Wikipedia talk:Harassment/Archive 6

This is an archive of past discussions about Wikipedia:Harassment. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 1

←

Archive 4

→

Meta complaint

I'm an irregular editor. After attempting to draw attention to what seemed a serious problem -- an editor apparently who had been systematically editing articles related to an individual to sweep clear from Wikipedia references to prior publicly reported scandals, discovered after yet another public scandal had hit the news, who therefore seemed likely to be either the individual or their agent -- senior editor (administrator?) KoshVorlon We are all Kosh claimed the phrasing would violate "a really strict" policy.

This page seems grossly unclear that enumerating the possibility that the editor might also be the actual individual rather than merely another acting as agent on their behalf would count as "outing" and a form of harassment, and that this policy is "really strict" in enforcement. Abb3w (talk) 15:18, 4 September 2015 (UTC)

WP:OUTING is "really strict" and WP:COI explicitly warns editors that they must not speculate about the identity of another editor. The reason for that is that the community would fall apart if there were real-world consequences of editing, even if only embarrassment. The solution is to speak in code. Instead of saying "Is editor A actually the subject of article B?" one says "The edits of editor A show an undue promotion of the subject of article B; is there a COI?". I emphasized "edits" to show that speculation needs to be based on evidence—you have to provide diffs showing a problem. In general, an issue like this would be raised at WP:COIN, not WP:ANI. Re admin: At the bottom of Special:Contributions for a particular user, you can click "User rights" which will show "administrator" if the user has that right. Johnuniq (talk) 03:08, 6 September 2015 (UTC)

The editor he's talking about is not an admin (it's me ). He's presented this issue on the COI talk page as well as this one. I'll AGF on this and just assume this editor is fairly new, but Abb3w be aware of Forum Shopping . Also, I never stated I was a senior editor or admin on his page, and this is the second time he's made that claim. It was explained to him on the COI talk page that I'm not an admin at all. If you really want to know what he's talking about see his contributions made on Sept 02. KoshVorlon We are all Kosh 17:12, 14 September 2015 (UTC)

Harassment survey

There was a harassment survey on a pagetop banner today. I started to answer it. It might be a good idea, before doing any more such surveys, to do a survey on why people don't complete surveys. There's a basic design fault.

Here's a hint on how to do it better. If answers are wanted, allow people to submit their answers when they get tired of going round the "You must answer this question" loop; because if the only way to end the loop is to close the window, that's what they'll eventually do. If answers are not wanted, don't do the survey.

This is probably the wrong place to post this comment -- sorry -- but I have no idea who put up the survey banner. Andrew Dalby 09:47, 3 November 2015 (UTC)

Hi Andrew, your question has received a response here.Kalliope (WMF) (talk) 14:27, 3 November 2015 (UTC)

Community discussion occuring on Meta

There is a community discussion on this topic occurring on Meta here: Meta:Grants:IdeaLab/Community discussion on harassment reporting. Hexatekin (talk) 20:17, 18 November 2015 (UTC)

Proposed clarification of Outing

About [1] and [2], I suggest adding the sentence: "Stating that such information exists online or elsewhere is as unacceptable as actually providing a link to it or disclosing what the information is, because such statements may lead others to search for the private information."

I can see from the revert edit summary that maybe the sentence should be worded differently, and I'm happy to discuss it. But I will say that I know for a fact that the outing policy is treated this way in practice, and that it would be helpful for editors to be clear about it.

The revert edit summary said in part "what about saying gamergate is discussed at reddit?". That is different than saying something like "I know from looking online that there is information about [name of editor]." --Tryptofish (talk) 00:43, 28 November 2015 (UTC)

Wikipedia works on the WP:DUCK principle—that essay is about sockpuppets but the idea applies generally. For example, there is no list of terms to define harassment—instead, anything which looks like harassment is treated as if it were. Assuming it wasn't known on-wiki, saying "Jimbo is Jimmy Wales in real life" would be outing. Therefore, anything which looks like saying that would also be outing—linking to a page which says "Jimmy Wales is editor Jimbo" would be treated the same as outing, and the policy does not need to be expanded.

The proposed wording moves the goal posts too far in favor of paid advocates and POV pushers. I am concerned that the proposed wording would be used to prevent discussions that are occasionally needed to show that off-wiki activity is responsible for an on-wiki problem. Perhaps a few extra words in the policy might say that any action that reveals personal information is also outing, but I don't think wikilawyers should be given ammunition to hide off-wiki activity if revealing that activity does not directly involve outing. Johnuniq (talk) 02:45, 28 November 2015 (UTC)

Agree. Smallbones_(smalltalk) 03:02, 28 November 2015 (UTC)

Double agree. Flyer22 Reborn (talk) 14:48, 28 November 2015 (UTC)

As the above editors have said, the proposed wording would make dealing with COI/Paid editors all but impossible. Personally I feel the OUTING policy needs to be weakened with respect to dealing with undisclosed-paid-editors rather than strengthened. Often is trivial to know someone is a paid editor from off site information such as job board postings and being prevented from saying 'hey I know you are being paid to edit XYZ because I saw X offer/you posted it in your resume etc. You need to declare it on-wiki' is silly and ultimately harmful to the project. Jbh ^Talk 14:58, 28 November 2015 (UTC)

Thanks everyone, those are helpful responses and I definitely hear what you are saying. But I'd like to please explore a little more whether there can be some better way to address what prompted me to raise the issue. For one thing, please take a look at what ArbCom is currently discussing at Wikipedia:Arbitration/Requests/Case/Genetically modified organisms/Proposed decision#Casting aspersions. Note how they are agreeing that the privacy protection of the outing policy "takes precedence" over any kind of investigation, including COI and paid editing. Thus, any local consensus here needs to be reconciled with what ArbCom is doing, because ArbCom blocks people for exactly what the sentence I previously proposed describes. I don't think it's always an obviously "duck" kind of thing. I get it that there are important kinds of investigations that deal with privacy issues, but I also think that it is established that those investigations require that private information be handled privately, not posted on-Wiki. I'd like to find a way here, that would allow us to make it more clear – because it isn't clear in the current language – that posting information in a way that, in a WP:BEANS-y sort of way, draws attention to the fact that one can find private information if you Google it, is something for which you will be blocked.

What do you think about this alternative possibility? Adding a phrase to an existing sentence: "Personal information includes legal name, date of birth, identification numbers, home or workplace address, job title and work organisation, telephone number, email address, other contact information, or photograph, or instructions for finding personal information, whether any such information is accurate or not." In other words, a posting on-Wiki that does not actually contain "legal name, date of..." and so on, but simply provides a road-map to it, is just as much outing as posting the legal name etc. would have been. --Tryptofish (talk) 17:11, 28 November 2015 (UTC)

If a mechanism is set up for handling COI/Paid issues off-wiki then there is no problem but those procedures need to be developed and published. For instance a group, which needs to include some admins to do blocks, can be established under COIN. Those people should sign the Wikimedia Access to Confidential Information agreements and can handle COI matters which refer to private information. There is still an issue with notifying/warning the COI/Paid editors. I suppose it is possible to simply drop warnings/inquiries with no explanation of why or ask them to contact the group handling private COI/Paid matters. That solution is cumbersome but if we are going to shield people making a profit off of editing here it is really the only way to go forward other than simply saying 'screw it' and giving up. Jbh ^Talk 18:06, 28 November 2015 (UTC)

My understanding (see the ArbCom link above) is that there are existing and known off-Wiki procedures in the form of (1) contacting Checkusers, and (2) sending an email to ArbCom. Beyond that, creating any new kind of mechanism is obviously beyond the scope of what we are discussing here, or can even decide here. I do not see any of that as being an obstacle to what I am suggesting here.

Let's consider this: there are three editors. Editor 1 keeps his or her real name strictly private. Editor 2 posts: "Editor 1's real name is [name]." Editor 3 posts: "Yes, that's right, and all you have to do is to Google it." I think that we can all agree that Editor 2 has violated the policy. But Editor 3 has also violated it, even if Editor 3 thinks "well, it was Editor 2 who posted the real name, and I never repeated the name, so I didn't out Editor 1." --Tryptofish (talk) 18:47, 28 November 2015 (UTC)

What important to remember is that outing is an intentional act done to harass somebody. We cannot codify every permutation. They key idea is, "Don't user personal information to try to intimidate somebody else." If personal information is being used to conduct a legitimate investigation of wrongdoing, we provide some leniency. For instance, if User:HochmanTennis starts adding spam links to Jon's Tennis Shop and Sneaker Emporium, it is fair to look up the website of that business and notice that the owner is one Mr. Jon Hochman, and to explain that the username strongly suggests COI editing. Outing is not a shield that prevents the identification of COI and paid editing. Jehochman ^Talk 21:39, 28 November 2015 (UTC)

As someone who works regularly on COI problems, I'd discourage changes to the "outing" policy which make paid editing easier. Paid editing is a big problem and getting bigger. Paid editors now have to make certain disclosures; see WP:PAID. Some paid editors make those disclosures. Others do so only when pressured. Promotional edits from IP addresses, without disclosure, are a constant problem. Don't make it harder to stop paid editing. Thanks. John Nagle (talk) 22:01, 28 November 2015 (UTC)

Jbhunley, I've been thinking of proposing such a group, and haven't only because of lack of time. It needn't even be people who have signed up to the privacy policy – if they're handling on-wiki material and material found on other websites, there would be no need for that, though it would be an added bonus.

The idea would be that issues too delicate for COIN could be sent to this group. The downside of the idea is that we don't want to encourage people to "investigate" other Wikipedians; if I had to anticipate the argument against setting up such a group, that would be it.

The current situation feels untenable. We're basically having to watch people who have a COI damage articles, and we can't say anything. SarahSV ^(talk) 22:30, 28 November 2015 (UTC)

SarahSV, Brianhe has mentioned something similar as well. I think it would be possible to manage potential 'Wikipedians investigating Wikidiand' downside through defining a solid set of procedures and a code of practice to prevent it being used to further on-wiki conflicts. There would need to be both screening and training for the people involved. (I would also like to see undertakings not to accept compensation for editing/participating in Wikimedia projects.) One reason for having people sign the private information agreements is as a confidence building measure for the community. It would also prevent people from using the group for opposition research or using the material they gain access to (old investigations etc) for other on or off-wiki purposes - writing articles or blog posts etc. I think it would be necessary because, even if material is publicly available, once it is collected into a case the aggregate may be more sensitive.

I certainly agree that the situation, as it stands, is untenable. Jbh ^Talk 23:17, 28 November 2015 (UTC)

Those are all good points, Jbhunley. Screening, training and trust would be essential, and you make a good case for requiring that they sign the privacy agreement. I would also suggest that it be mandatory to cc the editor who is being discussed. That is, the discussion would be conducted like any other COIN, but by email. Also, there would have to be a very good reason to suspect COI before the committee would become involved; otherwise it would be used as a harassment tool. SarahSV ^(talk) 23:27, 28 November 2015 (UTC)

I have to correct what I suggested above, namely that the discussion would be conducted like any other COIN, but by email. In fact, it would not be fair to expect the accused editor to correspond with the accuser, not least because s/he would have no guarantee that the correspondence would not be reproduced somewhere else. SarahSV ^(talk) 23:48, 28 November 2015 (UTC)

Doesn't the COI policy already point in this direction, without stating it explicitly? It states that the outing policy trumps COI investigations (although I've seen this flat-out ignored by some of our more aggressive COI "enforcers.") I do like the idea of a designated, trained group to handle COI off wiki though. As far as Tryp's proposal, I don't think it's a necessary addition since the outing policy already covers opposition research - and is specific enough that we can all recognize outing without listing every possible way an editor can accomplish outing. Minor4th 00:02, 29 November 2015 (UTC)

Break 1

Speaking as an oversighter who has occasionally had to handle oversight requests along these lines, I'm going to be no help at all and say that this is a really fuzzy issue and I'm not sure how we could word it accurately enough to add to the policy. A policy that says that "stating that such information exists online [is outing]" strikes me as far too restrictive (I'm not sure any admin would treat "There is information on the internet about Fluffernutter" as anything other than eye-rollingly silly) - but at the same time, I can also back up Tryptofish on the general statement that there are absolutely cases where referencing outing material without actually quoting it or linking to it will be treated as outing and actioned accordingly.
The trouble is where the line is drawn, and on that I think the fact is that there's a continuum of badness, and it's best handled on a case-by-case basis because continuums are, you know, continuous and really hard to quantify. "There is information on the internet about Fluffernutter" is almost certainly not going to get anyone blocked, because duh. But "there is information about Fluffernutter on [social networking site]" might, depending on how general that site is, the reason the statement was made, and what someone's presence there implies. And "I'm not going to link to what it says about her, but just look at the information you can find if you google 'Fluffernutter + bunny + murder'!" probably will, unless the only thing that search takes you to is this gif set.
The safest way to live your life onwiki is just to never reference offwiki content about anyone else - but even in typing that, I can think of cases where that's overly restrictive ("Fluffernutter lives in the United States" - well, duh, how many American wiki events have I showed up at?), or it's unnecessary (I gave an interview to a news site a few months ago about copyediting on Wikipedia, and gave permission for them to refer to me by username; if someone had mentioned that to me onwiki, I would object to them being yelled at for somehow outing me), or it would just be silly and not to the betterment of the project/user (for instance, the recent situation with a Wikimedian who's imprisoned). There are just too many factors in play for any given situation for me to be comfortable trying to explain it in a single policy sentence. A fluffernutter is a sandwich! (talk) 01:25, 29 November 2015 (UTC)

Thanks everyone for your comments. There's a lot to respond to there. I'll start with the idea, that is really unrelated to what I am trying to discuss here, of having a new group to investigate COI etc. I have two points about that. One is that the sensitive personal aspects of it can already be investigated by checkusers, without creating a new group (bureaucracy creep, and all that), and anything beyond what a checkuser can investigate can probably be discussed on-site. The other is that what I am suggesting here is something that is completely independent from whether such a new group is formed, so I think it would be helpful to treat the new group as a separate discussion.

Another point that I keep hearing is the view that adding anything to this policy will make it harder to investigate things like COI and paid editing. I hope that no one is just reflexively opposing any change because of an interest in prosecuting paid editing. How, exactly, would what I proposed most recently really make that harder?

Personal information includes legal name, date of birth, identification numbers, home or workplace address, job title and work organisation, telephone number, email address, other contact information, or photograph, or instructions for finding personal information, whether any such information is accurate or not.

Considering each of Fluffernutter's examples, I think that change would still work fine with each of them.

Let's say that "Editor 1" is someone who edits without choosing to reveal any personal information of any sort. As you encounter Editor 1's edits, you start to see a disturbing pattern that looks like some sort of COI agenda. You start looking around, and you find another website where someone posting under what is apparently their real name says outright that they edit Wikipedia as "Editor 1" and they are doing it for pay, to advance such-and-such. This is, beyond any doubt, something where you have every good reason to follow up on it. So, how should you do that?

One way would be to send an email privately to ArbCom or the checkusers and point out this information and explain your concerns. All of that would be private. On-site, you could post at the appropriate noticeboard that you are concerned about Editor 1's editing conduct (no mention of the other website), and say that you have also forwarded additional private information to whoever you forwarded it to, and whoever you forwarded it to then posts at the noticeboard that they have, indeed, found incriminating evidence. That's the right way to do it. And it would work with or without what I have been suggesting.

But another way one might handle it would be to go to the noticeboard and post: "At this website (URL), you can see that Editor 1 is actually (name), and is editing for pay." That violates this policy, and you should be blocked for it, unambiguously. And note that this would not be what Jehochman called "an intentional act done to harass somebody." It was done to uphold prohibitions on disruptive paid editing, not to intentionally harass. But it's still outing!

Now let's say instead that you posted: "At this website (URL), you can see that Editor 1 is editing for pay." That's still outing, because it will lead others to find the website where the real name is. You can argue that you did not post any of the following: legal name, date of birth, identification numbers, home or workplace address, job title and work organisation, telephone number, email address, other contact information, or photograph, and thus you were complying with policy, but that will not fly.

So let's say instead that you post: "A search engine search will reveal that Editor 1 is editing for pay." You haven't given the link, and you haven't given the name. But it is still outing, because you have given anyone reading what you said a road-map to the personal information. I guarantee you that editors get blocked for just that. But if you just say that "Some private information confirms that Editor 1 is editing for pay", then you are doing it correctly.

And what distinguishes the correct from the incorrect ways of handling it is, precisely, the issue of "instructions for finding personal information." It still depends properly on clue to evaluate what would constitute those "instructions", but that is the determinant between acceptable and unacceptable postings.

And the third issue that gets raised is that of "duck": that it's obvious to anyone with clue, and we cannot spell out every scenario, and that it's better to keep things simple to avoid providing ammunition for gaming the system. Broadly speaking, that's true, and I agree. But I will put it to you that for some good faith editors who read this policy page in its present form, but maybe do not have much familiarity with all the ancillary discussions about balancing privacy with investigations, they could entirely in good faith post something like "A search engine search will reveal that Editor 1 is editing for pay" without it crossing their minds that this violates this policy. It really is not obvious if all you have seen is legal name, date of birth, identification numbers, home or workplace address, job title and work organisation, telephone number, email address, other contact information, or photograph. Adding a simple phrase would remove that lack of clarity, and I cannot see how it would make enforcement of other policies any harder. --Tryptofish (talk) 20:43, 29 November 2015 (UTC)

I could live with that wording, Tryptofish. With my cynical hat on, I have to admit that I doubt it will entirely save us from the "oops" cases, but I think "instructions for finding personal information" is a pretty accurate description of something that should pretty much always be avoided. A fluffernutter is a sandwich! (talk) 19:52, 30 November 2015 (UTC)

Thanks. I agree with you that there is probably no wording that will completely eliminate such cases, but even a decrease in them strikes me as a net positive. And it really is something that should be avoided, but is something that is not spelled out currently. This is an easy way to spell it out. --Tryptofish (talk) 20:04, 30 November 2015 (UTC)

I would support the wording proposed by Tryptofish. It is relevant to cases of outing that are not about paid editing. Giving "hints" or "instructions" about how to find identity information about an editor who wishes to remain anonymized for Wikipedia editing, on the principle of it, is indeed outing just as much as it would be to explicitly state that editor's identity. nobody should be able to squeak out of an outing just because of using veiled language. If it quacks like a WP:DUCK, even if it is dressed up in sheep's clothing... it's still a duck. This is personal, because it's happened to me, and no action was ever taken in regard to it. It has been used as ammo by some people with apparent agendas to really injure my ability to edit on Wikipedia, even leading up to a current probably topic ban in the recent ArbCom case. So, hinted outing is as real as explicit outing. It's similar to how covert bullying is every bit as potent as overt bullying, perhaps moreso because it's not as easy to recognize or address. SageRad (talk) 19:21, 1 December 2015 (UTC)

Break 2

I think this goes beyond what's said in the policy now.

"Posting links to other accounts on other websites is allowable on a case-by-case basis. The fact that a person either has posted personal information or edits under their own name, making them easily identifiable through online searches, is not an excuse for "opposition research". Dredging up their offline opinions to be used to repeatedly challenge their edits can be a form of harassment, just as doing so regarding their past edits on other Wikipedia articles may be. However, if individuals have identified themselves without redacting or having it oversighted, such information can be used for discussions of conflict of interest (COI) in appropriate forums. If redacted or oversighted personally identifying material is important to the COI discussion, then it should be emailed privately to an administrator or arbitrator – but not repeated on Wikipedia: it will be sufficient to say that the editor in question has a COI and the information has been emailed to the appropriate administrative authority. Issues involving private personal information (of anyone) could also be referred by email to a member of the functionaries team."

You're saying that giving directions to finding posted "links to other accounts on other websites" is *not* allowable. Which would seem to contradict the first sentence above. I'd think if we were to change this part of the policy we could get rid of "The fact that a person either has posted personal information or edits under their own name, making them easily identifiable through online searches, is not an excuse for "opposition research". " What does this do other than make a value judgement, saying it is evil to research who an editor is? Does this apply to paid editors? Apparently not since you can " email (it) privately to an administrator or arbitrator." Except - to the best of my knowledge arbs won't accept this in a case (have they read this policy?) and then any accusation of paid editing will get you banned for making accusations that you can't support. As others have said, the current situation is untenable. Smallbones_(smalltalk) 20:53, 30 November 2015 (UTC)

That's interesting. I won't disagree with you at all about ArbCom sending mixed signals, and I also think that you are right about this passage that you quote from lower on the policy page being internally inconsistent. But I think that "internally inconsistent" is really what it is. The part of the first sentence that says "is allowable on a case-by-case basis" sounds like it was agreed to by a committee, and it is entirely unhelpful in deciphering which "case" is which. (But I do think that what it says about opposition research being inexcusable is talking about posting opposition research on site. If such research is reasonable in investigating disruptive editing, it's supposedly acceptable to email it privately – although I cannot vouch for how ArbCom has responded to it privately.) I gave examples above, including "A search engine search will reveal that Editor 1 is editing for pay." The passage you quote makes it sound like this is something that depends, case by case. And it's not even providing the actual link! It just says that the link exists. But I observed above that such a post will get one blocked by ArbCom, and that is absolutely true. I've seen it very directly, believe me! So I think that what you quoted needs to be revised, but that's a separate discussion from this one. And it's not a good reason to object to the revision that I am proposing. --Tryptofish (talk) 22:17, 30 November 2015 (UTC)

What Arbcom thinks is not relevant for what should be on a policy page—Arbcom follows policy and does not define it. I have seen several cases where one user has either posted personal information about another user, or linked to such information, and all that happened is that the post was oversighted and the offender told not to do it again. I have not followed the Arbcom case behind the proposed change but it is highly likely that a case before Arbcom involved protracted/repeated/egregious violations, not just a brief flurry. There is no need to expand the policy unless the community wants to hamstring itself. A wikilawyer could use new wording to prevent someone at a noticeboard saying "the influx of new editors in this topic may be related to posts on Elance", and the only reason for that would be that someone somewhere at elance.com has said "my name is John Smith [a bogus name] and I edit enwiki". Johnuniq (talk) 01:21, 1 December 2015 (UTC)

All I can find in the proposed decision is: "When investigating COI editing, the policy against harassment takes precedence; it requires that Wikipedians must take care not to reveal the identity of editors against their wishes." Is there a concern that this contradicts the policy? SarahSV ^(talk) 02:45, 1 December 2015 (UTC)

This discussion started over a clarification (diff) added to the policy to alert people about the fact that people have been blocked for posting links to personal information (but see my comment which you replied to—blocking probably occurs for egregious cases). I think the current proposal is to add "or instructions for finding personal information". Of course posting such instructions would be OUTING, and repeating it would result in a block. The question is whether the policy should spell that out, or whether such text would allow too much wikilawyering from off-wiki activists (to protect themselves, all they have to do is post on the off-wiki site "I am John Smith"). Johnuniq (talk) 03:47, 1 December 2015 (UTC)

Yes, it would depend on what the instructions are. If it's "User:Very Unusual Name appears to be on Twitter with the same handle, threatening to disrupt Wikipedia," then that user need only, as you say, write "I am John Smith" on his Twitter page, then request a block for the editor who mentioned Twitter.

The concern might be that this part of the above is misleading, namely "[the policy] requires that Wikipedians must take care not to reveal the identity of editors against their wishes," because the policy doesn't really say that. If someone has posted their identity onwiki and hasn't sought to remove it, then (as the policy says) editors can refer to it even if that person prefers otherwise.

I suppose the word reveal makes it accurate, so long as reveal doesn't include "refer to information already posted by the editor in question." SarahSV ^(talk) 04:30, 1 December 2015 (UTC)

I think this continues to be a very interesting discussion. At this point, I think the main thing I want to point out is that, during the ArbCom case, what is being decided there by the Arbitrators is very much different in spirit from what editors here say that the policy ought to be. I agree with Johnuniq that ArbCom is not authorized to override community policy. On the other hand, the explanation that I have heard from ArbCom is that the central Wikimedia Terms of Use override local policy at any language Wikipedia, and they are correct about that, and the Terms of Use apparently (I haven't checked for myself, but I AGF that this is accurate) place privacy concerns ahead of investigation concerns – absolutely and with no exceptions. For those who are not following the case, the origin of the Proposed Principle that we are discussing here is a chronic problem with some editors accusing other editors of having a COI without providing evidence or attempting to provide evidence. That is seen as basically like casting aspersions, so ArbCom is probably going to say that one can't do that, that instead one has to raise COI concerns through the appropriate channels, not just toss them around. I suggested to ArbCom that they simply link to this policy page, to indicate the appropriate channels, but they seem to be rejecting that suggestion and laying out the proper procedures in their own words, so that is what you see in the link. I'm going to ask them to look here, to see what editors are saying, and explain their own views if they want to.

But there's another fact that is important for me to make clear here, and it's the proximal reason for me starting this discussion. Where Johnuniq said above that violations have to be "protracted/repeated/egregious violations, not just a brief flurry", that has emphatically not been ArbCom's position. They have been very explicit in both discussions and actions (by "actions", I mean ArbCom-issued blocks during the case), that (per the Terms of Use) all it takes for a block is a one-time posting of what I am calling "instructions" and that it does not matter if it was done unintentionally. Let me repeat that: a one-time and unintentional posting is sufficient to require a block. Nothing about protracted, nothing about repeated. I have been motivated in this discussion by wanting to make this policy page reflect accurately what I am seeing at ArbCom. That's what brought me here. --Tryptofish (talk) 18:43, 1 December 2015 (UTC)

I left a note to ArbCom: [3]. --Tryptofish (talk) 18:54, 1 December 2015 (UTC)

To address the last point first, what happens at ArbCom ought to reflect this policy, not vice versa, because otherwise the committee is making policy, which they've said they don't do and don't want to do.

I recall that the ArbCom said of the terms of use, re: paid editing, that they were not able or willing to act as enforcers (though I'm writing that from memory). Also, I wonder whether the terms of use does place privacy above all else.

The spirit of this policy is "don't violate privacy and try not to be unkind about it," so if someone doesn't want you to say "he is John Smith," don't say it, and it's best avoided even if he has already said it himself onwiki. But, especially when the editor has already outed himself, there are cases in which repeating or pointing to the ID is difficult to avoid.

If someone called Rumpelstiltskin advertises for paid work on Elance, and a week later a new account, User:Rumpelstiltskin, creates articles on a small family business and its owners, we ought to be allowed to ask whether those events are connected – even if, buried somewhere on the Elance site, Rumpelstiltskin has posted a link to a page telling us who he is. That's why the policy says: "Posting links to other accounts on other websites is allowable on a case-by-case basis."

"On a case-by-case basis" means that everything depends on context. Was it necessary to offer the information? Was it done with care or gratuitously? Did it happen within an ongoing discussion about COI, disruption and similar? Was the absolute minimum information offered to resolve the issue, or did the poster go beyond what was needed? This all has to be factored in before deciding whether the policy has been violated. SarahSV ^(talk) 19:12, 1 December 2015 (UTC)

Because the edits that resulted in blocks during the case have been oversighted, I'm obviously constrained in what I can say about them, although I am familiar with what the edits contained before they were removed. I think the issue of ArbCom needing to follow the community, and not the other way around, is a good one, and I agree with you about it. What I can say is that the edits did not contain any of what the policy page now lists – "legal name, date of birth, identification numbers, home or workplace address, job title and work organisation, telephone number, email address, other contact information, or photograph" – but did fall in the range of "instructions", that they were arguably germane to the case discussion as opposed to being completely gratuitous, although that's a very subjective call and there are other ways in which it would be fair to call them gratuitous, and that ArbCom stated explicitly that it did not need to have been intentional, as in intended to harm another editor as opposed to providing information for investigation. I'd say that the blocks were in sync with your example about "John Smith", but that they would seem to indicate that what you said in the example about "Rumpelstiltskin" would be that the information must only be submitted privately, and that an on-site post saying that if you look on the web you will find information about Rumpelstiltskin editing for pay, even if you did not say anything about real life identity other than editing for pay and you did not actually provide the link to the website, just saying that a search would reveal it, that would result in a block and would be considered completely unacceptable. So, "I looked on the web and I think Rumpelstiltskin may be editing for pay" is enough to be completely unacceptable, unless it is only submitted privately. But "I think Rumpelstiltskin may be editing for pay, based on his editing patterns" is just fine. The difference is mentioning a web search, or mentioning however else you found information that also happens to include personal information (such as Rumpelstiltskin's real name or the fact that Elance is where Rumpelstiltskin works). It seems to me that "case-by-case" is applied by the current ArbCom differently than what editors are saying here. --Tryptofish (talk) 20:15, 1 December 2015 (UTC)

So why should the community change this policy to suit Arbcom? It should work the other way around. There are external factors such as laws that apply to the person making a post, yet we do not modify policy to reflect what various laws decree. If Arbcom believe some WMF procedure overrides this policy, they can request a clerk to add a hatnote. Meanwhile, there is no reason to race to the bottom and surrender to the activists. Johnuniq (talk) 22:52, 1 December 2015 (UTC)

I guess the shortest answer to that question would be: to make sure editors know not to do what will get them blocked. But your comment deserves a longer reply. I'm not an activist, and I'm new to discussions about outing versus COI, having just been immersed in it at the ArbCom case. It's not clear to me that it isn't the case that we have "activists" on both "sides" of this issue. I'm struck by the amount of strong feelings from editors here who oppose the proposal. But I agree with you that ArbCom needs to engage with the discussion here instead of dictating new policy to the community, and I'm sure that you've seen above that I've advised them of this discussion. It really bothers me that opposition to simply adding "or instructions for finding personal information" means that editors who think that they are complying with this policy could get blocked for it, even though a simple addition would make it sufficiently clear. Telling someone who had been blocked that we didn't make it clear because we don't want someone else to Wikilawyer seems to me to be cold comfort.

I guess that's what we are balancing here: making current practice clear – and policy should always be clear – versus giving disruptive editors a Wikilawyer gambit. And I'm having trouble with your scenario for how that gaming would happen: someone who wants to violate COI creates a post at another site saying that they have a COI so that anyone who calls them out gets unfairly accused of outing. That strikes me as a bogus argument. That would happen very rarely, and what would happen much more frequently is that the other website would simply be where the COI editor provides the evidence that they are violating our policy. It's appropriate to report that other website. But it's just a matter of reporting it privately, instead of posting it on-site. Why is that seen as such a horrible hobbling of COI investigations? Is there some reason that the sky will fall unless editors can post about that website on-Wiki? I keep getting the feeling that the opposition to the proposal arises from an unreasonable demand to be able to post what one wants on-site, and privacy be damned. --Tryptofish (talk) 23:26, 1 December 2015 (UTC)

I think we ought to distinguish between policy and how strict certain admins are about interpreting and enforcing it.

I'd be surprised if someone were blocked for the Rumpelstiltskin example I gave above, because I've seen that kind of thing happen many times, or I've seen editors say: "X posted about this issue on activist site Y, and now four new accounts have arrived." This policy does say: "Posting links to other accounts on other websites is allowable on a case-by-case basis." There was even a proposal recently that paid editors from Elance be required to give their Elance account name. It didn't gain consensus, but the argument was that it followed from the terms of use requiring editors to reveal "employer, client and affiliation."

Tryptofish, you say that it's just a question of reporting information privately, but Will BeBack did that, and it triggered an ArbCom case and became a major factor in his ban. There has been a chill in the air since that case, and it has seriously hampered COI work. Since the Foundation updated the terms of use in 2014, there has been a slight swing back to the pre-Will BeBack position, but only slight, and it very much depends on which admin is looking at what was posted. One ArbCom member did say (on COIN some time ago) that the rule of thumb is to post the minimum amount of information necessary to resolve the issue, but how people interpret that will vary. SarahSV ^(talk) 02:01, 2 December 2015 (UTC)

It should also be clarified whether or not we can include as evidence in a case a link to off-wiki forum posts by a WP user who uses the same pseudonym off-wiki to advocate for a cause but does not reveal their RL id. Unpaid advocacies can be equally as detrimental to the project. There are also cases where new editors might reveal their RL volunteer or work positions not knowing the repercussions of doing so, and then ask oversight to remove all the personal info. Sometimes there are remnants left over that no one is aware of which makes it all the more important to contact that editor privately first. If the editor requests that oversight redact the information their request should be honored. Of course, doxxing needs to remain an unacceptable practice. Atsme^📞📧 02:58, 2 December 2015 (UTC)

SarahSV, you make some very good points, and you make me aware of some things that I did not really understand before. You are quite right about the issue of different administrators interpreting and administering the policy in different ways. You are also quite right about ArbCom being, well, in my own words rather than yours, difficult to contend with. I didn't really follow what happened with Will that closely, so it's quite interesting to me to learn what you just said. Not that I can vouch for ArbCom, but I would hope that reasonable people would make a distinction between private submission of information in good faith, versus private submission of what might appear to be a campaign of harassment, but I obviously do not know what happened with Will. What I will tell you as an absolute fact is that the Rumpelstiltskin example parallels very closely a recent ArbCom block. If that surprises you, well, I guess ArbCom is full of surprises.

I came here intending to make things clearer in that regard, but I'm beginning to see that it may be a waste of my time to do so. And my personal opinion is that it should be fine to submit private information privately without fear of retribution, whereas private information should not be submitted on-Wiki, and that private information includes things like telling other editors to look for other websites. But even if we don't change anything here, at least I will alert any editors paying attention to this discussion that such a Rumpelstiltskin post will be treated by the current ArbCom as a violation of the outing policy.

As for Atsme's question about whether "instructions" are a problem if the external site, in effect, does not contain anything in the way of "personal information" as this policy page currently defines it, it seems to me that it isn't a problem, if the other site really does not have any personal information whatsoever. In other words, if all the other site has is an anonymous screen name, then that's not really outing, although there is also an issue of a "Joe job" or "false flag" (meaning that the person at the other website is someone else, maliciously impersonating the editor here – and in fact that's an issue in any case). My thinking in opening this discussion has been that the problem occurs when going to the other website leads to information of a private nature that is not revealed voluntarily here. --Tryptofish (talk) 19:54, 2 December 2015 (UTC)

What might be helpful, but a lot of work, would be a kind of roadmap, so that editors can see the kind of thing that they may, or probably won't, be blocked for. The problem is that there is inconsistency in how the policy is enforced, so we would have to choose examples of best practice to highlight, which might beg the question. SarahSV ^(talk) 01:59, 3 December 2015 (UTC)

Break 3

Extended conversation about motives behind previous posts, which doesn't add to the actual discussion on Harassment policy. Feel free to continue this discussion on any appropriate user talk page, but with due respect its not really useful here. -- Euryalus (talk) 23:00, 2 December 2015 (UTC)

The following discussion has been closed. Please do not modify it.

As I said above the policy already deals with the issue Tryptofish is trying to address. Relevant portions:

The fact that a person either has posted personal information or edits under their own name, making them easily identifiable through online searches, is not an excuse for "opposition research." and

In extreme cases, such as legal threats, threats of violence, or outing, protective blocks may be employed without prior warnings.

What Arbcom did in Tryptofish's case was according to policy and within their discretion. It has nothing to do with COI, as there was no alleged COI at all. What Tryptofish posted was to the effect: Editors John and Bob are clearly brothers who co-own a laundromat business. Just google their usernames and see for yourselves. John and Bob had not disclosed anything about their personal identifying information. Any alleged connections between the two editors could only be deduced through opposition research. As a result of Tryp's post, another editor apparently followed Tryp's instructions and posted even more specific identifying information about John and Bob. Also relevant is the fact that John and Bob were added as parties to an Arb case and against whom Tryptofish was proposing a site ban, and the post he made was on an Arb workshop page.

Any editor who does not recognize that as outing is either incompetent or is wiki lawyering and gaming the system. Minor4th 16:44, 2 December 2015 (UTC)

Wow, just wow. I cannot believe that you posted that. I'm going to reply briefly here, and then I am going to leave this talk page entirely, so it's up to other editors whether or not they want to continue discussing what I proposed.

Never in anything that I posted here did I ever frame it in terms of my own block or refer to anything about the conflict between you and me. Not once. You brought it up, not me. And in your characterization of my edit, you are blatantly lying about it, so don't anyone else believe it. I suspect that you did it to try to goad me into saying something that I shouldn't, but I'm not going to fall for that. So you have voluntarily said some things that you presumably would not have wanted to say, but I guess that your real agenda was about calling me "either incompetent or is wiki lawyering". Shame on you. --Tryptofish (talk) 20:04, 2 December 2015 (UTC)

You brought up your block, as you have continually talked about it for a month or two. And above you do clearly discuss that Arbcom has a different understanding than the community based on your own block. I want you to think about what that does to an editor who might feel victimized by your outing when you keep posting about the wrongness of the block, over and over again, for months. Stop and think. It has a seriously chilling and intimidating effect - I went through a case of more intentional harassment and outing years ago because of my Wiki editing, and I quit editing for four years because of it. I'll probably leave again because this has become so distasteful that it colors the entire experience here for me. And if you cannot understand that your post was outing, then, yes, I believe you are either being dishonest or are incompetent. It is not hard to understand - don't post things about anonymous editors that leads to revelations about their identity and personal information. How can you not understand that? Minor4th 21:31, 2 December 2015 (UTC)

I never brought it up here, nor anywhere else other than my talk page, which you are free to un-watch. And again, I need to make it very clear that you are lying. --Tryptofish (talk) 22:03, 2 December 2015 (UTC)

I also suggest that an uninvolved editor should "hat" this section, as it contributes nothing to the discussion of this policy page, and the other editor is doing herself no favors per the Streisand effect. --Tryptofish (talk) 22:08, 2 December 2015 (UTC)

Break 4

There's something about this discussion that really confuses me. It seems that people are assuming that sending private e-mails containing personal info to Wikipedia admins, arbs, and other functionaries is prohibited by this policy. It most assuredly is not, for the most part it is not even mentioned. The part of the policy about non-reverted personal info says that you can send that type of info by private e-mail. ArbCom sometimes seems to make the same mistake. The outing part of this policy is about posting this material on-Wiki and other publicly accessible sites. To clear this up I propose adding the following at the end of the outing section.

"Nothing in this policy should be interpreted as prohibiting the private e-mailing of personal information about paid editors or editors with other conflicts of interest to administrators, arbitrators, or other functionaries, or to the Wikimedia Foundation."

This is actually policy right now, simply because nothing is said about this in the policy. But it would clear up a lot of confusion. I'll likely add this to the policy in a few days if there are no objections.

There is one sentence in the policy that might tangentially relate to the above:

"The fact that a person either has posted personal information or edits under their own name, making them easily identifiable through online searches, is not an excuse for "opposition research"."

Frankly I don't know what it means in any operational sense. Does it say that doing any research on an editor here is prohibited? If so it is unrealistic, it would prohibit any enforcement of rules against paid editing, and is generally unenforceable. I'd just remove it.

Smallbones_(smalltalk) 02:43, 3 December 2015 (UTC)

I take it to mean deep probing into an editor's personal life - like posting links to their FB pages, using SYNTH to make a case against them because they have an opposing view, digging into their private domain registrations, etc. I vote to leave it in there with an explanation. Atsme^📞📧 03:08, 3 December 2015 (UTC)

posting info on a publicly accessible site - yes that is prohibited, but how can we prohibit research? Smallbones_(smalltalk) 16:37, 3 December 2015 (UTC)

I support that addition, but Typtofish drew the distinction above between private information sent by email with a good-faith intention to resolve a COI dispute, and private information emailed for the purpose of harassment. It's not always easy to tell the two apart. I wouldn't know how to word the policy so that we allow the first but not the second.

The paragraph starting "The fact that a person either has posted personal information ..." is a problem. I tried to copy-edit it earlier, but found I couldn't understand it. SarahSV ^(talk) 04:06, 3 December 2015 (UTC)

Maybe,

"Nothing in this policy should be interpreted as prohibiting the private e-mailing of personal information about paid editors or editors with other conflicts of interest to administrators, arbitrators, or other functionaries, or to the Wikimedia Foundation. Such information should only be used to help enforce Wikipedia's rules or help in dispute resolution. Making this privately conveyed information public may, in many cases, be considered harassment. "

That's what I mean, but sometimes adding all the littlw details just makes it more confusing. What do you think? Smallbones_(smalltalk) 16:37, 3 December 2015 (UTC)

I like Smallbones' ideas here, and I'm all in favor of making things clearer. At a minimum, I support adding the first sentence, "Nothing in this policy... Wikimedia Foundation." I think a solution to that other sentence, about research, would be: "The fact that a person either has posted personal information or edits under their own name, making them easily identifiable through online searches, is not an excuse for posting the results of "opposition research"." Telling editors what they may not privately read about is obviously unenforceable and unreasonable, so this policy should focus instead on what is posted on Wikipedia. There is indeed also that other issue, about privately submitting harassing material, but I tend to feel that it could be one of those things where we do not go into detail here. The correct way for administrators or whoever to handle that is to say something like: "At this point, I think that the information you are emailing me is too much, so I am asking you to stop it, and to stop sending it to other people as well", and if the editor sending the emails ignores that, then there is a valid reason for sanctions. Warn first. And I think that the improvement that Smallbones has suggested helps a lot with preventing sanctions that might come without a prior warning. --Tryptofish (talk) 17:28, 3 December 2015 (UTC)

How about:

"The fact that a person has posted personal information, or edits under their own name, or is otherwise easily identifiable through online searches, outing policy prohibits posting the results of or conclusions drawn from "opposition research."

. Minor4th 19:13, 3 December 2015 (UTC)

Although I think that "person" has been the longstanding word, it really should be "editor", as OUT is intended to prevent editors from being squelched or retaliated against. There are other policies like BLP and Notability that are supposed to protect the general public from Wikipedia. This actually matters when some quote from a purported expert is given that claims something. You're expected to be able to verify their credentials in a reasonable way, I don't think Google should be off limits in that sort of instance, and I think other policies assume that. Geogene (talk) 23:07, 3 December 2015 (UTC)

The whole thing is confusing. For example, Posting another editor's personal information is harassment, unless that person had voluntarily posted his or her own information, or links to such information, on Wikipedia. Voluntarily posted it where and what is considered "his or her own information"? Posted links where? Is an editor's workplace considered "personal" information? Let's say there's edit warring at Gut flora between User X and User Y. User Y becomes frustrated over what h/she perceives to be ineptitude on behalf of User X when it's really a matter of differing opinions. User Y starts waving his/her own credentials at User X, revealing that h/she has a PhD in clinical microbiology and teaches at a university in New Mexico; therefore he/she holds the trump card to end the argument. User X isn't convinced, so User Y takes it a step further and refers to a cited published online report that supports his/her argument. The report happens to be co-authored by User Y and was published by his/her university. User X reviews the report, connects the dots and then conducts a bit of opposition research. He returns to the article TP loaded for bear. He addresses User Y by his real name and criticizes the report. User Y is surprised and angry over being outed; he didn't think User X would figure out his RL ID based only on a tidbit of info and the link. User X alleges that User Y is POV pushing, has a COI and has self-cited, so he shouldn't be editing the article and should be TB. Ok - what's the verdict based on that scenario? Is the harassment policy clear enough to determine the outcome? Atsme^📞📧 06:04, 3 December 2015 (UTC)

Atsme, although I said yesterday that I should probably bow out of this discussion, I now think that the drama from yesterday has quieted down, so I've changed my mind and come back, and I trust that's not a problem. I think that I can actually answer your questions based on existing policy, but the very fact that you have asked really does indicate that this policy page should be made clearer and more user-friendly.

About "posted links where?", the policy as it is now prohibits posting the links at Wikipedia. I believe strongly that saying on Wikipedia that such websites are out there, if it is said in a way that anyone reading it would be able to find those websites, even without actually providing the link, is just as bad, is treated (at least by the current ArbCom) as just as bad according to the Terms of Use, and ought to be identified as such to avoid editors being confused – but other editors in this talk are pushing back against my suggestion.

Is a workplace personal information? Yes, unambiguously. If an editor says on-Wiki that their workplace is such-and-such, then it's public information, but if they have not done so and anyone else does it, that is outing. There have been real nasty incidents in the past when people have contacted other editors' employers and said bad things about them, and that's awful.

But your example with X and Y shows the gray areas of that. So Y has made public that he or she has that PhD and teaches at a university in New Mexico, so those things are no longer private. But if they did not say which university in New Mexico, then that is still private, although Y showed poor judgment in revealing as much as they did. Now when Y cites that paper, there's a critical point that comes up. If Y actually says that Y is an author of that paper, then they have effectively revealed their identity and there is no more issue of outing. On the other hand, if Y refers to that paper in a different context in the talk page discussion, simply presenting it as an important source, but not presenting it as Y's bona fides as an editor, then for X to put together that Y is an author based on the fact that the author list in the paper shows a university in New Mexico is technically outing, because Y has not said that Y is an author and Y has not said that Y's university is the same university as that of the authors of the paper. Obviously, Y has left him/herself very exposed however. But the correct thing for X to do is to act like X does not know Y's name. When X addresses Y by name, that's outing, and Y is entitled to have X's edit oversighted. X can, however, pursue the COI concerns without outing Y, by pointing to how Y has revealed that they are a PhD at a university and basing it on whatever else Y has said. X can also, presumably, email the information about the authorship of the paper to an administrator, without posting it on-site – but editors are discussing here whether it's really clear that it is safe to do so. That is one of the things that may not be clear enough now. --Tryptofish (talk) 17:17, 3 December 2015 (UTC)

Taking together what editors have been saying, I would suggest adding/modifying two sentences as:

"Nothing in this policy should be interpreted as prohibiting the private e-mailing of personal information to administrators, arbitrators, or other functionaries, or to the Wikimedia Foundation, for the purpose of identifying paid editors or editors with other conflicts of interest, or for resolving other violations of Wikipedia polices."

and:

"The fact that an editor either has posted personal information or edits under their own name, making them easily identifiable through online searches, is not an excuse for posting the results of "opposition research"."

For the first sentence, putting the "purpose" at the end actually helps rule out emailing for the purpose of harassment. And I think the second sentence is corrected for both the issue of editor-not-person, and for posting results of research as opposed to simply doing the research. --Tryptofish (talk) 23:22, 3 December 2015 (UTC)

Break 5

I've added my last version of the "Nothing in this policy" paragraph. More later... Smallbones_(smalltalk) 23:07, 4 December 2015 (UTC)

This is way too broad. At the moment, it implies it is okay to mass email 1000 admins with unsubstantiated allegations about other editors. This can (and will) be really easily abused. For instance, unsubstantiated allegations about an RFA or ArbCom election candidate to various admins likely to be voting. What for instance about mass emails that suggest that an editor is routinely drunk or on drugs or mentally ill and this affects their editing? Plus, and this is critical, the whole notion of anonymous denunciation is likely to be utterly repugnant to the editing community at large. This RFC is way too small to make such a broad change. Remember too that the last RFC on Outing exceptions rejected sleuthing, which this change encourages. Roger Davies ^talk 09:42, 5 December 2015 (UTC)

@Roger Davies: I'm sorry but you seem to misunderstand the current policy. It's all about posting on-Wiki and to some extent on other publicly accessible websites. The references to emailing (before my last change to the policy) are all along the line of

"Issues involving private personal information (of anyone) should be referred by email to a member of the functionaries team,"
"In some cases, the evidence will be submitted by private email" (referring to ArbCom), and
"If redacted or oversighted personally identifying material is important to the COI discussion, then it should be emailed privately to an administrator or arbitrator – but not repeated on Wikipedia: it will be sufficient to say that the editor in question has a COI and the information has been emailed to the appropriate administrative authority. Issues involving private personal information (of anyone) could also be referred by email to a member of the functionaries team."

Please read this seriously. Are there any policies on Wikipedia or in the terms of use that state that emailing an administrator, arb, etc. with private information is against our rules? Is there anything in this policy that says this is against our rules? If so, please provide a link or the text.

Since this policy does not state that privately emailing admins, etc. is against our rules, and since there are no other policies that say that privately emailing admins, etc. is against our rules, then "Nothing in this policy should be interpreted as prohibiting the private e-mailing of personal information about paid editors, editors with other conflicts of interest, or editors violating other policies, to administrators, arbitrators, or other functionaries, or to the Wikimedia Foundation." is current policy. While this does help clarify the issue to people like you who are reading something into the policy that is not there, it adds nothing new to the policy.

Your objection that this will lead to people sending 1000s of emails to admins might be dealt with further (your suggestions welcome), I think "Such information should only be used to help enforce Wikipedia's policies or to help in dispute resolution. Making this privately conveyed information public may, however, be considered harassment." pretty much covers it. Much of anything else would likely be a case of WP:Beans. In fact I'd guess that sending a 1000 emails to admins would be a sure way to get yourself banned, e.g. for clear disruption.

Your link above to the RfC on outing exceptions was all about posting links on Wikipedia, not about emails - which are obviously allowed. If you would like to change the current policy, feel free to start an RfC of your own. Smallbones_(smalltalk) 16:53, 5 December 2015 (UTC)

@Smallbones: This policy is about harassment, which is on the up and up. It can take many forms, direct or indirect. You are very relaxed about the idea of any user cold-emailing any number of admins with personal information about anybody, whether or not it breaches their privacy. That has never been acceptable conduct and you accept that it could lead to banning. One way to discourage mass emailings would be to make each of the referenced groups a singular: an administrator, an arbitrator etc etc. The unfortunate effect of stringing everything together is to offer a plethora of choice and a multitude of people to contact. Another would be to warn that emailing inappropriate material may result in sanctions. It is also worthwhile remembering that violating the privacy of others is explicitly forbidden on our sites (which I guess includes Wikimedia email as it's hosted entirely on the WMF's servers) under the Terms of Use. I'm not suggesting it's removed (as I clearly said above) but it does need to made more circumspect. As it currently stands, the bit along the lines of (I paraphrase) "nothing in this policy prohibits the private e-mailing of personal information" seems to me to be a get-out-jail-free card for TOU privacy violations and that's just plain wrong. Roger Davies ^talk 17:47, 5 December 2015 (UTC)

Hi Roger, thanks for the input. If we're going to have this section, I would prefer that we limit it to COI and paid editing, and perhaps child protection. I think we should also add caution about the community having rejected the idea that editors should be "investigating" each other. SarahSV ^(talk) 17:55, 5 December 2015 (UTC)

Hi SV Emailing to report outing and harassment (on- and off-wiki) is fine too. Roger Davies ^talk 18:06, 5 December 2015 (UTC)

Roger Davies, I'm glad that you joined this discussion, because I think that this is important input. I just reverted the new paragraph at least temporarily, because it clearly needs further discussion before it can be considered to have consensus. From the top of this discussion (before the section breaks), I have been concerned that editors in this talk may have been too resistant to anything that they perceive as privacy-based constraints on investigation. Could you please clarify for editors in this talk what the Terms of Use establish with respect to the balance between privacy and investigation of COI?

I agree with Roger that, at a minimum, we should word it in the singular, to rule out mass emailings. At this point, it sounds like it is appropriate to use email for reporting outing and harassment (and indeed to request oversighting), reporting COI and paid editing violations, and seeking protection of minors. As long as the emails are done in good faith, it's hard to see anything wrong with such use of email. It also seems clear to me that it is, instead, disruptive to use email to accuse someone of being a drunkard etc. But there are issues that fall in between, and we need to make it clearer. For example, I can think of situations involving sockpuppetry and meatpuppetry where email should be considered appropriate, and to be preferable to posting publicly. I also would worry about coming down too hard on an editor who believes in good faith that they are emailing something that needs to be investigated. Earlier in this talk, I suggested that improper emails should, initially, be met with a warning. I can easily see sanctioning an editor who continues to misuse email after a warning, but are there any circumstances when an editor should be sanctioned without warning after a single email? --Tryptofish (talk) 18:40, 5 December 2015 (UTC)

Suggested edits to the proposal:

Nothing in this policy prohibits the emailing of personal information about editors to administrators, functionaries, arbitrators, or the Wikimedia Foundation, when doing so is necessary to report conflict-of-interest or paid editing, harassment, or violations of the child-protection policy. Only the minimum information necessary should be conveyed and the minimum number of people contacted. Editors are warned that the community has rejected the idea that editors should "investigate" each other. Posting such information on Wikipedia would violate this policy.

SarahSV ^(talk) 18:45, 5 December 2015 (UTC)

Pretty good. I changed "e-mailing" to "emailing". "Editors are warned that the community has rejected the idea that editors should "investigate" each other" might be too broad. Does that rule out reporting by email something you saw on elance? Smallbones_(smalltalk) 18:51, 5 December 2015 (UTC)

(edit conflict) I would also make it singular, for administrators, etc. And I still have a question about, for example, sockpuppetry. I don't think that that should be considered improper to report by email. --Tryptofish (talk) 18:53, 5 December 2015 (UTC)

Smallbones, no, it wouldn't rule that out. SarahSV ^(talk) 18:56, 5 December 2015 (UTC)

In that case I'm all for it. I've tried editing out the plural forms, e.g. "administrators" and it comes out looking like you have email exactly one, or exactly 3 people which doesn't work well. The "Only the minimum information necessary should be conveyed and the minimum number of people contacted," does a better job. I'm not against including "sockpuppetry" but there are other ways to report that. Let's include Sarah's version and go back to the odd first sentence below. I include the second sentence only since a word or two might have to be changed if we remove the 1st sentence.

"The fact that a person either has posted personal information or edits under their own name, making them easily identifiable through online searches, is not an excuse for "opposition research". Dredging up their offline opinions to be used to repeatedly challenge their edits can be a form of harassment, just as doing so regarding their past edits on other Wikipedia articles may be"

I'd say Sarah's "Editors are warned that the community has rejected the idea that editors should "investigate" each other," covers what I imagine the first sentence means. So just delete that odd first sentence. And the second sentence could be "Editors dredging up the opinions of other editors, whether online or offline, to be used to repeatedly challenge their edits can be a form of harassment, just as doing so regarding their past edits on other Wikipedia articles may be."

Smallbones_(smalltalk) 19:29, 5 December 2015 (UTC)

Break 6

Way back in #Break 1 (ah yes, I remember it like it was only yesterday), I proposed expanding a sentence to say:

Personal information includes legal name, date of birth, identification numbers, home or workplace address, job title and work organisation, telephone number, email address, other contact information, or photograph, or instructions for finding personal information, whether any such information is accurate or not.

I still believe that something like this is needed to make the policy page accurately and clearly reflect current practice, but the discussion has taken a lot of detours, so I want to return to the original question.

@Roger Davies: @Euryalus: You are the two Arbs who have edited this talk page while the discussion has been going on, so I would like to ask you. How do you see this suggested addition in the context of how ArbCom currently views the issue? --Tryptofish (talk) 20:31, 5 December 2015 (UTC)

I can't speak for ArbCom as a whole but the spirit of policy (which is what this proposed change is about) has always been as important as the letter of it. Roger Davies ^talk 08:40, 7 December 2015 (UTC)

So if I understand you correctly, you seem to agree with me that the proposed change accurately reflects the spirit of the policy. That being the case, it would seem to me that the worst one could say about the proposal is that it isn't absolutely necessary. But that isn't the same thing as saying that it would cause harm, and I still cannot see any reason why it would not make things clearer to say this explicitly. --Tryptofish (talk) 19:30, 7 December 2015 (UTC)

Also, most of the concerns that were raised previously had to do with the risk of hampering investigations of COI etc. Since that time, the policy page has been revised to make it explicit that email submission of evidence is permitted. It seems to me that those changes take away the ability of a COI editor to Wiki-lawyer against legitimate investigations, and perhaps that removes the concerns about my proposal here. --Tryptofish (talk) 17:37, 8 December 2015 (UTC)

Offline or Online?

"Dredging up their offline opinions..." Shouldn't that be "online"? -- BullRangifer (talk) 05:59, 8 December 2015 (UTC)

Obviously, that's a mistake. I suspect that it should be neither, but instead should be "off-site", since that's really the key point. --Tryptofish (talk) 17:34, 8 December 2015 (UTC)

Yes, that's even better. -- BullRangifer (talk) 19:53, 8 December 2015 (UTC)

Harassment and disruption: suspect edit

This sentiment "The prohibition against harassment applies equally to all Wikipedians. It is as unacceptable to harass a user with a history of foolish or boorish behavior, or one who has been blocked, banned, or otherwise sanctioned, as it is to harass any other user. Wikipedia encourages a civil community: people make mistakes, but they are encouraged to learn from them and change their ways. Harassment is contrary to this spirit and damaging to the work of building an encyclopedia" was added in September last year with this edit by an editor who is no longer with us, with the edit summary "“Language from WP:CIVIL and WP:PERSONAL policies".
I have removed it, as there was no discussion that I can find on the need for it; nor can I find the wording in WP:CIVIL or WP:PERSONAL from which it is supposed to be copied (and if it is there I can't see why it needs to be repeated here); nor (considering the the person adding it was well on the way to the door at the time) would I particularly trust the reason for its addition. As it stands it suggests that responding robustly to vandalism, or sockpuppetry by banned users, could be classed as harassment, which I would argue with.
If anyone feels it should be retained, perhaps they could indicate why, here. Xyl 54 (talk) 23:03, 29 December 2015 (UTC)

I agree with deleting it, as it does sound more like editorializing than policy. On the other hand, the basic idea behind the first two sentences is a reasonable one, if one were to prune the catalog of types of users from the second sentence (perhaps replacing the long list with, simply, "a history of disruptive behavior"). --Tryptofish (talk) 23:29, 29 December 2015 (UTC)

The language is very similar to that at WP:CIV#No personal attacks or harassment and WP:PA#Why personal attacks are harmful, which have both been there for years. What they are saying, in other words, is that edits should usually be judged on their own merit, not suspicions about the editor who made them. I fully support this principle and oppose deletion. Reverting the edits of obvious vandals and sockpuppets is reasonable enforcement of policy, not harassment. Burninthruthesky (talk) 10:32, 30 December 2015 (UTC)

Thanks for replying; I was going to say:

@Tryptofish:: yes, editorializing was probably the word was looking for. And yes, the list was part of what was troubling me. If we were to keep any of it, the first sentence “The prohibition against harassment applies equally to all Wikipedians” and maybe parts of the third and fourth “Wikipedia encourages a civil community; harassment is contrary to this spirit and damaging to the work of building an encyclopedia” would be OK.

@Burninthruthesky:: Thank you for the links; part of me would say that if this sentiment is already stated in at least four other places, is there any real need to repeat it here? But (my first reason for removing it) there wasn't any discussion on adding it originally, and it seems a pretty cavalier way to create binding guidelines just letting anyone chuck stuff in. I'd also note that the wording here goes some way further than the format on the other pages; if we are to have some expression of this view here then it'd make better sense to copy it verbatim. But I'd still be interested in a justification for adding it here.

And I would agree, reverting edits of vandals and socks should be reasonable enforcement, but that isn't (to my mind) how it reads. If we are to have a statement of this principle here, it needs a caveat along those lines.

Any thoughts? Xyl 54 (talk) 20:13, 31 December 2015 (UTC)

As I said earlier, I was OK with your deleting of it, and I'll say now that there wasn't really consensus for putting it back. But I shortened that editorializing list, and I'm pretty comfortable with it, the way it is now. Maybe we could shorten it a bit further, but I think it is OK to leave some of it there. --Tryptofish (talk) 20:42, 31 December 2015 (UTC)

Firstly, I apologize for my mistake in assuming this discussion had finished. My thinking was that the principle had had silent consensus in this document for over a year since it was WP:BOLDly added, and much longer in its other locations. Yes, it is redundant to an extent, but there is an inherent overlap between our conduct policies and I think some parts are sufficiently important to be explained in all of them. I've compared again the language used here and in the two places I linked above but nothing particularly strikes me as going "some way further than the other pages". Can you elaborate? Burninthruthesky (talk) 21:18, 31 December 2015 (UTC)

I don't quite follow how this language could be read as implying "responding robustly to vandalism..." is harassment? We shouldn't be harassing anyone, including vandals - in the latter case it's probably better still to just WP:RBI. Burninthruthesky (talk) 21:18, 31 December 2015 (UTC)

@Tryptofish: @Burninthruthesky: Well, I should probably apologize also, for not responding sooner, but, no harm, no foul.

The wording as it stands is OK by me, if having discussed the matter the conclusion is that some form of words there is necessary (and, looking at it again, WP:PA does specify “personal attacks and harassment...” so probably something should be here, if anywhere).

What I was bothered about was the appearance of a list of editors “a user with a history of foolish or boorish behavior, or one who has been blocked, banned (highlighted with a link), or otherwise sanctioned ...” who seemed to be offered special protection (and yes, I acknowledge it is only the same protection as everyone else gets, but it is spelt out here). The list isn't there now, so I'm fine with it.

I was mindful also of a recent TfD where the BMB template (which to my uninformed mind seemed only to be giving notice that BMB was being enforced in someone's case) was deleted because it constituted "harassment".

I get the fact that we should maintain the moral high ground; that if someone is behaving like an arsehole we should refrain from calling them one; that if that person stops behaving so and gets their act together we should refrain from being vindictive.

OTOH the concern is, if I come across a bit of vandalism and revert it, that is merely RBI; but if I then check the editors contributions and find more, and then revert that, or simply rollback everything if indicated, that could be construed as harassment; and if I put a warning on their talk page, and keep an eye on them, that would more or less confirm it.

But isn't that the responsible thing to do? Xyl 54 (talk) 12:47, 3 January 2016 (UTC)

Yes I agree keeping an eye on vandals is the responsible thing to do. To me, policy is clear that that is not harassment. I'm fine with the wording either with or without the recent edit. Burninthruthesky (talk) 15:27, 3 January 2016 (UTC)

I think that this ended up working out just fine, and thanks to both of you. I agree that the biggest problem with that earlier language was that list ("boorish behavior"? really?), and that is why I deleted that part, and I think it is OK now. As for watching vandalism, simply checking whether or not a vandal has also made other vandalism edits that need to be reverted should in no way be construed as harassment, and it would be downright bizarre to insist against reverting vandalism or BLP violations. The line into harassment would only be crossed if the reverting editor was either aggressively incivil to the vandal over a period of time, or if good edits were also being repeatedly reverted as a sort of payback. But all of that should be self-evident.

As for something that might be less self-evident, I'd like to invite each of you to consider commenting on the issue that I raised in the talk section just below. Thanks! --Tryptofish (talk) 20:26, 3 January 2016 (UTC)