The Wikimedia community has a problem with harassment. The situation is that a significantly large number of people say that they would participate in Wikimedia community activities, except that they perceive the Wikimedia community to be hostile, unfair, and sometimes that it explicitly targets them for harassment. I recognize that there should be some response to these complaints.
My opinion is that the biggest part of the complaints is not harassment – it is the perception of harassment. The chief cause of complaint is when a new Wikimedia contributor comes to Wikipedia, tries to engage the website in an inappropriate way, then is told in a polite way by some community member that what they did is not allowed. The new user is very likely to feel that if their submission is rejected then they have been treated unfairly. Most new contributors with rejected submissions have no understanding of what happened – they typically assume that the rejection was arbitrary, inappropriate, malicious, and personal.
The typical experience of a rejected submission being perceived as harassment is that a user adds sentences, but fails to provide a citation to the source from which those sentences came. The rule on Wikipedia is that the encyclopedia is supposed to be a summary of published thought and not original research, but most people including academics, librarians, and communications professionals do not grok this. “Grok” is the only correct word – very often noncompliant contributors can repeat the rules but even though they say “follow every sentence you add with a citation”, they will not practice or understand the concept they are able to communicate. It takes a few years for people to reflect on Wikipedia for them to understand this, so for this kind of harassment, we need civility policies, better communication, training, tailored messages for new users, and other responses.
The most traumatic kind of complaint and it is still fairly common is actual targeted personal attacks by one user on another. Most commonly this happens by men to women, and gender is part of the reason why this happens. The Wikimedia community has an extensive set of processes and protocols for managing harassment, and the system of support for people in Wikimedia projects with harassment complaints is extensive, deep, and compares favorably with any other support system. The problem is that it still is not enough to manage the burden of the problem, and the current system would be impossible to scale because it depends too much on too many people’s personal attention. It is a flattering, efficient, and effective system but new infrastructure is required to meet the demand.
Considering these two problems as example cases which need to be addressed by any new harassment response system which is developed, it is my opinion that the chief deficiency in going forward is the lack of means of reporting that problems exist. I propose that Wikimedia projects have a single communication channel by means of which anyone may report any kind of complaint, and that certain individuals – paid staff whose discretion can be trusted – route the complaints to various response queues. In all cases, anyone who has a complaint receives a complaint that their comment will be sorted. In those cases in which the response will be inaction the persons at the queue will tell the person directly that they are unable to assist further, and give them a canned response describing available resources which the person may research, use, or develop on their own. Finally, trusted researchers who meet a high standard – perhaps the United States standard for clinical or social research – may access a database of the complaints so that basic data which is currently unknown may be described. Some interesting questions which are currently impossible to answer include how often does someone complain about a Wikimedia process, when people complain how common is it that they complain about harassment, are certain users more frequently the target of complaints in multiple unrelated circumstances, and in what places or circumstances on the website are complaints most likely to happen.
Backing up a bit more to generalize the issue – I am not convinced that harassment or complaints happen more often in Wikimedia projects than they do in other online forums. A recent realization just starting to be voice more since about 2012 is that all online communities of a certain size with open-ended user interaction will have members who report experiencing harassment. Harassment in online forums happens with much greater frequency than it does in face-to-face social forums away from online spaces. Definitely the big players are dealing with this – Facebook, Twitter, Google – where they provide spaces for the public to interact. What is more relevant to me is that volunteers for nonprofit organizations experience harassment and unfriendliness in the places they create for collaboration in activism and projects. Communities which have spoken about this include Tor; Electronic Frontier Foundation; open source or free software projects including Debian and Ubuntu; and most social groups which have successfully recruited a user base such as women’s online forums, hobbyist groups, and entertainment discussion forums. The harassment and unfriendly talk on Reddit has gotten special media attention. I do not see other people saying that all of these problems have the same cause, but I say this, and I feel like I have enough experience with enough of these projects to be sure that a solution for any one of these would probably work for the others, assuming that it could be adapted for other cultures and scaled to size. None of these projects have workable solutions in place at this time in my opinion. I feel that what Wikimedia websites provide now is the best attempt at a solution. I do not think Wikimedia projects should be specially criticized for the harassment problem when it is a common problem inherent in online spaces.
It would be undue for any Wikimedia community group or the Wikimedia Foundation to invest too much money or resources into addressing harassment. Typical proposals are to hire paid staff moderators to assist with complaint investigation and to police bad actors. This response instinctively seems best to most people, but in my opinion, they come to suggest this response because this is what works for off-line communities in physical spaces. Reasons why this would not work in Wikimedia projects include that there is no rank or hierarchy in the community, and the community resists regulation when in many disputes both the target and accused have a claim for justice. Not infrequently, the most contentious harassment cases are the ones in which the victim was clearly treated unfairly but at the same time there are broad claims that the victims engaged in offensive or aggressive behavior themselves. In my opinion, it is irrelevant whether victims really do sometimes have bad behavior or whether the accusation of bad behavior is just another harassment action – what is relevant is that narratives in which the roles of victim and attacker should be anticipated, and that a trend in harassment stories is that large crowds routinely develop in opposition to each other for some contentious cases. Having a paid moderator step into these situations to pass judgment probably will rarely lead to peace or a favorable resolution, so while online communities might demand moderation of this sort, stewards of a community should consider the costs and likely outcomes of providing paid moderators.
Ideally, online communities should be able to police themselves and maintain their own peace. In many ways they do. All online communities socialize their members to behave in a way aligned with an online culture which will be unlike any offline culture. For a community to monitor that its participants are socializing as they should, and to police the members who are not aligned with its social expectations, I propose that infrastructure be developed to permit anyone who wishes to report a noncompliant member, and for the complaint to have some chance for being reviewed and acted upon.
The major reason why no online community forum has a communication channel for receiving social complaints and crisis notices is because of the equivalent precedent for brick and mortar organizations. If a brick and mortar organization gets a complaint or is informed of some crisis, like sexual harassment, threat of physical harm, discrimination, illegal activity, stalking, creepy behavior, or general rudeness, then for physical spaces and offline communities, the expected response is that the managers of the physical space should govern their space, do what is necessary to eject wrongdoers and bring safety to their space, and resolve the problem themselves quickly and effectively. In online forums such a response is not possible. In offline spaces, like for example a bar, if a woman says that she is being threatened and stalked then without a lot of controversy, the bar would ask the offender to leave, maybe call the police, and perhaps even the accused would be comfortable saying “Regardless of whether I am right or wrong – I should go and leave this situation.” What works very well in a workplace, or bar, or on the street, or in other physical contexts does not make sense online where the victim can be in one country and the offender can be in another. Simply removing the players from a virtual space does not end the ties between them, as ties in virtual spaces can be more persistent and entangling than even close offline social or workplace relationships. No manager of an online space wants to put themselves into the position of responsibility that managers of offline spaces routinely and happily accept. Society still has not caught up to the concept that online “places” are not the same as offline places. “Online places” are a metaphor for offline places – it is a mistake to think that going somewhere online is like physical travel, or that online personas and relationships are like offline human interaction, or that what works offline would work online. Because of this, an implication of making a communication channel for complaints about the online behavior of others is that such a channel would receive complaints for which there is no social precedent for effective responses. Likely, a communication channel for receiving complaints would not be useful in a response model that imagined that the moderator or manager of an online space should be responsible for resolving the complaint, because it probably is not possible for the manager of the online space to give any meaningful response to bring resolution. Furthermore, if there were a communication channel for receiving complaints about an online space, then the manager of that space would need to be comfortable with the existence of an accumulation of the complaints about goings-on in that space. The manager would need to be comfortable saying, “We get suicide threats, deadly stalking reports, awful sexual harassment reports here… and we do nothing. We are comfortable with that because we have another process. We are not responsible.”
Of course I want all complaints to get a response, but the existence of a complaint reporting system should not unduly obligate the manager of the space to have a new burden. Something which the Wikimedia Foundation already does is make referrals for suicide threats that happen in its community spaces. It happens in online communities that, no matter what sort of community it is, inevitability someone will come to the online space and say that they want to commit suicide. This is often an easy problem to address – the manager of the space, whether the Wikimedia Foundation or Club Penguin, reports to local police whatever user data they have to identify the person and get them help. In the meantime the person gets a notice directing them to resources like a crisis hotline which they can call for counseling. They may also get an email containing reading materials and links to more services. A bigger problem is when the threat comes from an underserved region. Some organizations, like the Wikimedia Foundation, are large and well-funded enough to have contact information and social connections for crisis services in the various districts of Iran or Uzbekistan or China, and if some Wikimedia participant says that they are feeling suicidal because someone found out that they were gay, or a political dissident, or a Wikimedia contributor in those places, then there is a little that the Wikimedia Foundation might be able to do only because it is prepared to serve its international community. Many online collaborative communities have management which only think about services local to them, even though the multiculturalism and internationality of online communities has a diversity of participants which is unprecedented in history and unimagined in arts and mainstream media like magazines, books, and movies. The experiences that people can have interacting with online communities still have almost no representation in popular social narratives. For this reason, there is no social infrastructure which dictates that community groups based in San Francisco should provide counseling to individuals in rural areas on the other side of the world for their supporters who live in economies which are embargoed by the United States and that the support should be in a language which is not a focus for routine activities in the organization’s management.