User:Tothwolf/rescued essays/Decision-making

This essay will focus on how decisions are made on Wikipedia regarding content, by ordinary editors, and administrative enforcement, by administrators. Top-level decision-making, by the Wikimedia Foundation, will not be covered except in passing.

Individual decisions
Editors are theoretically independent, and, in the absence of article protection, free to make decisions about content, without the need to consult others. Only when conflict appears or policies are violated does this freedom come to be restricted.

Most process decisions, involving the use of administrative tools, are made individually by an administrator, either ad-hoc, as needed, or as a judgment on the outcome of some more formal process, such as an XfD. While the process may solicit comment and even votes from individual editors, the actual outcome is decided by an adminstrator. There is no formal process for choosing which adminstrator closes such a discussion, and closes that do not involve the use of administrative tools are sometimes made by non-administrators, though there is some controversy about this.

A closing administrator should not have a conflict of interest regarding the decision, should not have been involved in prior conflict regarding it; generally, the administrator should not have voted on the issue.

It's often stated that we do not vote on Wikipedia: this means that decisions are not made rigidly based on votes, but on evidence and cogency of argument. This is often stated with a lost performative: It is not stated who makes that judgment. Generally, it is a closing adminstrator, who makes the actual decision, presumably having been advised by the evidence and arguments presented by the community, but not being limited to such. Sometimes a closing admin will decide contrary to the preponderance of the votes.

However, in some cases, the administrator does not appear to make a personal decision: instead, the administrator will simply report a "consensus." "Consensus is Delete." Or, "Consensus is to topic-ban the editor." And herein lies the problem that this essay is being written to address.

When a decision is made by an individual administrator, there is then someone responsible for that decision. There is a person to whom the decision may be appealed. If the closing administrator is, in fact, neutral, such an appeal may have as much opportunity to be accepted as it would with some more complex process, and thus the system of individual decision on which Wikipedia mostly functions remains efficient, possibly avoiding more contentious and disruptive process. That is, if one disagrees with a decision, there is an *individual* to whom one can appeal, and if the appeal is denied, there is then a dispute, possibly, with the possibility of following the gradually escalating process of WP:DR to resolve it. With XfDs, there is WP:DRV, which will bring in a new administrator to decide based on a new discussion. But it's less disruptive if a deletion, for example, can be reversed simply by discussion among two or a small number of editors.

Normally, editors are blocked, again, by the decision of an individual administrator, and thus this administrator is the first person to go to if there is disagreement about the block. The same would be true of topic bans, in theory. In practice, though, topic bans are more likely to be decided after a discussion. Generally, when there is a discussion, an administrator (or sometimes another editor?) will take responsibility6 for closing it and, if a ban has been decided, informing the editor regarding the ban, as well as, perhaps, logging the ban at WP:Editing restrictions.

If the administrator takes responsibility for that close, i.e., has made a personal decision based on the evidence and arguments, which would include an independent review of the evidence either before making the decision or sometimes afterwards, if the decision is questioned, then that administrator, who had the right to decide for or against the ban, initially, can reverse the decision either based on a change of opinion about its appropriateness, or upon changed conditions.

But if the administrator has decided to close based on "consensus," then sometimes the administrator thinks that the decision can't be reversed except by going back to the community which supposedly made the decision.

If this understanding is correct, then a decision was made by vote. And there is no individual responsible for it, the only decision the administrator made was to decide on what level of agreement constituted "consensus."

In the case I have in mind, a topic ban on an editor was supported by a strong majority of voters in a poll taken on AN/I. There was no formal close and the discussion was archived. Later, however, an editor, the one who had created the report resulting in the poll, complained that the editor hadn't respected the ban; an administrator took it upon himself to inform the editor, and when she protested, he said that "It wasn't my decision, don't shoot the messenger." When it was pointed out that this was a problem the administrator decided to take responsibility for the close as his decision.

But when he was asked for the evidence on which the ban was presumably based, he could only point to the AN/I report, plus a prior one from the same editor, and neither report contained evidence that would reasonably support a ban, though they contained implications that such evidence existed, and a number of voters had voted saying that they were assuming that these implications were accurate. But nobody had investigated; a number of voters had actually asked for evidence; when there was no evidence provided, they did not come back and vote against the ban, this kind of oversight would be pretty common. If evidence had been provided, it might have been debated, they might have been attracted to come back, and the "consensus" may have gone quite differently.

When the wisdom of the ban was questioned, the administrator, again, claimed that he could not change the decision, it had been the community's decision; apparently he saw his role as closing admin to be only to decide what the consensus of the community was, not to make an original decision for which he would be personally responsible. (And which he could then presumably reverse based on new evidence, impeachment of old evidence, or new arguments.) And he went to AN for confirmation that he had closed correctly.

At AN, comment was initially confined to supportive notes that, yes, consensus to ban was obvious, and he seemed to think that sufficient. Only later was this questioned, research was done into the evidence, an independent review was sought, an editor was blocked as a result of his criticism of this administrator, and, ultimately, the ban was lifted, on the thinnest of evidence that the banned editor had become responsive, a single request for assistance or independent opinion.

The disruption could have been avoided if the administrator had simply taken responsibility for the decision, and this is, from practice with nearly every other issue, standard practice. First stop to challenge an AfD: the closing admin. It might be brief, and sometimes this step is bypassed, but it's always there. First stop to challenge a block: the blocking admin. (Though there is, for this, a special bypass: the unblock template, for immediate additional opinions.) First stop to challenge article protection: the protecting administrator. Etc.