Wikipedia talk:Centralized discussion/Removing warnings

Introduction

The results of a poll on the subject are disputed, and since polls aren't the way to make policy it seems like what this issue needs is some good discussion. I've listed what appear to me to be some of the main issues below; feel free to add more headings as required, of course. JYolkowski // talk 21:51, 25 August 2006 (UTC)[reply]

I don't believe that a consensus should be recognized here without consideration of the comments in Wikipedia:Removing warnings poll. Furthermore, no policy supports the claim that "polls aren't the way to make policy". John254 04:57, 26 August 2006 (UTC)[reply]

However, Wikipedia:How to create policy states that two ways include a) consensus approval for a specific proposal or b) long-term consensus and precedent. The poll in question achieved neither of those, which is the point people keep pointing out that you're missing. You can't just run something for a week, close it when you like the partial results, and call it policy. Not here, anyway. There's a lot of precedent already for doing it the other way, and the poll distilled very little information, given the number of caveats employed in the responses there/contradictory positions taken. -- nae'blis 05:47, 26 August 2006 (UTC)[reply]

I'm not attempting to create policy. I'm merely attempting to restore policy that was eliminated against consensus. The prohibition of the removal of legitimate warnings has been a part of Wikipedia:Vandalism well before I made my first edit on Wikipedia -- see the March 28, 2006 version of Wikipedia:Vandalism, for example. It's not as though I created a poll for a completely new policy proposal, then attempted to enact the policy a week later. The purpose of the poll was merely to show that there is still a consensus for prohibiting the removal of legitimate warnings. "Caveats employed in the responses" such as "This vote is for VANDALISM and DISRUPTION warnings" [1] are quite obvious as to their effects. John254 06:24, 26 August 2006 (UTC)[reply]

Your purpose here is concensus, but you need to have more opinions than those found in the previously conducted poll (Wikipedia:Removing warnings poll) before you can claim concensus. In fact, statistics would suggest that any concensus you reached that was contrary to the conclusions of that poll would require a contributory population several times greater than the poll. I could calculate the exact amount if you wish, but I would hazard a guess that you need at least 3 times the number of respondants.

Is there a difference between vandalism warnings and other warnings? Why?

I don't think so, although I might be willing to agree to block warnings being different from non-block warnings. JYolkowski // talk 21:51, 25 August 2006 (UTC)[reply]

I believe that there are definite qualitative differences in warnings. In my view there are two that are serious but both are different in nature. The first is vandalism (typically by supposedly new users or anons). The second is disruption and edit warring by new and more experienced users. Each of these requires a different approach. I suppose that there may also be other warnings, but (perhaps I am inexperienced) these seem to be far less frequent and of lesser concern.

There are also differences in the nature of the warning for the particular offense. A vandal may simply be "warned" at first that their efforts were considered unuseful and reverted. But later, the warnings should escalate. These are qualitatively different. --Blue Tie 16:34, 26 August 2006 (UTC)[reply]

How important is it to warn vandals as quickly as possible?

Fairly important. If an editor is simply experimenting or just goofing around, a quick warning may stop them before they have irritated too many other users. -- Donald Albury^(Talk) 01:21, 26 August 2006 (UTC)[reply]

Very important. Agree with Donald Albury. --Blue Tie 16:35, 26 August 2006 (UTC)[reply]

- I would tend to agree. Next question that I'll throw out: Is it so important to warn vandals as quickly as possible that people warning them should not pause to check their contributions or the talk page history? JYolkowski // talk 17:03, 26 August 2006 (UTC)[reply]

- - Basically, yes. It is so important that these things be done quickly that they should not have to do so. --Blue Tie 18:15, 26 August 2006 (UTC)[reply]

Should non-logged-in users be treated differently from logged-in users?

I suspect that most non-logged-in users do not look at the talk page for their IP address (for example, if the next time they log in their IP address has changed, they'll probably never see any messages on the first IP address's talk page). On the other hand, that means that they will probably never remove any warnings left on the original talk page either. Therefore, any policy concerned with users removing warnings from their talk page should focus on logged-in users, since they are the ones who are likely to receive the warnings and likely to want to delete them. --Metropolitan90 04:42, 26 August 2006 (UTC)[reply]

Treat them the same. I disagree with Metropolitan90. I think anons should be treated the same as other users and vice versa. Assume good faith but do not tolerate bad behavior in either case. However, I also agree that anons should be watched differently because so many do not know what they are doing or are malicious.--Blue Tie 16:38, 26 August 2006 (UTC)[reply]

Personally, I would tend to think that they should be treated the same. Would we all agree that anonymous users should not be treated any harsher than logged-in users with respect to warning removal? JYolkowski // talk 17:07, 26 August 2006 (UTC)[reply]

I agree with that. Also no more leniently either. We do not know if they are new or experienced users. They should be encouraged to log on, but that does not have to be a condition of anything. --Blue Tie 17:22, 26 August 2006 (UTC)[reply]

Should new users be treated differently from experienced users?

How should "new" and "experienced" users be defined?

Classes of users:

New user: Less than 3 months of active editing AND fewer than 200 edits.
Intermediate User: More than 3 months of active editing but fewer than 700 edits.
Experienced User: At least 6 months of editing and at least 700 edits (both are necessary).

New Users should be given more benefit of the doubt if they claim that they acted ignorantly or did not understand something. Experienced Users should get less slack. However, there should be an overall parity in the way that they are treated -- a sequence of warnings leading to action if they fail to meet reasonable conditions for participation.

--Blue Tie 17:17, 26 August 2006 (UTC)[reply]

How long should warnings need to be left on a talk page?

I would say that they don't have to be left for any amount of time. If a user removes them, they've seen them. If they then reoffend, give them a sterner warning about the behaviour in question. JYolkowski // talk 21:51, 25 August 2006 (UTC)[reply]
It seems to me the main function of a vandalism warning is to alert the vandal that his behavior is being observed and needs to stop. Once that has been done, it doesn't matter if the warning continues to appear. It is especially pointless if it works, and the editor stops vandalizing. What about assuming good faith? Rbraunwa 22:09, 25 August 2006 (UTC)[reply]
Users participating in RC patrol often rely on the existence of prior vandalism warnings on the current version of the vandal's talk page to indicate whether and how the vandal should be warned again, or whether the vandal should be listed on Wikipedia:Administrator intervention against vandalism. Allowing vandals to hide the warnings in the history of their talk page frustrates countervandalism efforts. From my personal experience in countervandalism efforts, I am able to state that RC patrol is a highly time-pressured activity. Every minute that must be spent on combing through a vandal's talk page history is a minute during which a vandalism warning or a report to Wikipedia:Administrator intervention against vandalism must be delayed, an additional minute during which the vandal will remain unblocked and may commit further acts of vandalism. Furthermore, while I identify the placement of vandalism warnings in my edit summaries, some users who place vandalism warnings do not supply such edit summaries. Thus, identifying such "no-summary" warnings in a talk page history would involve viewing every diff between edits, a particularly time consuming process. As a practical matter, the time-critical nature of countervandalism efforts demands that the integrity of talk pages as an quick reference to legitimate vandalism warnings be maintained. John254 00:25, 26 August 2006 (UTC)[reply]
I oppose any attempt to legislate a time period during which warnings cannot be removed. We don't need such a rule to deal with disruptive editors, and forcing warnings to stay on talk pages when editors are not being disruptive serves no useful purpose. -- Donald Albury^(Talk) 01:28, 26 August 2006 (UTC)[reply]
I agree with John254 and disagree withJYolkowski, Rbraunwa and Donald Albury. In my view, warnings do not have just the purpose of stopping the bad behavior but they also serve as an adminstrative tool in case the problem continues. Many editors have chips on their shoulders and just do not even see how their behavior is wrong. They must be handled eventually. Warnings help record the process by which this occurs. The warnings should remain for a reasonable amount of time. This could include a month or two for vandalism, or for as long as the editor is participating in the discussion that led to the warning in the case of non-vandalism activities like disruption, 3rr, etc. I am opposed to permanent warnings though. All warnings should be removeable after some period of time.--Blue Tie 16:45, 26 August 2006 (UTC)[reply]

If a user's talk page is long enough that it should be archived, and they then receive a warning, should they still be permitted to archive the page?

Sure. Just don't archive the warning if a reasonable time between its issuance and the archiving has not taken place.--Blue Tie 16:47, 26 August 2006 (UTC)[reply]

Should removing legitimate warnings be treated as vandalism, making reverts to restore them exempt from the three-revert rule and allowing users who repeatedly remove legitimate warnings to be blocked and have their talk pages protected?

No, I don't think that we should give people unlimited licence to edit war on other people's talk pages. JYolkowski // talk 21:51, 25 August 2006 (UTC)[reply]

Users participating in RC patrol often rely on the existence of prior vandalism warnings on the current version of the vandal's talk page to indicate whether and how the vandal should be warned again, or whether the vandal should be listed on Wikipedia:Administrator intervention against vandalism. Allowing vandals to hide the warnings in the history of their talk page frustrates countervandalism efforts. From my personal experience in countervandalism efforts, I am able to state that RC patrol is a highly time-pressured activity. Every minute that must be spent on combing through a vandal's talk page history is a minute during which a vandalism warning or a report to Wikipedia:Administrator intervention against vandalism must be delayed, an additional minute during which the vandal will remain unblocked and may commit further acts of vandalism. Furthermore, while I identify the placement of vandalism warnings in my edit summaries, some users who place vandalism warnings do not supply such edit summaries. Thus, identifying such "no-summary" warnings in a talk page history would involve viewing every diff between edits, a particularly time consuming process. As a practical matter, the time-critical nature of countervandalism efforts demands that the integrity of talk pages as an quick reference to legitimate vandalism warnings be maintained. John254 00:21, 26 August 2006 (UTC)[reply]

Removing warnings should not be considered vandalism.
Removing Vandalism warnings too quickly should be considered violation of the vandalism policy
Removing other warnings should be considered violation of those policies.
Restoring vandalism and disruption related warnings that were removed too soon should not be subject to 3rr.
Users who repeatedly remove legitimate warnings should eventually be blocked for a day each time that happens.

--Blue Tie 17:23, 26 August 2006 (UTC)[reply]

If a user stops their problem behaviour, is it okay for them to remove the warning, or should it remain there if there's a risk that they may reoffend?

Vandalism warnings should remain for a determined period. Disruption and edit warring warnings can be removed if the person leaves the subject that created the disruption, or if a suitable period of time without further trouble has passed.--Blue Tie 17:26, 26 August 2006 (UTC)[reply]

What if the sender and receiver of a message disagree as to its validity?

I think that this is the biggest problem with prohibiting the removal of warnings from talk pages. Because most people usually act in good faith or at least believe they are even when they're behaving in a manner against community norms, there are probably doing to be disputes as to whether a warning is "valid" or not. This is just going to cause conflicts to escalate. JYolkowski // talk 17:17, 26 August 2006 (UTC)[reply]

Criminals always disagree with the law. There is a law on wikipedia. Disagreement with it does not make it invalid. --Blue Tie 17:32, 26 August 2006 (UTC)[reply]

How should such disputes be handled, or is it instruction creep to even define a process?

I would say that if someone removes a message, and they don't agree with it, just ignore it unless they repeat the same behaviour that caused the message in the first place. JYolkowski // talk 21:51, 25 August 2006 (UTC)[reply]

I disagree. If there is a dispute about a warning, pull in an administrator. However, warning tags should contain either instructions or links to instructions about how to dispute it or remove it. --Blue Tie 17:32, 26 August 2006 (UTC)[reply]

Which party, if either, gets the benefit of the doubt?

I do not see why it is necessary that there be a "benefit of doubt". It seems like a nonsense concept here and more fit for a court of law where facts may be in doubt. The records exist that show the behaviors. But if there is some weird case where a benefit of doubt must be given, then, between two editors, the more experienced editor. Between an editor (or two editors) and an admin, the admin. --Blue Tie 17:32, 26 August 2006 (UTC)[reply]

How often do people actually remove warnings?

How often are people able to get away with problem behaviour because they remove warnings?

My suspicion would be very rarely. JYolkowski // talk 21:51, 25 August 2006 (UTC)[reply]

It is irrelevant how often they are able to do so now.--Blue Tie 17:33, 26 August 2006 (UTC)[reply]

Why do people remove warnings?

Should we assume good faith?

Yes, always, unless it can be shown beyond a shadow of a doubt that they're removing warnings in bad faith. JYolkowski // talk 21:51, 25 August 2006 (UTC)[reply]

Yes, but not beyond a shadow of doubt. By the preponderance of the evidence. --Blue Tie 17:35, 26 August 2006 (UTC)[reply]

Do they do so out of ignorance, out of embarrassment, or as a purposeful attempt to deceive?

All of the above. Everyone is different. --Blue Tie 17:36, 26 August 2006 (UTC)[reply]

If the warning takes the form of a personal attack, can it be removed?

Warnings that use the templates should not be considered personal attacks unless a pattern of harrassment can be shown. --Blue Tie 17:38, 26 August 2006 (UTC)[reply]

Should the results of this discussion be documented as part of an appropriate policy or guideline?

Is a guideline fine, or does it need to be policy?

We need a policy. --Blue Tie 17:39, 26 August 2006 (UTC)[reply]

Where? No policies or guidelines deal specifically with user talk pages, although there is a guideline that deals with talk pages.

In the policies that deal with the infractions. For example Vandalism warnings and their removal should be on the Vandalism Policy page. --Blue Tie 17:40, 26 August 2006 (UTC)[reply]

Does the previous poll on this subject reflect consensus per Wikipedia:Voting is not evil, or are polls evil and a poor way of determining consensus?

Wikipedia:Removing warnings poll does show a consensus for the inclusion of the following language in Wikipedia:Vandalism:

The removal of legitimate vandalism warnings given within the last month is prohibited. If users remove legitimate vandalism warnings given within the last month from their talk pages, the warnings may be restored and additional warnings about removing warnings added. Such restorations are an exception to the three revert rule. Users who repeatedly remove legitimate vandalism warnings given within the last month from their talk pages may be blocked and may have their talk pages protected.

Wikipedia:Removing warnings poll currently contains 25 comments in favor of "Deleting valid vandalism warnings is always wrong", and 19 comments in favor of "Deleting valid, recently given vandalism warnings is wrong", which gives us 44 comments in favor of at least prohibiting the prohibiting the removal of valid, recently given vandalism warnings. By contrast "Deleting valid vandalism warnings related to ongoing disputes is wrong" currently only has two comments, "Deleting valid vandalism warnings is discouraged but should be tolerated" only has 4 comments, and "Deleting valid vandalism warnings is acceptable after reading said warning" has only 1 comment, which gives us only 7 comments in favor of allowing the removal of valid, recently given vandalism warnings under at least some circumstances. Thus, at least prohibiting the prohibiting the removal of valid, recently given vandalism warnings has the support of 86.3...% of the established users who have commented at Wikipedia:Removing warnings poll, which results in a consensus for this prohibition. Wikipedia:Removing warnings poll currently contains 28 comments in favor of "The warning should always be restored and an additional warning about removing warnings added" (in the case of inappropriate warning removal), and only 6 comments in favor of "Issue additional warnings when and if appropriate, but do not try to restore warnings that a user has deleted" with respect to vandalism warnings (there is one comment in this category that only applies to non-vandalism warnings.) Thus, "The warning should always be restored and an additional warning about removing warnings added" (in the case of inappropriate vandalism warning removal) has the support of 82.4...% of the established users who have commented at Wikipedia:Removing warnings poll, which results in a consensus for this language. With respect to the removal of vandalism warnings, Wikipedia:Removing warnings poll currently contains 20 comments in favor of "Repeatedly removing warnings should lead to blocks and/or talk page protection, even in the absence of other ongoing disputes". By contrast, with respect to vandalism warnings, the other categories on this topic only contain 6 comments. Thus, "Repeatedly removing warnings should lead to blocks and/or talk page protection, even in the absence of other ongoing disputes" (in the case of vandalism warnings) has the supoort of 76.9...% of the established users who have commented at Wikipedia:Removing warnings poll, which results in a consensus for this language. There is currently a clear consensus for including the above quoted language in Wikipedia:Vandalism. Furthermore, there's no policy basis for the claim that a poll cannot indicate consensus or be used to make policy. Wikipedia:Voting is evil is an essay, not a policy, and has Wikipedia:Voting is not evil as a counterpoint. Finally, in the templates for deletion discussion for the warning removal templates, there was a strong consensus to retain the templates, and, by extension, the policy which authorizes their use. Note that the TFD discussion was not organized as a "poll". John254 22:26, 25 August 2006 (UTC)[reply]

m:Voting is evil is a long-standing essay that's been around for ages. Its counterpoint is much newer, and its editing history demonstrates its relative newness. Voting is evil is something worked out by a number of pillars of the Wikipedia community. They cannot possibly be treated as equivalent. The fact of the matter is, Wikipedia has a longstanding tradition of avoiding votes, and not taking votes all that seriously. The amount of major policy that has been decided by vote currently stands more or less at one piece - blocking for 3RR violations. Phil Sandifer 23:47, 25 August 2006 (UTC)[reply]

Clearly wikipedia does not have a "long standing tradition" of avoiding all votes. Requests for admin take votes and if you fall below a certain percentage is will almost certainly fail. Members of the board of trustees are appointed by vote. So clearly votes are not ignored when it comes to people or power. Is the argument that votes do not count when it comes to disputed ideas? I have seen this claim that votes are not important on wikipedia before, but I do not see any Policy or any guideline that says "Voting is Evil" or anything like that. Can you point me to one? As a matter of fact, voting is a method of establishing concensus. It is a fair and open method. The problem in wikipedia is that it often involves just a few people -- not a sufficient population. But when you get 50 or 100 people voting the results are more valid. I suspect that people who do not like the idea of concensus by vote are people who hold obscure or unpopular views and typically lose to the majority view and want protection.--Blue Tie 17:48, 26 August 2006 (UTC)[reply]

Is it acceptable for this discussion to come to a conclusion that contradicts existing policy?

I would say that any conclusion that unreservedly encourages edit warring on user talk pages is unacceptable. Small variances with existing policy may be okay, IMHO. JYolkowski // talk 21:51, 25 August 2006 (UTC)[reply]

Policy is created and modified via consensus; thus, this discussion may indicate a consensus to alter existing policies. John254 22:30, 25 August 2006 (UTC)[reply]

There's also a big fish/small fish aspect to policy. A small policy like this, to my mind, cannot alter a major policy like the prohibition against edit warring or the tradition that people's talk pages are their own. Phil Sandifer 23:43, 25 August 2006 (UTC)[reply]

I agree. -- Donald Albury^(Talk) 01:24, 26 August 2006 (UTC)[reply]

"Small" policies can create exceptions to "larger" policies. John254 00:24, 26 August 2006 (UTC)[reply]

No - that's the same logic that leads WikiProjects to decide that they can redefine NPOV as it applies to their pet field. Small policies don't get to go against the big ones. Phil Sandifer 01:34, 26 August 2006 (UTC)[reply]

If it becomes policy, replacing warnings will fall under the same headings as vandalism - exempt from 3RR. It will not therefore contradict eedit warring policy. Viridae ^Talk 01:44, 26 August 2006 (UTC)[reply]

There is widespread consensus among the community that users should assume good faith, not bite newbies, and that users can manage their talk page in any reasonable, non-disruptive manner. It would take even more widespread consensus to agree to ignore these ideas when it comes to removing warnings. Nothing, including the recent poll results, demonstrates that there is agreement to anywhere near the degree required. JYolkowski // talk 13:59, 26 August 2006 (UTC)[reply]

You could argue that hiding vandalism warning is potaentially disruptive. The majority of user that I have come across dleteing warnings are those that are in the middle of a vandalism spree.Viridae ^Talk 14:26, 26 August 2006 (UTC)[reply]

I think it's a fairly acceptable axiom that nothing a user can normally do to his own talk page should be considered vandalism. The only thing that comes close to being an exception that I can think of is redirecting it, and that's because it causes a major technical problem. Phil Sandifer 02:15, 26 August 2006 (UTC)[reply]

They do not own the userpage though, it is for the community not a personal webspace. Viridae ^Talk 02:17, 26 August 2006 (UTC)[reply]

There is longstanding tradition that a user controls the maintenance of their own userpage. Phil Sandifer 02:31, 26 August 2006 (UTC)[reply]

It is however used for communication with the community and if that communication involves warnings placed by other users then so be it. Viridae ^Talk 02:37, 26 August 2006 (UTC)[reply]

Agreed, it's used for communication with the user. It's not a permanent record of misbehaviour. If someone removes a warning, they've seen it, so the purpose of communicating with the user worked. JYolkowski // talk 13:02, 26 August 2006 (UTC)[reply]

It is also, in the case of warnings given used for communicating among the community about the user. Warnings allow others (as pointed out) to warn the person appropriately. Do you think the suspected sock templates should be removed because the user (who may or may not be a sock of someone else) has read them? Viridae ^Talk 13:56, 26 August 2006 (UTC)[reply]

It is appropriate for this discussion to come to a conclusion that disagrees with policy. However, it may not be relevant to anything if it does because....
This discussion will not be superior to the poll unless it can be demonstrated that it contains opinions from more people than the poll previously conducted.
Guidelines may not alter policies and policies may not alter the Pillars of Wikipedia. However, guidelines are equal to other guidelines, policies are equal to other policies and Pillars are equal to each other. Sometimes there will be conflicts. In this case it is up to the Admins or higher authorities to judge the matter.
Removing warnings from talk pages may not be vandalism but they should still be against policy. Even user pages fall under policy and Pillars (But perhaps not guidelines).
I am not in agreement that all warnings are equal. I do not think that there should be a warning policy but rather that each policy that involves warnings such as vandalism, 3rr and others, should include a topic on warnings and the features that pertain to the warnings under that policy. I think, for example, that a persons first instance of 3rr when they are a new user, should not result in a block but rather a warning and a direction for them to read the policy. That process of warning and so forth should be included in the 3rr policy.
Purposes of warnings are A) Inform the editor, B) Better administration of policy leading to a well-regulated wikipedia C) Enhance communication within the general body of users.
I disagree with JYolkowski, Phil Sandifer, and Donald Albury in most of this discussion. I agree with Viridae and John254 in most of this discussion. But in either case, not exclusively.
I am concerned that this discussion asks loaded questions.
I consider it a bad thing that this relatively long-standing concept and helpful method to keep wikipedia well-regulated is under attack despite the clear concensus to the contrary.

--Blue Tie 18:09, 26 August 2006 (UTC)[reply]

Results of Previous Discussion

Removing warnings is Vandalism = 7 votes
Removing warnings is against policy but not Vandalism = 14 Votes
Removing warnings is discouraged but not against policy = 10 Votes
Removing warnings is Great 1 Vote

21 out of 32 (66%) said it should be against policy.

You can do whatever you want to your talk page = 2 Votes
No you Can't = 14 votes

--Blue Tie 19:07, 26 August 2006 (UTC)[reply]