about this. If more than 70% of those who contribute in the debate support the candidate, the request passes. If less than 60% supports the candidate, the request fails. In the range in between, the trouble begins. The closing bureaucrat has to determine if there is enough consensus, and weigh the gravity of the arguments, and how they discussion represents the community.

The process of AfD has been formed by community consensus, so we can say, that there is (or should be) community consensus that if 70% of the community supports something, there is a consensus for it. But no. WP:RfB requires a much higher percentage of support to call it consensus. Again determined by... Community consensus. In the end, there is not even consensus on what consensus is. That has to be bad news for the community. If there are doubts that the person closing the debates has correctly determined consensus, we repeat the whole process, and determine this problem by community consensus.

There is no community

So apart from that there is no objective measure of consensus, the only way to determine if and individual correctly interpeted the consensus at a debate, is by community consensus, with the same problems. Can it get any worse? Yes. Apart from that there is no consensus, also, there is no community.

In the same way that consensus remains undetermined, so is the community. At least, discussions that require community consensus, have very little to do with that community. In it's broadest sense, the community consists of everyone who makes an edit on wikipedia. Unfortunately, that includes many trolls, vandals, single purpose accounts, and people who have not edited more than once. If we limit the community to autoconfirmed accounts, we still have many users that don't edit anymore. In it's narrowest sense, the community might be the almost equally vague "regulars". Some would make the point that the community are those people for which the community has consensus they are part of the community. In the light of this essay a fairly disasterous definition. So who does make up the community consensus? Whoever knows their way to the said process, and feels like showing up on the debate. The representation of the community, however it is defined, is completely arbitrary.

So what do we have? An individual, who is supposed to determine if the non-defined community has a non-defined consensus based on the opinions of an arbitrary group.

Impact on policy

• • expand**

Guidelines are desprictive. That used to work while we were still a single clear community. Now that we are not, we use them to refer to as an authority, and nobody knows where the authority comes from.

How did we get here?

• fix, format, integrate*

In the old days, long before I was an editor, the community was much smaller. It was so small, that people pretty much knew who the other people on the project were. At that point, all the community consensus processes still worked pretty decently. When someone requested adminship, you knew that person. You knew who he was, what he did on the project, and if he should be trusted with the tools. These days we don't. I don't know any of the people who are currently up for adminship. I have never interacted with them. Our processes were never designed to handle so large a community, even if it were well defined.

What exactly is the problem with that?

• integrate*

The problem is, that because our processes are not designed to handle the current community, they don't work anymore. There is a very strong bias towards the Status Quo. Every change that lacks a streamlined process for that change on Wikipedia will simply never happen. We don't know if there is consensus for it, because we don't even know what that consensus might look like, or how to poll it. The only thing that kept changes happening lately is by Divine Intervention of Jimbo or ArbCom. If they say it happens, it still has a chance of happening otherwise, we have come to a grinding, squeaking halt. And we can't properly reform our processes, because there is no process for changing them. People will simply say that the change doesn't reflect consensus, and there is no way of denying.

How do we get out of this pinch?

• rename, expand, rephrase*

It is going to get worse before it is better. We first need to recognise that our system of making changes is unsustainable. And we're not at that point yet, not by a longshot. The biggest problem at the moment lies with the undefined processes, for example, the do we allow advertisement in any form on wikipedia. Those are relatively minor issues. The currently existing processes still run pretty well. Despite complaints about them, AfD works, RfA works RfC works. Then we have no so community consensus based parts, that also run pretty well. BAG works, ArbCom works, GA and FA nominations work. Together, we are getting by. Only when we really stop getting by, when we hit a crisis, only then will there be enough community consensus to do the unspeakable: move our processes away from being driven directly on community consensus.