Archive 15 ← Archive 18 Archive 19 Archive 20 Archive 21 Archive 22

RfC: Clarification of OUTING

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.
TLDR
This RfC presents a disagreement on whether it should be a violation of our harassment and outing policy when User A posts personal information about User B when User B has previously voluntarily publicly shared that personal information on another Wikimedia wiki. Supporters of Option #1 believe that policy should not prohibit User A from doing so. Supporters of Option #2 believe that doing so is still outing and should constitute a violation of policy. An alternative proposal made after the beginning of the RfC tries to strike a middle ground, and would permit users to refer to information voluntarily shared on the English Wikipedia, Wikimedia Commons, Meta-Wiki or Wikidata.
The current policy, many participants assert, is ambiguous, and there is disagreement as to what it currently allows; some participants believe that the plain language of the policy currently allows references to information shared on other-language Wikipedias (such as the French Wikipedia), while others believe it only applies to the English Wikipedia. Still others believe that SUL projects are already functionally "Wikipedia" for the purpose of this policy and the policy as currently applied does not prohibit references to information voluntarily posted on other public Wikimedia projects.
24 users supported Option #1, including one user who would also prefer no change and one user who only commented in the alternative proposal area but clearly meant to support #1, and excluding one additional user noting that they would "trend" towards Option #1 if Wikipedia were starting from a blank slate. 22 users supported Option #2, including one supporting it "weakly", one advocating a prohibition on all links/references to personal information, and one preferring no change.
Of the users commenting on the alternative proposal, 14 supported, including four second-choice preferences to Option #1, six second-choice preferences to Option #2, two additional supporters of Option #1 (including one who "weakly support[s] at best"), and one additional supporter of Option #2. 14 opposed, including six users who support Option #1 and six users who support Option #2.
In short, this was a close RfC. Normally, in such a close policy RfC, the default is to close as "no consensus" and let the current policy remain. In this case, there is a lot of disagreement as to what the current policy means, so even if none of the words of the policy change as a result of this RfC, it would be ideal to have a definitive statement on what the community will allow and what it will sanction at a noticeboard as a matter of practice. Of course, this does not 100% rule out a "straight" no consensus close, but it makes such an outcome less favorable as I begin my analysis.
At the outset, I should note that it was clear to me that supporters of Option #1 and Option #2 were generally looking at the policy from different lenses, with some potential misunderstandings between the two: Option #1 supporters seemed to be thinking substantially from the perspective of someone who might run afoul of a strict policy here (e.g. someone who’s worked with a user on multiple wikis and usually uses a real name on other wikis and mentions the user’s real name on enwiki by accident, or someone who is cleaning up spam/PAID violations and sees a Commons "own work" declaration for a company-produced photo and wants to use that as evidence for TOU enforcement on enwiki). Option #2 supporters seemed to empathize more with the perspective of victims of harassment and outing, who might find their edits on other wikis "researched" and pored over for personal information. One exchange toward highlighted this difference in perspective: one editor worried that "If some good-faith editor fails to read the policy change and violates the new policy, he/she will get indeffed, although he/she meant no harm. E.g. outing someone who disclosed his/her own identity at de.wiki." An administrator disagreed in reply, noting that in her experience, administrators have much more discretion and may choose shorter blocks, or no block, for good faith violations. This flexibility is how outing enforcement works in practice and guides the outcome of this RfC, as where there is no clear consensus to change it, our policies are meant to reflect our accepted practices.
The arguments behind Option #1 are pretty simple to boil down: All the Wikimedia sites are linked, and if someone has voluntarily disclosed information about themselves on another Wikimedia site, we should be allowed to reference it for the same reasons that we should be allowed to reference information voluntarily disclosed on the English Wikipedia. It’s a unified login, after all, and many Wikimedians interact with each other on multiple wikis, all of which have the same visibility and access as the English Wikipedia. Supporters note that there are legitimate reasons that editors might reference personal information voluntarily posted on other wikis, including to help enforce policies like PAID or copyright; to paraphrase, users should not be able to immunize voluntary public statements simply by posting them on another Wikimedia wiki.
Supporters of Option #2 had a number of objections to Option #1. Supporters note that different projects have different policies, cultures, norms, and protections surrounding disclosure of information, and it’s possible that posting information about others on one project might be treated very differently than on another. Supporters argue that editors might expect that the projects are separate entities and that some might not want to apply the same standards to all wikis; that it might be very easy to make a mistake and accidentally reveal one’s identity on another wiki; that adopting Option #1 would encourage harassers to fish through other Wikimedia sites for personal information; that they wouldn’t want information disclosed on enwiki discussed on other Wikimedia wikis and thus we should adopt the inverse rule to protect those who work on other Wikimedia projects; that some Wikipedians would prefer to go back to a time before SUL. Supporters argue that Option #2 is easier to administer and enforce. And in any event, some supporters argue that even if some of these individual problems are addressable, our dedication to user privacy means that it would be safer to err on the side of privacy (the precautionary principle, as one editor referred to it) when in doubt in drafting our policies.
Supporters of Option #1 noted, in response to some Option #2 supports, that Option #1 refers only to information that was voluntarily publicly disclosed by the user in question. Further comments clarified an understanding that Option #1 refers only to non-enwiki accounts that have been voluntarily, publicly disclosed to be the same user as the enwiki user (typically through linking on SUL), and at least one Option #2 supporter said they would "be amenable to a version of #1 that clarified that only publicly connected accounts are included".
The alternate proposal was supported by some users from both sides. Its supporters note that it addresses the problem of users who consider different cultural contexts and share information (e.g. their name) in the context of a different linguistic community that they would be opposed to sharing here. It was few users’ first choice, though, and many participants on both sides opposed the alternate proposal. Overall, it was clear that the alternate proposal did not receive consensus and was not seen as a good compromise between the two options because it failed to address many of the concerns each side had with the other proposal.
Stepping back from the specific wording of the proposals, there was a lot of common ground which helps define what the community wants to allow and prohibit, even though at the surface appears substantial disagreement. Many participants on both sides agreed (implicitly or explicitly) that editors should not be digging through ("researching") old contributions to try to find outing or harassing material; many participants (except those who oppose any links or references to personal material, from enwp or not) also seemed to agree (implicitly or explicitly) that editors should not be penalized for information posted on a global/meta user page, or places of similar prominence.
I found it noteworthy that very few participants chose to disagree with the current rule that if someone posts "their own personal information, or links to such information" (emphasis added), at least on the English Wikipedia, then referencing that information is not outing. If the provision about links to information remains policy, interesting results may ensue: What counts as a link? If I say See my meta talk page for a relevant discussion, does that count as a link? What if I use one of those charts with my accounts across SUL? How about a ((User admin somewhere since)) userbox? Or an essay in my userspace on another wiki? Link to a grant proposal that clearly says my Meta account is the author? If links to other wikis don’t count, that also has interesting consequences: if I link my Twitter handle and my Meta userpage, is it fair game to mention things I put on Twitter, but not Meta? That result seems inconsistent with the intent of our policy. And just as importantly, isn’t the point of SUL that all the accounts are linked? By voluntarily using the same SUL account on multiple wikis, by voluntarily merging them, isn’t that voluntarily posting links to the other accounts? I’m aware of no policy that says you can’t use one username on the English Wikipedia and another on Wikidata. This isn’t just a discussion about the technicalities of policy; this exception reflects the purpose of the outing policy in the first place, and the broader community’s intention that voluntarily publicly connecting one’s English Wikipedia account with another link or identity makes it much less categorically unacceptable to refer to that other identity.
Ultimately, when determining the outcome of this RfC, because our policies should reflect our community’s actual practices, I have envisioned some scenarios and imagined the outcome of a noticeboard thread had the participants at this RfC responded to it, based on the arguments and positions presented at this RfC:
  • It’s likely there would be consensus to sanction for harassment a user who has searched through ("researched") the contributions of someone on another project, especially in the context of a dispute or apparent retaliation, or any other factor that would make it seem like the personal information is revealed with the intent to harass. A revert-revdel would likely have consensus too.
  • It’s likely there would be no consensus to sanction a user who revealed personal information voluntarily posted on another project in good faith (e.g. forgot where they learned it, or thought it was public on the English Wikipedia) or used it in order to enforce policies and norms (e.g. Commons author information for PAID enforcement). In some circumstances, a courtesy revert-revdel may be appropriate. In some cases, depending on circumstances, the community would trout for bad judgment in posting personal information.
  • It’s likely there would be no consensus to sanction a user who posted personal information first voluntarily disclosed on another project in a prominent location, or with the clear intent of dissemination, or which was later linked to on the English Wikipedia.
A number of Option #2 supporters argued that the precautionary principle should apply in privacy matters, which can be taken as an argument against a more equivocal, circumstances-determine-the-outcome policy approach. This argument is real and reasonable. However, that’s not how most of Wikipedia works; because, as I’ve set out above, there are circumstances that the community probably tolerates or even accepts sharing personal information voluntarily revealed on other projects, a bright-line rule against such sharing either would be inconsistent with practice or would result in acceptable/tolerable actions being now viewed as outing. We do have bright-line rules on Wikipedia, but only when all (or nearly all) of the underlying conduct is considered harmful, such as 3RR. And even then, there are a litany of codified exceptions from which arise somewhat frequent disputes (such as 3RRBLP). All in all, there is no consensus for a bright-line rule that posting personal information about an editor is outing, even if it was voluntarily publicly disclosed on another Wikimedia project.
Having considered all of the above, the result of this RfC is:
  • It remains a bright-line violation of OUTING to post personal information that has not been voluntarily disclosed on a public Wikimedia project. It is also a bright-line violation to post personal information if it has only been disclosed on accounts that are not publicly connected to the English Wikipedia account in question (i.e. by SUL).
  • Circumstances will determine whether a particular edit referencing or revealing personal information that was not previously posted on the English Wikipedia is harassment in all other cases. It is generally more acceptable to reference information if it is clear the user does not mind wider dissemination (e.g. posted on a user’s public userpage) and less acceptable if it requires much "research" to find (particularly information later removed by the user in question). There was no consensus that referring to such information is always or usually a violation.
  • Editors are urged to take care to err on the side of privacy, and to ask users before posting their personal information if there is any doubt. This RfC does not constitute a green-light to snoop around, and in fact demonstrates the community’s continued emphasis on the importance of privacy. One sentiment that arose was that posting information which might not constitute outing per se, such as information shared on another project, can still be very unwise and reflect poorly on the poster’s judgment.
  • Administrators continue to be entrusted with the discretion to act with judgment, including e.g. revision-deleting personal information without finding that the user who had posted it had intentionally engaged in outing and blocking them as a result.
I will be making some changes to the harassment policy to implement these results, but the consensus was for the principles as established above, not for any specific wording that I will be adding to the policy. I’d like to take this chance to thank everyone who participated in this RfC; it’s a difficult, personal topic, and I’m grateful for everyone who contributed so thoughtfully. Also, I’d like to apologize for the delay in closure; I revised and rewrote this closure a few times. If anyone has any questions, please don’t hesitate to let me know. Best, Kevin (aka L235 · t · c) 01:25, 7 September 2019 (UTC)

Current wording of WP:OUTING:

Should this be revised to instead say one of the following:

  1. unless that person has voluntarily posted their own information, or links to such information, on the English Wikipedia or another public Wikimedia project.
  2. unless that person has voluntarily posted their own information, or links to such information, on the English Wikipedia.

(See also previous discussion at Wikipedia:Administrators' noticeboard#Question about outing) GMGtalk 19:04, 13 May 2019 (UTC)

Survey

Unless an individual has chosen to work in Wikipedia only under their real name, the correct way to address that individual is by their chosen Wkipedia-pseudonym.
We are here to improve the encyclopaedic content of Wikipedia. That's it. The social interaction encouraged by the existence of user pages and talk pages, as well as the existence of community projects, often confuse people. Well, Wikipedia is not a message board. -The Gnome (talk) 09:20, 17 May 2019 (UTC)
I understand the argument that someone using a unified log-in account and declares their identity on one project should be considered to have waived their interest in keeping that info secret on all projects, and should not be able to rely on a bureaucratic technicality to call out someone else who references that information in good faith. However, if we adopt option 1 without further caveats, we are clearly opening the door to doxxing. I almost !voted for option 2 as a consequence, just to pursue the precautionary principle on this issue and avoid creating the possibility of people using the technical wording to avoid sanction for clearly inappropriate disclosures of another's private information. But I think with some effort, we can come up with wording that clearly expresses a rule that people with expressly linked accounts are deemed to have disclosed their own identities on en.Wikipedia if they have disclosed it on any Wikimedia project, but only where the accounts in question were clearly meant to be expressed as a common identity by the person operating them. It is a clear principal of the broader Wikimedia community that people should have the right to operate different identities across multiple projects if they feel this is essential for their privacy and well-being--provided of course that they are not abusing multiple accounts for any disallowed purpose on a given project, of course. its essentially that we safeguard this broader Wikimedia community right against any inadvertent eroding as result of rule creep here that is meant to clarify one of our local policies, even one as important as our harassment policy. Snow let's rap 22:01, 8 June 2019 (UTC)
Changing my !vote to #2. You know, with the benefit of a little more time to think about this, I've decided that the safest course of action to preserving the privacy and safety principles embraced by the broader Wikimedia community (and indeed, the local en.Wikipedia community) is to just utilize a simpler standard here. I do believe there is a way to approach option 1 in such a way that those interests are protected, but A) I don't want to endorse an option which may not receive that additional protective language, and B) writing that language such that the standard couldn't be abused would be difficult. So I'm shifting my preferred approach to be in line with the KISS principle. Snow let's rap 22:21, 8 June 2019 (UTC)

Alternate proposal

Per discussion above and below, I'm proposing an alternate compromise wording:

unless that person has voluntarily posted his or her own information, or links to such information, on the English Wikipedia, Wikimedia Commons, Meta-Wiki or Wikidata.

*2nd Choice to #2 above.. Mostly support though. Tbh I feel like this is probably going to be the best choice in the end. (Also I didn't forget to sign this time c:) --NikkeKatski [Elite] (talk) 13:56, 20 May 2019 (UTC)

Unless an individual has chosen to work in Wikipedia only under their real name, the correct way to address that individual is by their chosen Wkipedia-pseudonym.
We are here to improve the encyclopaedic content of Wikipedia. That's it. The social interaction encouraged by the existence of user pages and talk pages, as well as the existence of community projects, often confuse people. Well, Wikipedia is not a message board. -The Gnome (talk) 09:20, 17 May 2019 (UTC)
The Gnome Let's play that out. Say someone has identified who they are on a different Wikimedia project. They come here, and, bypassing AfC, write an article about themselves. This would be promoting themselves but we would be banned from pointing this out because to do so would be OUTING by your standard. How do you reconcile this? Best, Barkeep49 (talk) 02:27, 4 June 2019 (UTC)
Exposing a violation of policy should never be inhibited. Therefore, that would be the sole exception to the rule I support, since, by creating an article about themselves, the editor has violated Wikipedia policy. Thanks for this very valuable comment, Barkeep49. -The Gnome (talk) 09:02, 4 June 2019 (UTC)

Discussion

Note concerning comments on Unified login above: Should I remove my email account, cancel my global account, and place a disclaimer on my user page? I only have one user name. I suppose the "User committed identity" is also not a good idea? I started the Wikipedia:Unified login ("allows users to use a single global login on all public Wikimedia Foundation projects. This allows users to maintain a consistent identity throughout Wikimedia and work on different projects without having to sign up and log in to each project individually.") for ease of navigation. It now seems this could have been a bad idea. I HAD NO IDEA that by linking accounts there would be the remote thought I would be subjected to possible privacy infringement or that somehow I have given some proxy agreement that it might be alright just because I linked accounts. I sort of had the apparently mistaken idea, mentioned by user:David Eppstein "I firmly believe that editors should be allowed to edit pseudonymously here without fear of being revealed", and I thought this was a given but apparently not. I would have though there was a concrete expectation that personal private information ALWAYS BE PROTECTED ---EVERYWHERE on Wikipedia or connected Wikimedia Foundation projects to the best of our ability. If any editor chooses to use a real name, or otherwise reveal their real identity, that would be their prerogative but it still boils down to my above comments that outing should not occur. Since there are possible legal implications here should we seek a WMF opinion? Otr500 (talk) 04:10, 9 June 2019 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Implementing the RfC outcome

Hello all, I recently closed the OUTING RfC above. It's getting fairly late here and I'm having trouble translating what I wrote into actual edits to the policy, so if anyone wants to take a crack at implementing the RfC closure, I'd be grateful. (Alternatively, perhaps it doesn't need to be really integrated into the text of the policy; a footnote quoting an excerpt from the close might do just fine.) Best, Kevin (aka L235 · t · c) 04:42, 7 September 2019 (UTC)

I have added a footnote with some of the conclusions of the RfC, but would appreciate anyone who has time to integrate it into the text or at least make the note more elegant. Best, Kevin (aka L235 · t · c) 22:23, 10 September 2019 (UTC)

"Overriding reason" vs. "constructive reason" in WP:HOUND

In June, Tryptofish changed "overriding reason" to "constructive reason." I think the long-standing "overriding reason" is better because it's saying "for no essential reason/unnecessarily." In other words, going by the definition of "overriding," it's saying for a reason that is not about adhering to a policy or guideline. By contrast, saying "for no constructive reason" is subjective because an editor may feel that they are being constructive by following the other editor around. It's been made clear countless times before that an editor following another around because that editor thinks they are being constructive is not enough of a reason to keep following the editor around if it's annoying the other editor/causing the other editor distress and the following isn't actually necessary/helping at all. Besides, disruption to the project is usually considered unconstructive.

We could also change "for no constructive reason" to "for no policy or guideline-based reason." But following an editor around to enforce a guideline has also been criticized. Following an editor around to enforce a policy, especially one like the BLP policy, has more support. Maybe change "for no constructive reason" to "when it is not necessary for improving the encyclopedia"? Flyer22 Reborn (talk) 03:23, 22 September 2019 (UTC)

You could argue that what is and isn't constructive could be based on community consensus :P, but no I do agree we should do something about this. --NikkeKatski [Elite] (talk) 16:42, 22 September 2019 (UTC)

It would have been nice to ping or notify me, if you were going to refer to me by name. It seems to me that "overriding" is a very strange locution, because the word usually refers to a situation where something is being overridden. What is being overridden here? I'm not very persuaded by the argument that if someone who is hounding someone else thinks that they are doing it for constructive reasons they could use that to wikilawyer. It should be obvious that "constructive" means "considered to be constructive by consensus". To say otherwise is to say that someone could only be violating the harassment policy if they believe themselves to be violating it: "I thought I was doing something good, so how could it be harassment?" In any case, I find the timing of this proposal rather strange, coming a day after the Fram ArbCom case was closed. There, much of the controversy was over whether or not Fram was hounding someone who was repeatedly violating policy by repeatedly telling that person that they were doing so. To incorporate the nonsense that came out of T&S into en-wiki policy is a horrible idea. In any case, ArbCom has just announced that there will be a community RfC about how hounding (and private information) will be defined at en-wiki, and that seems to me to be the proper venue for proposing such changes. --Tryptofish (talk) 21:15, 22 September 2019 (UTC)
I didn't ping you because you commonly edit this page and clearly watch it. I personally don't like being pinged to pages I am obviously watching. But since you'd rather be pinged, I'll keep that in mind in the future.
As for your reasoning for changing the text, I still feel the same way I did when I made the above post. I explained why "overriding" is better and what seems to have been meant by it. It's my opinion that the text should either be changed back to the long-standing "overriding" text or to something else, like "when it is not necessary for improving the encyclopedia." I'm not going to point to a recent example of an editor arguing the "constructive" angle in the way I described, since it will only give that editor an actual reason to show up to this section after I made the above post (although I'd use their showing up in this section as further evidence of hounding). The timing of this post has to do with that. It has nothing to do with WP:FRAMGATE, which I haven't been involved with at all. I noticed that the wording was changed, and apparently not for the better, and I made the above post. This is the proper venue for suggesting that the wording you changed be changed backed to the wording that was there for years. It is the appropriate place to propose any change to this policy. By suggesting that "overriding" be restored, I'm not suggesting we incorporate anything that wasn't there before....for years. By suggesting that the subjective "constructive" be replaced, I'm not suggesting anything that would enable "the nonsense that came out of T&S."
Elitematterman (NikkeKatski [Elite]), since you agree with me, would you rather "overriding" be restored, that we use "when it is not necessary for improving the encyclopedia", or do you have some other suggestion? Flyer22 Reborn (talk) 05:05, 23 September 2019 (UTC) Updated post. Flyer22 Reborn (talk) 05:13, 23 September 2019 (UTC)
Just about the ping thing (I'll reply about the other issues below), I take stuff on and off my watchlist all the time. Given that you were disagreeing with me about something, and referring to me by name, a single ping would have been appropriate. Once I'm aware of a talk page section discussion, there's no need for repeat pings. Thanks. --Tryptofish (talk) 17:41, 23 September 2019 (UTC)
I prefer the word "constructive" because I believe that word is much easier to understand, and to debate if necessary, than "overriding" which I perceive to be vague, legalistic and almost impossible to define or debate in context. Our purpose here is to build and improve an encylopedia. Anything that contributes to that goal is "constructive", a word and concept we frequently use to evaluate whether specific behaviors or long term patterns of behavior either improve or do not improve the encylopedia. If asked to evaluate whether a pattern of behavior is "overriding", I would be bewildered and would have no reasonable metric for separating "overriding" from "not overriding". What do those things even mean? Cullen328 Let's discuss it 05:25, 23 September 2019 (UTC)
Cullen328, I'm not necessarily suggesting that "overriding" be restored (although "overriding" was there in the policy for many years without any problem), but I am suggesting that "constructive" be replaced. You and I both know of what case I'm referring to where "constructive" is already being invoked. Like I stated, anyone can believe that their following another is constructive; it's been argued times before by those who have hounded. And now it's in the policy for people to use, and I don't think that's beneficial. Flyer22 Reborn (talk) 05:37, 23 September 2019 (UTC)
We should not be writing the language of policies and guidelines based on one specific case where an editor may have misunderstood or misused the plain meaning of a word. Instead, we should use language that is easy to understand and commonly used and properly understood on this encylopedia every day. Trying to confuse or complicate the language to foil people who are not acting in good faith is a fool's errand. Clarity comes first and and should always be the top priority. Cullen328 Let's discuss it 05:57, 23 September 2019 (UTC)
We write and change policies and guidelines based on specific cases of misunderstandings all the time. We do this for clarity, and that includes limiting or stopping the misuse of a policy or a guideline. It's why after this recent discussion at Wikipedia talk:Manual of Style/Words to watch, we changed the wording of that guideline. I am never about "trying to confuse or complicate the language to foil people who are not acting in good faith." I am about having clear language so that editors can't wikilawyer in any way. I don't find the subjective "constructive" -- the exact word that has been used by hounding people to excuse their abuse and to actually confuse and complicate situations with debates about what is or isn't constructive following -- to be clear language. Why state "constructive" and not be clear what we mean by it? Assuming that "it should be clear" what we mean has failed countless times with regard to our policies and guidelines. That is why we tweak, tweak, and tweak them. Flyer22 Reborn (talk) 06:12, 23 September 2019 (UTC)
On the contrary, as an administrator, I find the word "constructive" to be easy to understand and a useful tool for evaluating editor behavior. "Overriding" is a word that I simply cannot understand in the context of evaluating editor behavior and any policy language that uses that term in a general sense is a policy that I will be unable to implement because of the baffling and unclear language. If you do not like "constructive" then please propose another word or words that increases clarity rather than radically reducing it. Cullen328 Let's discuss it 06:41, 23 September 2019 (UTC)
In my first post above, I proposed two different wordings. I stated, "We could also change 'for no constructive reason' to 'for no policy or guideline-based reason.' But following an editor around to enforce a guideline has also been criticized. Following an editor around to enforce a policy, especially one like the BLP policy, has more support. Maybe change 'for no constructive reason' to 'when it is not necessary for improving the encyclopedia'?" And when replying to you, I stated, "I'm not necessarily suggesting that 'overriding' be restored (although 'overriding' was there in the policy for many years without any problem), but I am suggesting that 'constructive' be replaced." I'm open to different wording. Per my arguments above, I just don't agree with "constructive" replacing the long-standing "overriding." Flyer22 Reborn (talk) 06:48, 23 September 2019 (UTC)
  • Thanks everyone for the comments, and I've read and thought about all of them. For starters, I decided to do some, um, archaeology, and look into the edit history of the passage in question. Here is the edit, from October 28, 2008, that put the word "overriding" into the policy: [2]. Here is the talk page section from around that time: Wikipedia talk:Harassment/Archive 1#Wikihounding. At that time, editors were changing the term from "stalking" to "hounding". There was never any discussion that I can find about the word "overriding". It's worth noting, I think, that the wording just after it was, and continues to be now, wording that places hounding at the level of "tendentiousness, personal attacks, or other disruptive behavior". I feel that there is a significant difference between that, and following someone who is violating policy, in order to enforce policy.
Regardless of whether the discussion was intentionally started just as the Fram ArbCom case ended, we must inescapably take into account what is happening at en-wiki right at this time. Look, for example, at comments just today at WT:Arbitration Committee/Noticeboard#Wikipedia:Arbitration/Requests/Case/Fram – including [3] and [4]. These issues are about to be discussed at Wikipedia:Arbitration/Requests/Case/Fram#Arbcom RfC regarding on-wiki harassment, where the discussion will be community-wide.
Where JBW says it's obvious that what gets "overridden" are "the negative effects which may be regarded as harassment", that's not obvious to me. Are those "negative effects" the harm to the project, or are they the harm to the happiness of the person making the complaint? Is it possible to make that person unhappy, while also advancing the good of the project? Yes, and that was very much at the heart of the Fram case. Someone who was repeatedly violating policy and guidelines, and was not hearing complaints about it, considered Fram's repeated pointing out of those violations to be hounding. Do editors have a "right" to violate policies and guidelines without having to respond to editors who have concerns, or to administrators who keep trying to enforce our norms? No. Are those administrators acting "constructively"? Yes. (At ArbCom, they decided to almost no one's satisfaction that it was "borderline hounding" because of the way in which Fram did it, as opposed to the fact that he did it. And the issue is sufficiently murky that ArbCom have decided to have that RfC.) In any case, if some of us find "overriding" unclear, and some of us do, it's worth rethinking it.
But JBW's comments about a small constructive element were very helpful to me in understanding the potential shortcomings of "constructive". So that got me thinking. Just being "constructive" might not, in itself, be sufficient. It has to be, sort of, very constructive. And "overriding" is clearest when we know exactly what has to be overridden.
And that gave me this idea: Each of those two words actually corrects the other word's limitations. I'd like to suggest that we change it to: "overridingly constructive". --Tryptofish (talk) 19:02, 23 September 2019 (UTC)
A very good suggestion, Tryptofish, and better, I think, than anything else that has been suggested. I support that. JBW (talk) Formerly known as JamesBWatson 20:03, 23 September 2019 (UTC)
I'm glad. Thanks! --Tryptofish (talk) 21:34, 23 September 2019 (UTC)
I've read this entire discussion, and have come to agree that "overridingly constructive" is the best option here. -Crossroads- (talk) 22:12, 23 September 2019 (UTC)
JBW (especially JBW) and Crossroads, thanks for giving your thoughts. Tryptofish, thanks for thinking this over. Where JBW says "Suppose I do something which has large and extensive aspects of 'disruption to another user's own enjoyment of editing, or disruption to the project generally', but which also has a very small element of doing something constructive." is what I referring to when I said "But following an editor around to enforce a guideline has also been criticized. Following an editor around to enforce a policy, especially one like the BLP policy, has more support." I mean, an editor could be following another to make sure that the other is adhering to MOS:BADHEAD. But is it really necessary for that editor to do so, especially if the one doing the following has a bad history with the other editor? In that case, bringing MOS:BADHEAD to the offending editor's attention once on the editor's talk page and bringing in others to assess the matter if it continues is enough. With regard to "Is it possible to make that person unhappy, while also advancing the good of the project?", my point has been that the "at the expense of the editor's happiness" aspect should be necessary, such as stopping BLP violations. But even in those cases, the community feels better about uninvolved admins continuing to address a specific editor, such as with warnings, rather than an editor with a bad history with the other taking on the task. I'll go with "overridingly constructive" as a compromise. Flyer22 Reborn (talk) 00:32, 24 September 2019 (UTC)
I have to hand it to trypto, thats probably the best thing we can do here. --NikkeKatski [Elite] (talk) 13:26, 24 September 2019 (UTC)
Thanks everyone; I'm genuinely flattered. About the points that Flyer22 makes, I agree that these are things that the community needs to come to a consensus about, whatever one's personal opinion. If I understand correctly, the planned RfC is going to be the start-off point for such a community discussion. --Tryptofish (talk) 21:12, 24 September 2019 (UTC)
I want to reiterate that, for me, this discussion had nothing to do with Fram/FRAMGATE. I haven't been keeping up with that case. With it mentioned in this discussion, however, I wondered if "overriding" was replaced with "constructive" because of it. But I see that it was changed days before FRAMGATE. Still, as seen by the provided links above, the "constructive" wording has affected the Fram case. So I understand why you brought it up. Flyer22 Reborn (talk) 23:40, 24 September 2019 (UTC)
Are you going to implement the agreed-upon wording, or would you rather someone else implement it? Or are you waiting for the RfC you mentioned? I don't see that we need to wait for that. Flyer22 Reborn (talk) 23:53, 24 September 2019 (UTC)
Just so you know exactly how I came to make that edit, if you go a few edits back in the edit history, someone else made some changes to that section following a discussion that I had not (to my memory) been involved in, and I made a series of copyedits to clean up after that, and I made that edit along with those. At the time, I truly thought that the change was just a minor change that didn't alter any meaning. Go figure. Had I realized that there would be any issues about it, I would not have made that edit without prior discussion, and I would not have made it after the Fram thing had begun.
I didn't make the edit yesterday, because I figured I would give it one more day in case of any late objections. But I'm happy that JBW took care of it. --Tryptofish (talk) 22:10, 25 September 2019 (UTC)
Yeah, Tryptofish, I'd examined the series of edits leading up to the change from "overriding." And, again, you made that change before FRAMGATE happened. I know that the disputes regarding Fram happened before FRAMGATE, meaning they are what led up to FRAMGATE, but your "overriding" change wasn't in response to FRAMGATE. Flyer22 Reborn (talk) 22:44, 25 September 2019 (UTC)

I am the editor who is being discussed here who has been accused of hounding. [5]. I've read this discussion, and I feel like what's being missed is the word "reason". Either the reason to follow an editor is constructive or it's not; it's not acceptable to follow an editor with the intention of causing them distress while also making constructive edits which override their bad faith reasons. What I also find problematic about the language is the focus on appearances and the perceptions of the effected editor: "an apparent aim of creating irritation.". I believe hounding is when the aim is to cause distress, and regardless of whether that behavior is accompanied by constructive edits. Administrators can decide whether apparently hounding behavior is hounding behavior. So then the question is, if the "reason" is purely constructive, but the followed editor feels distressed, is that simply not hounding, or must the constructiveness outweigh the distress felt by the other editor? I would say that's an impossible thing to measure, and that harassment is actually defined by intentions.
In summary, I would change the language to "for anything other than constructive reasons" and remove the word "apparent". Kolya Butternut (talk) 03:24, 25 September 2019 (UTC)

Administrators, including JBW and Johnuniq, were clear on your talk page. And you unnecessarily showing up here for more confrontation, just like you unnecessarily watch my talk page to keep tabs on me (and drop in at talk pages I'm at) and try to justify that (watching my talk page) with some claim that I'm focused on/after you, is further proof that you have an issue when it comes to understanding WP:HOUNDING. Stating "I'm not going to point to a recent example of an editor arguing the 'constructive' angle in the way I described, since it will only give that editor an actual reason to show up to this section after I made the above post (although I'd use their showing up in this section as further evidence of hounding)." and "You and I both know of what case I'm referring to where 'constructive' is already being invoked." is not truly discussing you. In fact, it is a clear effort to have you stay away, as I do not want to interact with you at all. But, alas, you couldn't resist. The only reason I even noted that I was not "going to point to a recent example" (you) is because Tryptofish stated, "I'm not very persuaded by the argument that if someone who is hounding someone else thinks that they are doing it for constructive reasons they could use that to wikilawyer." Otherwise, I wouldn't have alluded to you at all. But even if I hadn't, we both know that you would have showed up here anyway. I'm not stating anything else in this section to you, except that I oppose any proposed wording you suggest. Compromise wording has already been worked out, and for valid reasons (which JBW couldn't have been clearer about). Flyer22 Reborn (talk) 05:49, 25 September 2019 (UTC) Updated post. Flyer22 Reborn (talk) 05:58, 25 September 2019 (UTC)
I'm having an issue seeing where someone contributing in a truly constructive manner can cause warranted distress. It may very well be possible but it seems like an odd situation to me. Would anyone be so kind to point me towards an example case so that I may understand better perhaps? Or am I missing the point entirely? --NikkeKatski [Elite] (talk) 21:15, 25 September 2019 (UTC)
I hope that I don't sound like a broken record, but those are exactly the kinds of questions that came up in the Fram case and they are likely to be discussed more systemically than we can do here in the forthcoming RfC. I don't want to get drawn into the dispute just above, and so for that reason I don't want to parse what I see as the differences between the Fram example and this one. But they do need to be parsed eventually. For now, I hope that we can close the discussion here. --Tryptofish (talk) 22:23, 25 September 2019 (UTC)
Elitematterman (who signs as "NikkeKatski [Elite]"): It happens all the time. I can assure you that in my administrative work I have come across that situation thousands of times. Here is one example. An editor with strong antisemitic views goes round posting Jew-hating statements in many articles. Other editors follow their editing history and revert their edits. Editors post warnings to the editor, which escalate into very strong warnings. Eventually the editor is blocked from editing. The editor who did it in the first place, who sincerely believes that Jews are evil and that he or she is doing a good job by publicising their wickedness, is genuinely distressed to find his or her good faith work destroyed, and to be subjected to what he or she perceives as personal attacks, but the actions which cause that distress are fully warranted, and the fact that they cause distress is an unfortunate downside that we have to accept. That is a fairly extreme example, rather than a typical one, but I have chosen it because that very fact makes the point clearer. However, every day there are situations where one person's actions are totally unacceptable, so that steps have to be taken against them, even if those steps run the risk of causing upset for that person. JBW (talk) Formerly known as JamesBWatson 10:27, 26 September 2019 (UTC)
For what it's worth, I support the compromise language that Tryptofish came up with. Thanks. Cullen328 Let's discuss it 22:55, 25 September 2019 (UTC)
Thank you for the insight. Its quite clear to me now that this can be a very complex matter. --NikkeKatski [Elite] (talk) 15:19, 26 September 2019 (UTC)

question about outing

If someone has posted on their Twitter account their Wikipedia information, and also post hate speech there against someone whose article they then try to get deleted, can that information be used against them in an ANI? Is every single thing said off of Wikipedia simply blocked out no matter what? Dream Focus 17:53, 5 November 2019 (UTC)

They would need to post their twitter account on Wikipedia, otherwise linking a twitter account and a Wikipedia account on Wikipedia is outing. That being said you can email the information to ARBCOM if you think it needs to be dealt with. --Kyohyi (talk) 18:00, 5 November 2019 (UTC)
Seconding what Kyohyi said. The connection needs to be made on Wikipedia* by the owner of the Wikipedia account - in part this is to protect against Joe jobs, but you believe the account is controlled by a Wikipedian then you can email the information to arbitration committee. You can, and should, also report the account to Twitter as hate speech is against their terms of service. *It's less clear whether it is outing if the connection is made on another WMF project - if it's linked to prominently from en.wp (e.g. their userpage here says "see my userpage at Commons" and the link is made there) then it's almost always going to be fine though; where it is less clear cut (e.g. the connection is on a subpage on the Esperanto Wiktionary and they don't make a big thing here about contributing to that project) it's better to err on the side of caution and send the info to arbcom privately. They can post the information publicly if they determine the connection is public and they feel there is benefit to doing so. Thryduulf (talk) 19:23, 5 November 2019 (UTC)
The above replies are correct. I would just add that if the editor has, on Wikipedia, given any clue that they control or otherwise approve of the Twitter account, that should be included in an email to Arbcom (see User:Arbitration Committee). Make the email brief with the point in the first sentence: should editor X be sanctioned due to Y? Unfortunately Arbcom is snowed under and an email might not be answered for a long time, but it is best. Johnuniq (talk) 22:21, 5 November 2019 (UTC)

Does anyone recall a detailed discussion clarifying the boundary between hounding and disruptive editing?

Just putting this out there as I wonder if more clarification is needed (and whether we have discussed it before...?) - We have under Wikipedia:Harassment#Hounding "Correct use of an editor's history includes (but is not limited to) fixing unambiguous errors or violations of Wikipedia policy, or correcting related problems on multiple articles." - hence clarifies when we can review edits for copyvios etc. and then at the other end of the spectrum we have stalking/hounding of editors (i.e. harassment). How about the grey area in the middle? Has this been looked at before and discussed?

e.g. possible scenario - left-leaning editor A edits an political article and the source and text are (let's say) a little bit open where the text in the wikipedia article is a bit more "left" (positive/whatever) than the source. Not grossly wrong but possibly a little. Now, right-leaning editor B comes along and they have a discussion and afterwards the article is reworded so that the text better matches the source. Now both editors edit political articles. Is editor B justified in looking at other edits of editor A on political articles to check and possibly reword text? Cas Liber (talk · contribs) 00:50, 7 November 2019 (UTC)

I don't recall anything exactly like that, but there have been two recent discussions about that boundary: #Addition to WP:HOUND and #"Overriding reason" vs. "constructive reason" in WP:HOUND, both above. I think there has been sentiment that the exact boundary is something best left to administrator judgement, rather than being something that can really be spelled out, and that the boundary is based on whether the following of an editor is about advancing polices and guidelines, or about causing unpleasantness for the person being followed, and that it needs to be clearly the former. I think if editor B in your example is just making corrections based on what editors A and B agreed upon, that would be fine. But if editor B is focusing on editor A in a single-minded and nitpicking way, that could potentially be hounding. --Tryptofish (talk) 01:07, 7 November 2019 (UTC)
Also, I'm under the impression that ArbCom is working on having a community RfC that will, in part, touch on that issue. --Tryptofish (talk) 01:08, 7 November 2019 (UTC)
My eyes went square looking at the walls of text. I think some sort of RfC to get more of a quantitative feel would be good. Yeah....about that community RfC...... Cas Liber (talk · contribs) 12:19, 7 November 2019 (UTC)
  • The question of "could I reasonably get ANI to reach a consensus that the user's edits are a problem, taken as a whole" is a good start. Another possible way of looking at it - suppose, after user B makes a bunch of changes to user A's edits, user A takes objection to the changes and goes back to revert all of them. If this then ended up at WP:ANI, would there be a general consensus that A was in the wrong and B was in the right? Or would they call it a bunch of content disputes? If it's the former, B is probably in the clear; if it's the latter then what B did has to be stopped, because otherwise we'd see content disputes spill out over the entire wiki in ways that make WP:BATTLEGROUND problems even worse.
  • I think it is almost always acceptable to check closely related articles for the same extremely specific problem or dispute, as a one-time thing - eg. if there's a content dispute about one specific quote, or the coverage of one specific event, or some other very narrow thing that you've realized one editor just added to a bunch of articles, I think it's always reasonable to check their edit history and go over what's essentially the same edit made on multiple pages, since that still falls under the general locus of one dispute and discourages things like WP:COATRACK or people trying to push something through by editing a bunch of articles all at once (generally I feel mass edits without a clear consensus behind them are questionable due to the difficulty of dealing with them - of course that also applies to mass-reverts, but eg. putting together a list of every place the discussions touch on and placing notices to centralize discussions is reasonable.) Particularly when an editor makes a bunch of very similar edits on related articles, it's reasonable to object on all of them or to try and centralize discussions. But that requires that everything fall reasonably under one core reasonably-narrow dispute that B could articulate - it doesn't extend to eg. obsessively double-checking everything A does in an entire topic area. As a general guideline, you shouldn't even need to check their edit history most of the time for this because the related articles will be obvious anyway.
  • Perhaps related to the above, the length of time you do something for matters as well. Which isn't to say it's the only thing that matters - checking every single edit an editor made just because you're in a dispute with them is not OK, even if you only do it once - but the most serious types of WP:HOUNDING continue for extended periods of time. And, conversely, it's hard to justify continuously following a user's edits like that - if there's a constant problem that requires correction, it should be brought up on an appropriate editor-conduct board, and if there isn't then the "fix one specific problem" and "catch all instances of one specific narrow dispute" justifications rapidly disappear when you're following someone's edits for an extended period. So "repeatedly seems to follow the user over an extended period of time" would be a major alarm.
Also of course sometimes you'll bump into another editor all over the place because you edit the same topic area. --Aquillion (talk) 04:51, 8 November 2019 (UTC)
I'm generally of the mind that this is the kind of thing that cannot be quantified. And a caution that comes to mind about the ANI test is that ANI is sometimes a lousy measure because warring camps often show up to yell at each other. But, that said, I do like the idea of trying to list some of the things that can distinguish acceptable from unacceptable behavior. If nothing else, I think it's useful to find out whether there are things editors tend to agree on, or things that turn out to lack consensus. Just throwing out some thoughts that I haven't really worked on as they pop into my mind:
Things that tend to be acceptable:
  1. The editor being followed is violating policies and guidelines, and the editor doing the following is simply seeking to uphold those norms.
  2. The editor doing the following has previously been a frequent contributor in the topic area where there is overlap.
  3. The following is done in a polite and friendly way.
  4. The interactions take place only briefly, and are not a chronic pattern.
  5. The editor doing the following is also directing attention to a lot of related edits by editors other than the editor being followed, rather than focusing narrowly.
Things that tend to be associated with hounding:
  1. There is no obvious and overriding policy reason to be monitoring the other editor's work.
  2. The editor doing the following suddenly takes an interest in an area they never edited in before, in order to follow the other editor.
  3. There is an unfriendly or menacing tone to the way the following is done.
  4. The following of the other editor is a chronic pattern over time.
  5. The editor doing the following is conspicuously focused on the editor they are following.
None of these criteria, taken alone, proves much of anything. Rather, it's the preponderance of the pattern. To be acceptable, there should be a clear preponderance of the acceptable criteria. --Tryptofish (talk) 19:48, 8 November 2019 (UTC)
Automated detection of Wikipedia misconduct!, 15-minute video presentation of the research paper!
@Tryptofish and Aquillion: If you converted your ideas of what hounding is to a human readable policy, then humans could apply that in a few cases to stage the field for bots to supplement human review in the future. If we had labeled cases which were specifically hounding then that would help make the automation more precise. Blue Rasberry (talk) 21:32, 8 November 2019 (UTC)
What happens if a bot starts hounding me? --Tryptofish (talk) 21:59, 8 November 2019 (UTC)
Got you covered - we are already making bots to seek out and punish earlier designed bots gone rogue. Blue Rasberry (talk) 22:14, 8 November 2019 (UTC)
Hehehe. Will have a read. Cas Liber (talk · contribs) 22:23, 8 November 2019 (UTC)

Is suggesting harassment (outing) an indef ban offence or less?

In relation to the incident discussed at Wikipedia:Administrators'_noticeboard/Incidents#Unblock_request_now_at_two+_weeks,_editor_apologized,_seems_simple_enough?, where an editor suggested that another editor's identity should be revealed, how seriously do we wish to reprimand them? In this particular case, the editor was indef banned for making such a suggestion (to be clear, no outing has occurred, but a suggestion was made, effectively saying that "a newspaper that mentioned editor X should have revealed his identity... revealing it would be informative".). Should such a suggestion, made on wiki, merit 1) indef ban 2) shorter duration ban 3) warning 4) no action 5) something else? There is no consensus in the thread about this (so far only two admins have commented on this issue). I think it might be helpful to discuss this here and formulate a policy for future incidents. Personally I think that anyone who suggests some form or harassment should be sternly warned, and repeated offenders might merit an interaction ban, but I wouldn't go as far as a full editing ban for merely making a tactless suggestion. Thoughts? --Piotr Konieczny aka Prokonsul Piotrus| reply here 02:51, 10 November 2019 (UTC)

  • You keep saying it depends. On what? On who editor X's friends are? On what has been rumoured to have happened off-wiki in BADPLACES you can't link to, only summarize (as one does)? What does it depend on JohnuniQ? 🌿 SashiRolls t · c 03:46, 10 November 2019 (UTC)
  • whose context? yours? mine? the deleted one? the banner-bot's blocker-bot's?
& the suprasegmental data, where is it stored? As I recall, the case Piotrus speaks about is complicated. Bots (simply the best at go & chess) like rules. 🌿 SashiRolls t · c 14:35, 10 November 2019 (UTC)
More