Talk:Harassment consultation 2015/Ideas/Hire a harassment expert

This is a no-brainer. We have the money, we have the problem. We lack the expert, so hire one. This does not stop us from trying to come up with alternative routes to a solution, and a hired consultant remains hired subject to acceptable performance. Targets must be compiled for deliverables, invitations put out for tender etc. I see this as the tricky part. · · · Peter (Southwood) ^(talk): 06:56, 17 November 2015 (UTC)

Completely agree. Only this week, we have ARBCOM members being educated by the community as to why it's not ok to punish someone for the actions of others. We expect them to know this. Burninthruthesky (talk) 09:19, 17 November 2015 (UTC)

This is a great strategy: having someone responsible for a) regularly training eidtors in trusted rights positions, on what harrassment is and b) provide oversight/consultation with ArbCom and other community groups through deep knowledge of online harrasment, and managing it. Sadads (talk) 16:50, 17 November 2015 (UTC)

Also of importance, we can reasonably expect someone who is not involved, and has no history with any of the players. · · · Peter (Southwood) ^(talk): 17:14, 17 November 2015 (UTC)

Just to make it clear, this idea is lacking general impact, as it is rather English-language-oriented, or - in case WMF can afford hiring more experts for more languages - major-language-oriented. I cannot imagine, how the expert can follow, evaluate and help solving cases of harassment on non-English-language wikis, or simply on wikis he does not speak their languages... --Okino (talk) 23:38, 17 November 2015 (UTC)

The English language Wikipedia is where the biggest problems appear to exist, and it is the biggest project. It would be a reasonable disposition of funding to tackle the biggest problems first. If it turns out to be worth the money, it could be expanded. The principles should be universal, and translators should be available for urgent cases in other languages. A potential solution does not have to be universally applicable to be worth trying or using. It would be irresponsible to decline a potentially viable solution to a large and persistent problem on a major part of the projects because it might not be applicable to every part, some of which may not even have the same problems. · · · Peter (Southwood) ^(talk): 08:52, 18 November 2015 (UTC)

I did not say that I decline this solution or that the solution should be declined, I just noted an argument that this one aspect should be taken in mind.

And - I do not follow the opinion that "English language Wikipedia is where the biggest problems appear to exist". In fact, we had some cases with signs of harassment or stalking on Czech language projects, the users even tried to bring them in front of wider public to Meta, but they struggled getting any significant response because of the lack of interest from the non-Czech-language community. It did not mean though they were not cases of harassment and it did not mean they were not big problems, it only meant the problems failed to break through the language and community barriers. So I feel the opinion is based on a number of people that can take care of en.WP and its problems and who really do care, and not on the true scale of the problems themselves. --Okino (talk) 18:54, 18 November 2015 (UTC)

Fair enough, I accept that there may be other projects with equal or bigger problems. As no-one has made a comparative study, there is no data, and I am unable to read Czech language so cannot form an opinion from personal experience. If reasonably convincing evidence can show that another project would be a better place to try this out I would not object. Without that evidence, I think that English Wikipedia has enough of a problem to justify the experiment. · · · Peter (Southwood) ^(talk): 08:12, 19 November 2015 (UTC)

• I disagree that we should have someone "to act as an on-call consultant to arbitration committees, administrators and the Foundation." What we need is help establishing training, policies and protocols. If we have someone dealing with these things day-to-day then all the harassment across wikimedia becomes their problem, which is not what we need. Stuartyeates (talk) 20:38, 18 November 2015 (UTC)

• Stuart, I hope education would be part of what a consultant could offer. I agree completely that we need training for editors and admins. The single biggest problem with harassment on the English Wikipedia is even getting people to recognize that it's happening, and what the effect of it is on the target. SarahSV ^talk 22:55, 18 November 2015 (UTC)

• Actually, I'm going to modify my statement above. As part of good governance, the governance level of WMF may need "an on-call consultant" to help them 'get it right' when things escalate to that level. Stuartyeates (talk) 18:37, 22 November 2015 (UTC)

Consider someone who has proven can withstand and overcome harassment (has been long successful despite the attacks), an expert.--ManosHacker (talk) 23:56, 25 November 2015 (UTC)

A harassment expert should have as a main duty to train users - his ambassadors - and probably not to resolve conflicts one-by-one. --FocalPoint (talk) 21:06, 28 November 2015 (UTC)

Yes, Yes, YES. And combine with Rosiestep's suggestion that all admins be required to receive training to keep their admin status. Lightbreather (talk) 17:26, 31 January 2016 (UTC)

Some questions about specifics

Hi SlimVirgin! This is a really interesting idea. I'm trying to think through how an expert would actually function in this role, and I have a couple of questions about how you (or anyone else reading this) envision it working, etc.

• Where in the organizational hierarchy would a harassment expert sit? That is, would they be a member of the WMF's Community Advocacy team? Would they be a WMF contractor whose entire job is "look for harassment issues and jump in"? Would they be essentially a community member, an volunteer who is given some sort of advanced user right and then talks a lot? Would they somehow be slotted into bodies like local arbcoms as non-voting member?
• What particular expertise would you be looking for? Internet harassment straddles more than one academic/professional specialty - sociology/social psychology, law enforcement, social media trust & safety, and so on - and I wonder if one of those is more of what you're imagining than the others
• What influence would the expert have? Would they be simply an advisor, with no policy-making power (if so, advisor to whom, and how far down the line? english wikipedia functionaries? all functionaries? all community members?)? Would they have input into or veto power over community processes like RFCs and Arbcom cases?
  • If they had some amount of voting/decision-making power, how much? A voting arbcom membership? Veto power over arbcom decisions, like Jimmy Wales has?
• How could we help ensure that communities actually accepted the expert's input, and didn't just ignore it (or even disempower/block/etc the expert) as "not how we do things" or "stupid friendly-space SJW stuff"? We can talk until we're red in the face about what is and isn't best from an expert's perspective, but our communities still run on the power (and vote) of the crowd, not on top-down expertise

I can probably think of a million more of these as I try to play devil's advocate inside my head, and I don't expect you to have answers for all, or even most of these, but I think it would benefit the proposal (and the WMF's ability to understand and act on it) if we could get some discussion going about how people envision this stuff playing out. Kbrown (WMF) (talk) 20:46, 30 November 2015 (UTC)

And replying to myself to point out that since I wrote this, someone has suggested training community members to train others using Ally Skill workshops like the one that happened at this year's Wikimania. I wonder if that could be rolled into this expertise idea - would a community-sourced expert be more likely to be trusted by their community? Could community leaders (admins, arbcoms, functionaries?) be specifically asked to attend this sort of training (if it's available online? somehow?) and become their communities' local experts? Kbrown (WMF) (talk) 21:10, 30 November 2015 (UTC)

I'll jump in, speaking only for myself of course:

I'll suggest an independent position, reporting only to the WMF Board. Certainly working closely with Community Advocacy, but being independent.

How would this person be brought into play? Probably the best is via e-mail, or phone calls from the person claiming to be harassed. The role probably should be just as a recorder at this point - note that a complaint was made, instructions or suggestions were given on how best to proceed, which might include on when to report back, etc. So the role would not be an advocate, only as an outside trained observer and investigator. Helping to train the community would be possible, and making it clear to the community that they were available could definitely be part of the job.

Working with ArbCom might be the biggest challenge, but could have the biggest effect. I'll suggest that at the start of any ArbCom case (and during if necessary) the independent observer just state the facts as they've observed them, present their expert opinion on whether harassment has occurred, answer any questions from the arbs. This might be either on the arbcom pages or via private e-mail. At the end of the case they could over-rule any decision they believe is inconsistent with WMF policy on harassment.

The alternative to having an impartial observer would be to have a victim's advocate, but that would preclude working with ArbCom.

The question of English-language only or multiple languages? Let's start with English and try to work up to other languages as needed. Maybe the WMF could hire somebody who speaks English and also Spanish-Portuguese and Chinese or German? Probably not, but perhaps other WMF employees who speak other languages (e.g. a Czech programer) could be trained to record the complaints in that language, and give the basic advice on how to proceed and report back, while relying on the expertise of the independent observer.Smallbones (talk) 18:33, 1 December 2015 (UTC)

@Kbrown (WMF), SarahSV, and Smallbones: I discussed this in some detail at Grants:IdeaLab/Comprehensive harassment and privacy policy. The ideas were heavily borrowed from (and reworked to fit online and Wikipedia situations) from a NOLO book on preventing employee lawsuits, the title of which escapes me at the moment. Preventing sexual harassment has been standard in HR departments for at least 30 years, and there should be no need to reinvent the wheel. See for instance here, and I'm sure government agencies would be an even better source of examples.

For specific answers to KBrown's questions:

• Where in the organizational hierarchy would a harassment expert sit? A high as possible, and an independent office, probably reporting to the ED. There also needs to be more than one reporting path. One reason for this is in case there is a situation involving a high-ranking individual, for instance an arbitrator or WMF employee--you need someone with enough authority to deal with high-ranking people. Who do you call when the problem is the person you are supposed to report the problem to?

To answer Smallbones about the board, you need someone who is hands-on to solve the technical and cultural problems; it's a nuts-and-bolts how-to-make-the-policy-work thing, and should belong to someone who is answerable to someone, and therefore motivated to solve the problems before they become public. The board is more for political direction than technicalities; this is not a political thing, it's about making things work.

This need to be a paid position. It needs to be a multiple approach, both with training and responding to problems, and it needs to be confidential, in case that isn't already very and painfully clear to everyone.

Forget ArbCom, this has already proved unworkable. Also, the problem of harassment is one of enforcing the Foundation's Terms Of Service, which ArbCom may find outside their remit, as they are often very reluctant to step outside the specific enwiki policies.

• What particular expertise would you be looking for? this is a straight-up administrative / HR problem and typically, in an agency, the new-hire training and any problems with individuals are dealt with by HR, as this is seen as a performance evaluation problem in the case of paid employees.
• What influence would the expert have? They would enforce the WMF policy. Are WMF employees also members of the community? Sometimes, but that's always tricky. You do need someone who understands the WP processes.
• How could we help ensure that communities actually accepted the expert's input...our communities still run on the power (and vote) of the crowd, not on top-down expertise The harassment officer / ombud would have the power to immediately remove harassing statements, perhaps in response to a private flag, which would have an immediate positive preventative effect, also perhaps the authority to remove someone from the community and establish prerequisites for their return, for example, require further training before being returned to the community. It would not have to be a block button, I believe there are codes that can prevent editing, as I saw an admin do this once when they retired. In any case, much of this would be a technical question for the programmers and developers. I have heard people saying privately they are surprised the foundation hasn't been sued yet over this issue--enforcing the TOS should be a matter of law and not open to mob rule, so IANAL, but is WMF Legal not weighing in on this? But again the technical solutions will be the best ones, and the least controversial.
• Could community leaders (admins, arbcoms, functionaries?) be specifically asked to attend this sort of training (if it's available online? somehow?) To me that's the most obvious first step, to do it for fun, and to start a dialogue, so that people voluntarily arrive on the same page, but I know the arbcom wouldn't agree with me. You see quizzes everywhere online, people take them for entertainment, and I'm sure the critics would love a chance to mock it before it went live some months down the road. A prospective admin could take it as a preliminary step towards an RfA, or you could have a barnstar or user box for it. This type of online (or rather intra-net internal system) test has also been done in agencies as part of mandatory HR training, for instance a two-hour refresher on conflict of interest policy where you answer questions before you get a printable diploma in case you need to prove to your director you completed the requirement. —Neotarf (talk) 01:39, 2 December 2015 (UTC)

@Smallbones again, I know I have left Arbcom out of this equation, but I do think there is an idea here that needs to be honed. If you look at this Charlie Rose interview with the dean of Harvard Business School, you will see how they took the problem apart and how they solved it. [1] I have noticed over and over at arbcom the women are ignored completely, even the admins and former arbs. It's as if they didn't exist. In one recent case, an arb even went to far as to hat the ideas of the women users, while leaving obvious trolling vandals untouched. No wonder there is such low morale, when arbcom isn't willing to even listen to the women. But Harvard Business School has an explanation and a solution; I was going to try to cut out that audio segment and post it separately, but I would recommend you listen to it, it's worth digging out if you are really interested in the topic. It would take very little to apply this to arbcom proceedings. —Neotarf (talk) 01:55, 2 December 2015 (UTC)

Thank you Smallbones and Neotarf for your thoughts so far! I will reply more extensively later, but for now just leaving a placeholder to say that I'm reading, I appreciate the input, and I'm going to watch the interview Neotarf suggested and see where that takes me along these lines, too. Kbrown (WMF) (talk) 15:26, 2 December 2015 (UTC)

┌─────────────────────────────────┘
Hi Karen, I'd like to see an expert who reports to the executive director. I'm thinking of someone with a proven track record; the two most obvious backgrounds are law and sociology, but I can think of non-academic backgrounds too.

A huge issue on Wikipedia is that our dispute-resolution processes sometimes (perhaps often) end up constituting harassment, and may even put editors in danger. For example, when you're being stalked online, experts will tell you that it's important to disengage. Don't respond (except perhaps for one message saying you want to be left alone); don't acknowledge. But if that person takes you through dispute-resolution on Wikipedia, the arbitration committee will penalize you for not responding. You're expected to submit evidence, respond to evidence, everything is posted in public, and it might continue for months.

We need an expert who can point out how flawed this is, and who can bring our thinking about harassment and dispute resolution up to date. There are multiple examples of editors being placed in awkward positions because of things done by people in positions of influence on Wikipedia, and it happens because those people are not familiar with how other organizations handle harassment. SarahSV ^talk 02:35, 3 December 2015 (UTC)

@Kbrown, most of this would be in the field of public administration. If you don't have a position for a staff person budgeted at the moment, you might consider looking for a grad student from a local public administration program who needs an internship project for spring semester. You might look here. Not sure about all the accreditation there. Some typical programs in other areas are midwest and east coast. A person might be hired into an HR position with this type of credentials, or if someone is promoted from within they might enroll for this type of program. As far as setting up a new program, that's not the same as stepping into an established program and running it. You might check with some of your friends in government service, who can talk to their department director to find out who set up the harassment programs. If you just make a cold call, you might end up talking to anyone--there are all kinds of people in public service, only some of them are the ones who can make something work; so if you can get an informal referral from a friend and work your way up from there, I daresay the Wikipedia name might open some doors to information.

@SarahSV, thank you so much for your comments about the arbitration process. It's so hard for me to think about this objectively, even now. One of the huge problems of arbcom, since they are elected by a majority, is that they see themselves as representing the "consensus" or majority position, not protecting emerging minority groups, or preventing discrimination, or protecting the encyclopedia itself. No one represents those interests. Is this another good place for a grad student, as a stop-gap measure? Do law schools do this?

Another problem that impacts women disproportionately is that of privacy. There is a privacy policy in place, quite a comprehensive one, but no means of enforcement. —Neotarf (talk) 01:52, 8 December 2015 (UTC)

Neotarf, I wouldn't want to see grad students in these positions. We need professionals from outside the movement, people able and willing to stand up against the pushback they're going to experience. Re: having an outside observer on the committee, this is a good idea. Several people have been suggesting it for a while. SarahSV ^talk 02:46, 8 December 2015 (UTC)

Yes, Sarah, of course they need experienced professionals for the position(s), I'm talking about the planning phase--what we are trying to do for free in our spare time, without even being able to see or talk to the people. They would at least benefit by plugging into the local universities and government institutions for whatever knowledge is to be had in their area. This is not rocket science, there is no need to invent the wheel, but it does need to be specific to this particular unique situation. I have scanned something for Karen, along these general lines, but I have misplaced it for the moment, will have to upload later.

As far as the committee, you may remember I was an outside observer of sorts when I covered it for the Signpost in 2013. It was a huge amount of work, 20-30 hours a week at least--I read all the pages and more than half the diffs, and had all the talk pages watchlisted, and a lot of times I still couldn't figure out what was going on, at least not right away. It is very easy to tell what is going on when you are part of an editing group, as I was with MOS, and can watch the edits unfold in real time, or as happened with the Tea Party case, which was suspended for a time, while the dispute continued, but you would be amazed how very, very hard it is to tell from the outside, by backtracking old edits. You know how people can play with the archiving, or make edits disappear from the visible text, or how a dispute can unfold over several talk pages, and how you have to use the history to properly track a talk page, or how to know if someone tweaking a bot's settings is malicious or helpful. How do you teach someone about that? But I suppose that's not what an observer would do. I was looking at the dispute itself, they would be looking at process.

Earlier I said "position(s)". Just thinking out loud here that some people are not be able to talk to a man about some situations, others might not be able to talk to a woman, maybe you would need one of each eventually, or maybe cross-train whoever is currently doing the harassment stuff for the regular employees as a back-up person. —Neotarf (talk) 03:50, 8 December 2015 (UTC)

About a professional in the SF area, also professional definitions of harassment, I have a specific possibility, if someone in an official position is interested they can contact me. --Neotarf (talk) 23:42, 15 December 2015 (UTC)

Sarah, I've added a new page for your idea about the dispute resolution professional. IMO this is a completely separate topic and deserves separate consideration. —Neotarf (talk) 18:53, 19 December 2015 (UTC)

Imposition vs dialog

The way that this proposal has been fleshed out by some contributors on this talk page, has resulted in an expanded proposal that is not the normal way Wikimedia projects do things, and is not in keeping with the original proposal. Admittedly, the normal way Wikimedia projects do things seem to be failing us here in some cases. But I think we've had great success with (semi-)-consensus-based written policies and fair, public, accountable enforcement of those policies in the past, and we should see how far we can go with that kind of approach in future.

I am suggesting a compromise - WMF hires an expert, but does not give the expert a dictatorial level of power above ArbCom, even on matters which allegedly "solely" concern harassment. That would be a PR disaster (the likes of Gamergate would probably regard this structure as fatally compromising the ethics of wikimedia, as soon as it was used in a way they didn't like) and probably an actual disaster as well. Rather than having the power to impose "nuanced" (i.e. arbitrary, not based on policy) decisions on Arbcom, admins and editors at the drop of a hat, or even the power to impose policy ahead of time (which wouldn't fly, I don't think), I would suggest that their role would be to advise the community as a whole, and Arbcom in relation to their non-public activities, on how site policies should be adjusted to better deal with harassment, and how the WMF could support the community and Arbcom in implementing their choices. Again, this would be far more in keeping with the traditional WMF approach of deferring to the wishes of the community - an approach which is adopted, I believe, in order to not deter existing editors.

Having a harassment czar with super-admin powers (have we forgotten the "super-protect" fiasco so soon?) and an overriding goal to shut down harassment, risks attracting all kinds of unscrupulous users who might claim (even falsely!) that because of their race, religion, gender etc. that they are being harassed, and that therefore cases against them should be dismissed and their opponents in an edit war alleged harassers should themselves be sanctioned or blocked. Such users might deliberately mention their race or gender somewhere on a wikimedia site, then deliberately post flamebait content or discussion messages in order to elicit "harassment" in response, and then claim that they are being oppressed... obviously on grounds of their race or gender... and use this to try to short-circuit people's rational thought processes and obtain special treatment. It might seem unlikely now, and I hope it would be unlikely to succeed in any case, but it is a risk we should bear in mind. In case anyone thinks this a hypothetical situation which never happens, I wish to point to the recent (offline) example of the Goldsmith's University Islamic Society, who severely disrupted a talk that they didn't like, at one point turning off the projector, and then brazenly claimed that they had been harassed - when arguably they were the ones who were actually doing the harassing. This example is especially relevant because the material those people were objecting to - criticism of Islam and its treatment of apostates - is precisely the kind of controversial material that our users ought to be able to post, both in main pages and on talk pages, as long as it is relevant in the context it is posted and it meets a site's policies and guidelines.--Greenrd (talk) 15:44, 20 December 2015 (UTC)

Anti-harassment training has been standard feature of HR departments for at least 30 years, and I have yet to see an agency taken over by Islamists as a result. Leaving the issues with the arbitration committee has already failed miserably; they have no training to deal with the problem and they would probably be the first to say they do not wish to be tasked with responding to it, not to mention the negative public relations aspects. This is exactly the reason we need concise definitions of harassment and standardized ways of dealing with it. —Neotarf (talk) 01:46, 4 February 2016 (UTC)