Wikipedia:Remove personal attacks

From Wikipedia, the free encyclopedia

This is an essay. It is not a policy or guideline. Please update the page as needed, or discuss it on the talk page.
Shortcut:
WP:RPA


This is not official policy. A clear consensus did not emerge from a discussion and vote on the talk page. "The remove personal attacks guideline (and the application thereof) is controversial. It has often been abused by malefactors, and may not have community consensus. It should, at most, be interpreted strictly and used sparingly." - Wikipedia:Requests for arbitration/AI (emphasis original). It is left up to individuals to decide whether to apply it themselves, and if they do they may find themselves held accountable for questionable uses.

Personal attacks are the parts of a comment which can be considered personally offensive and which have no relevant factual content. While simple insults or personal comments can often (but not always) simply be removed, partly factual comments can often be edited to preserve the constructive content, placing the new text in [square brackets].

Contents

[edit] Proposal

Personal attacks are not allowed on Wikipedia. Although users can ignore such attacks, repeat offenders may be banned.

A different approach is discussed here: Instead of resorting to drastic measures, or taking personal attacks for granted, why not just remove or rewrite them? This is a wiki, after all -- discussions on talk pages have disadvantages compared with bulletin boards, but also have the advantage that everything can be "mercilessly edited". See Wikipedia:Refactoring talk pages.

[edit] Refactoring instead of banning

Letting personal attacks linger may contribute to a deteriorating discussion climate, to edit wars, and to users being scared away from editing certain pages. It obscures factual and useful discussion in a mess of accusations and counter-accusations.

When you want to remove personal attacks on article talk pages, you should generally refactor text rather than removing it altogether. "Stop being such a pain! Your writing is awful, let real historians edit this article" is a personal attack, but not free of content. Removing an entire comment is almost always poor form.

When refactoring personal attacks, always start with your own. At the same time, take care not to be one-sided to make yourself appear good and your "opponent" bad; editing your own comments can be misconstrued as a desire to rewrite history.

Check backlinks to your user page to give a list of pages where you've signed your name - see if any of them contain personal attacks written by you, and get rid of them. Leave sensible notes in place of anything you remove.

A discussion of refactoring proposals follows, written largely by those who favor refactoring.

[edit] Where to refactor?

This policy, in all variants, only concerns comments on talk pages. Personal attacks on article pages are of course problematic as well, but should simply be edited in usual NPOV fashion.

Some users feel that this policy should apply to all talk pages, including user talk pages. They argue that personal attacks disrupt Wikipedia wherever they occur, and since no part of Wikipedia is private, every part should be open to editing and refactoring.

Others want to see refactoring limited to article talk pages, because these are more directly linked to the encyclopedia creation process, and read by more people.

[edit] Who should refactor?

Should a person directly involved in a personal attack, e.g. a victim or someone who has expressed strong support for his position on the same page, be allowed to refactor a personal attack?

Some users are strongly opposed to that. They feel that allowing this to happen would make it very easy for personal bias to influence the refactoring process. Others respond that biased refactoring can always be reverted, just like biased edits on an article page. They feel that limiting the refactoring to uninvolved parties will make it so rare that in the few instances when it occurs, whoever does it will be attacked for meddling. They believe that if everyone can refactor personal attacks, those doing it will be justified by a widely accepted policy, and that bias can be kept at a minimum by using clear definitions and restrictions on what can be refactored.

[edit] When to refactor?

Many users consider it a good idea to refactor later - wait until the hot blood has died down a bit, and then get rid of the junk. If you refactor as you go, that can inflame the discussion - in a flame war, people often fail to assume good faith.

Others argue that the whole point of refactoring, under a generally accepted policy, is to cool down a discussion and bring it back to the issues. They feel that when refactoring is generally accepted, there is really not much to get inflamed about -- the user whose personal attacks have been removed should in fact be glad that someone else is cleaning up after them. Under this logic, personal attacks are like spelling mistakes and should be removed whenever they occur.

[edit] When not to refactor

It might be better to not refactor if there is any kind of voting going on. Tampering with comments that are attached to votes may be perceived as in some way tampering with the vote itself. There is also the possibility that refactoring will make disreputable users think that it's ok to change someone’s comments and so abuse this policy. Of course bad edits by people abusing the policy may easily be reverted.

[edit] Don't refactor and rebut accusations

One way to cause a problem by refactoring is to mark up another person's comments while you're engaged in rebutting them. If you change another person's words and then "respond" to them, it's likely to look like you're deliberately setting them up to look bad, or belittling them. Consider the following dialogue:

  • Joe User is abusing Wikipedia yet again by removing monkeyboxes! -- Jane Editor

... changed to ...

  • Joe User is [personal attack removed] removing monkeyboxes! -- Jane Editor

By altering Jane's words and then responding to them, Joe User comes across as a bully who's unwilling to take criticism. Even though Jane's comments were less civil than they could have been, it would have been better for Joe to simply respond calmly.

[edit] What to refactor?

Personal attacks are the parts of a comment which can be considered personally offensive and which have no relevant factual content. Examples:

  • "Your prose is horrible, it could have been written by a third grader."
  • "Once again, you have shown that you have no interest in being neutral."
  • "You clearly don't know what you are talking about."
  • "This is just typical for you and your ilk."

Counter-examples which are not personal attacks -- they are not nice, but they are not affected by this guideline:

  • "This article is complete crap."
    • Does not refer to a specific author.
  • "Wrong! Kaiser Wilhelm was never in Malaysia, and if he had been, he would not have worn a woman's dress!"
    • Factual and not personal.
  • "You Wikipedians can't get anything right."
    • Collective insult -- not nice, but not affected by this guideline.

[edit] How to refactor?

The trickiest part is, of course, how to do it. In general, it is safe to simply remove any comment that is purely a personal attack. Comments that do not reflect a substantive on the subject matter, but merely an attack on other editors, fail to contribute to the discussion. Comments like this can be removed immediately:

  • You moron!
  • Anyone who thinks that is evil.

Editing another's comments is more difficult. It's important to make sure people can tell that comments have been removed; that the words you leave behind are not the original words of the previous poster. One way to do this is to put the new text in [square brackets], as is done in newspaper articles, books etc. when text is not directly quoted.

Including a link to the original revision of the comment helps, too. This makes refactored parts easy to spot and allows interested persons to look up the version history to see what has been changed.

Sometimes, an entire dialogue can be evidently unproductive, for instance when two or more participants descend into unmitigated personal attacks:

  • You fascist! -- Person A
    • You commie! -- Person B
      • You Nazi-loving Hitlerbater! -- Person A
        • You sadistic Stalinist S.O.B.! -- Person B

In such cases it makes sense to remove the entire sequence of attacks, leaving a note such as "flamewar removed" with a link to the relevant entry in the revision history. It's probably not a good idea to do this if you are one of the participants in the flamewar, since the other person is unlikely to take that very kindly.

Remember, the purpose of doing this is not to remove matter offensive to your sight, or to remove one side of a nasty argument. It's to clean up a discussion so that all participants (including the fellow who called you a rude name) can proceed more productively.

[edit] Don't destroy context

Whenever you refactor, do not destroy the context of a conversation -- all statements in the discussion should still make sense after the refactoring. So if you refactor a personal insult, but do not refactor an insulting response, that destroys the context in which the response has been made. Don't do that.

[edit] Answers to concerns

What about accusations that someone's breaking the rules?

Pointing out that a user is breaking a rule is not a personal attack and should not be removed. If someone claims that you're breaking the rules, and you don't think you are, please don't respond by deleting their accusation as a "personal attack". If people can't discuss whether or not folks are following the rules, then the rules would be pointless.

Instead, try to understand their point of view -- including their point of view on the rules. It's your job (as well as the other fellow's) to try to build consensus; deleting each other's words probably isn't going to help here.

If you disagree with someone else on what the rules are, or what should be done in a situation, please do make use of the dispute resolution systems. By inviting more people to take an unbiased look at the dispute, you can avoid the problems of repeated accusations going nowhere.

Shouldn't people who make personal attacks just be banned?

Personal attacks are not allowed and people can be banned for making them repeatedly. Cleaning up a discussion can be a lot friendlier than trying to get someone banned. Sometimes people do benefit from a clean start.

That said, it's important not to use threats to resolve a dispute -- and that includes the threat of having someone banned. "I'm going to delete your attacks -- but if you keep arguing the point, I'll have you banned!" is an unacceptable way to conduct yourself here.

Can't we just ignore personal attacks?

We can certainly try, but it is not that simple. Recent research has shown that social rejection elicits similar brain activity as physical attacks [1]. To oversimplify things, trying to ignore constant personal attacks can be like trying to ignore constantly being hit in the stomach. In case of non-physical attacks on discussion pages, the intensity with which they are felt varies greatly according to the victim's own psyche, whether they consider discussions on the Net fundamentally social in nature or more like abstract conversations with a machine, whether there exists a history of conversations with the attacker, and what the attacker's (and the victim's) standing in the community (in our case, Wikipedia) is.

So what may seem simple to you may be very difficult to someone else in a different situation, or even in the same one. Aside from that, ignoring personal attacks leaves them in the public record -- whenever someone visits the talk page in question, the attacks are there again, creating negative impressions, harming the reputation of the participants and of Wikipedia. They will show up on Google and on every Wikipedia mirror. They are like an untreated lesion.

But I really mean it! That user is a jerk.

See Wikipedia:Staying cool when the editing gets hot and try to talk calmly and reasonably to the user. When nothing else helps to remedy the behaviour you consider problematic, try using one of the established banning or mediation procedures.

Isn't editing a comment a kind of attack?

Yes, it certainly can be! As noted above, people with bad intentions sometimes misuse this process to attack others. [2] Altering another person's words -- even when well-intentioned -- invites accusations of censorship and misrepresentation. Moreover, because removing personal attacks is itself a controversial guideline, any instance of doing so invites further controversy about whether it's appropriate.

Because of this, anyone who chooses to remove personal attacks should do so with caution, and be prepared to defend the removal calmly and respectfully. Getting into bickering matches about whether a specific comment was a personal attack is rarely productive; in such cases it's best to let the disputed comment stand, allowing other editors to judge for themselves.

How do we make sure that facts and context are not lost?

If you think about it, questions like these are already answered on Wikipedia:Replies -- the same principles that apply to general wiki editing also apply to the refactoring of comments and the removal of personal attacks. It's just that we are more careful (and rightly so) not to distort what other people have said and signed. But a principle like NPOV is much harder to maintain consistently than trying to make sure that you do not accidentally refactor a factual statement. If we can write a neutral encyclopedia larger than the Britannica, we can certainly be careful with each other's words!

Aren't we rewriting history?

To put it bluntly, yes. Removing or refactoring discussion necessarily changes the record of what people have said. However, it would be dishonest to misrepresent what another person said. So if you change another person's comments, you're taking on the responsibility of representing their position fairly and accurately.

Remember that anyone who wants to see the full record can view prior revisions, all the way back to the first. One way to ensure that the record isn't obfuscated is to include a note stating that comments have been refactored, and include a link to the version just prior to your changes.

Won't this make banning repeat attackers harder?

Not really. The attack is still in the page history and can be linked directly to the old version of the page. On the other hand, isn't it a good thing if we have to ban fewer people? Consider the possibility that people might learn, through positive reinforcement, that personal attacks will get them nowhere.

What about attacks elsewhere?

Personal attacks on the mailing list, by email, or off-wikipedia, cannot be solved by this process. All we can do is encourage people to stay cool, and not say something that they'll later regret.

How about mediation instead?

You may want to find a mediator and ask him/her to refactor personal attacks in a dispute on your behalf. In general, however, you can think of mediation as prevention and refactoring treatment.

[edit] See also

In other languages