QUOTE(Kelly Martin @ Sun 20th September 2009, 11:27am)
![*](style_images/brack/post_snapback.gif)
Perhaps the problem is that the software allows for blocks of definite duration. That is, if all blocks had been "indefinite" in the first place, there would not have developed a mythos around blocking that treats an indefinite block as a "life sentence".
Law of Unintended Consequences, I suppose. I imagine the reason for timed blocks in the first place was to allow for blocks to deal with what were presumably transient disruptions without having to force an admin to remember to go back and remove such blocks when the disruption had presumably passed. In retrospect, this probably should have been handled differently, as the present scheme has obviously created difficulties.
Definitely. When the police officer yells "Stop!", and someone doesn't stop, they might get arrested. An arrest has no specific period associated, though there are legal limitations. What makes sense to me is that all blocks would be indef, with no presumption that they are correct. If an editor can't respond to a warning and stop a behavior, at least pending discussion, then the editor is in a dangerous condition and should be blocked, until there is a reason to unblock. Which leads us to
the real problem: not blocking procedure, but unblock procedure.
Which leads me to the practically universal problem: how does the community make decisions? Technically, the community doesn't decide, it advises, WMF decides, perhaps through Jimbo, but, of course, in practice, the community, through a mass of rather uncoordinated administrators and editors, makes all routine decisions, and the WMF's "decisions" are mostly to leave it alone.
(Another way to state this is that all Wikipedia decisions are made by individuals with the power to implement their own personal decision, and, then, with an assumed responsibility for having made the decision. This is why I argued against closes of discussions with an admin saying "the consensus was ..." instead of stating why that admin was making the decision, personally, because some very good closes have been made that differed from 90% of those commenting, and I've seen bad closes where it actually snowed -- participation bias plus me-too !votes, with no research. There are lots of excellent procedural reasons for no decision ever being made purely on a vote basis, though there are also good reasons to avoid closing with a decision that disrespects an uninvolved majority, because it's likely to be disruptive, even if the close is "correct." Don't agree with the majority, don't close that way! Close favoring the minority, if you are neutral and that is your conclusion, or don't close. And then you can defend the close personally, and, as well, change your mind. The problem with "consensus" closes is that there is then nobody responsible for the decision, and to change it requires a new discussion, instead of the much simpler request to an admin to reconsider, based on new argument or evidence, which I was able to do a number of times, successfully. No DRV needed.)
Which then leads me to possible solutions,
methods for large groups of people to seek and find true considered consensus, efficiently and fairly, given the nature of the situation. Because the community consists entirely of volunteers, centrally coercive solutions, while very common because they are so easy when centralized power exists, are known to fail, to divide and reduce effectiveness; in the long run, such solutions weaken voluntary organizations and often ultimately cause them to fail. They are unstable, but because of the apparent absence of alternatives, failure tends to oscillate between similarly unstable solutions. Overthrow a fascist government, what do you get? Maybe you get something better, quite often you get something worse. There is another approach.
Some of those reading this, and knowing my history, might be surprised at my conclusions: they imply that
involved admins may block, unless they have been enjoined not to do so. "Involvement" is highly associated with knowledge, and excluding those with knowledge from blocking or banning (i.e., strong warning re a topic area, as I interpreted it) is to exclude action by the knowledgeable. Until we have more efficient means of advising the uninvolved, it is necessary to enable involved action, but, then, to set up much better structures for reviewing such actions promptly and reliably. And, let's not forget, efficiently.
The present methods for review, when there is real controversy, are horribly inefficient. Were they reliable, the inefficiency might be tolerable. But they are not reliable, precisely because they are inefficient. Who has the time to review all the evidence and argument presented in RfAr/Abd-William M. Connolley? In an efficient system, most of that text would never have been written, or, if written, written in very small-scale discussions seeking consensus, with escalation only over points of disagreement, until major specific disagreements might be considered at whatever top-level forum exists, restated through consensus summaries by each "faction." You might see an ArbComm case where there are two parties presenting a case and answering questions from arbitrators, with general community comment partitioned off and filtered into the case by the parties or arbitrators.
Structure. Deliberative process. Filtering. All of these are necessary for community process to exceed in intelligence the intelligence of the most perceptive of the participants; instead, because of the lack of structure, overall intelligence is greatly reduced.
ArbComm itself has some structure, but not nearly enough.