This series proposes the establishment of a universal protocol for reporting online abuse. The intention of the protocol is to handle the entire lifecycle, from the initial complaint to resolution; it should specify a standard data structure which would allow for outside reporting.
Machines and humans see the Internet differently. At the machine level, two systems that are communicating are able to do so reciprocally. One system can send a message to the other with the expectation that it can get a response. At the human level, however, this does not hold: one person can send a person a message without any return address. This basic asymmetry has been at the heart of most of the abuse on the Internet.
Originally, administrators focused on “network abuse,” whereby network resources were overburdened by delivering large quantities of unsolicited commercial email (spam). Efforts like the Network Abuse Clearinghouse have been focused on spam. As the Internet grew in popularity in the 1990s the domain of abuse has expanded to include interpersonal complaints (harassment, stalking, exploitation) as well as intellectual property concerns (copyright, trademark). Any petitioner seeking redress for abuse had to rely on the simple expectation that the Internet WHOIS database contained a working email address (and other contact information) for the domain administrator– and that the administrator would address the complaint. (Under the Digital Millenium Copyright Act, complainants may bypass the publisher and ask the ISP simply to take down the website).
With the growth of usage– in the number of users, in the diversity of usage, and in the intersection of communities– it is possible the range of complaints has expanded as well. In particular, there may be claims of harassment that may not be perceived as litigious, and thus these do not receive the necessary prompt attention of an administrator or publisher.
We consider here that the use of email alone may be inadequate for managing complaints. Since the growth of the web, software architects have developed higher-level content and workflow standards for the exchange of business or other structured information. There are two leading advantages to this approach. First, data is well-formed by the original petitioner, without requiring third-party formatting. Second, the structured format lends itself to being stored in a database for convenient perusal. Thus a tardy administrator or publisher who has in the past sat on emails now can face public pressure – whether from community members, advertisers, or other press institutions – to respond to complaints.
We call this the Protocol for Online Abuse Reporting, or PONAR and describe it within.
In my research at Civilities, I’ve come across several cases — three in the last many months– which inform the development of PONAR (Protocol for Online Abuse Reporting). Each involved an aggrieved party (in some cases, still mostly anonymous) who was harassed by anonymous aggravators online. In all cases, I have been able to contact at least one of the parties, in order to understand the case better.
- In 2004, a Delaware man contributes posts to an online guestbook run by the local newspaper. An anonymous poster, who, the man suspects is his neighbor and fellow city councilor, hurls insulting posts at the man (accusing him of “obvious mental deterioration”). The man and his wife were able to get a court order to force the ISP to disclose his harasser’s identity, and the man appealed to the courts to prevent it. The state Supreme Court decides not to force disclosure. (“Cahill” — I have corresponded with Julia Cahill)
- Over 2006-07, multiple female students at various law schools throughout the U.S. have their pictures posted, are written about, are ridiculed, and sometimes threatened with assault, in a discussion board that they never joined. Two Connecticut women initiate a public outreach and have filed a civil defamation suit against 28 anonymous posters and one former administrator of the board. (“AutoAdmit” — I have corresponded with Jarret Cohen, admin of the site)
- In 2007, a Colorado woman, a popular author, and blogger finds herself the target of harassing speech– via private emails, on her own blog, and on a new blog set up by people she knows which allows anonymous comments. Perceiving these together as death threats, she opens up a police report and pursues a public outreach as well (canceling her talk at an industry conference). The campaign reaches enough people through the national press that several considerations are proposed as institutional remedies. She withdraws the police report and withdraws from blogging. (“MeanKids” — I have corresponded with Joey K., one of the admitted aggravators)
- In 2007, a California woman, and municipal employee is “outed” by anonymous posters on an anonymously-published blog for allegedly having what would be a lawful relationship with the retiring city manager. The woman complains about the allegations, and the anonymous publishers excise her name from the comments out of deference to her privacy. The city attorney later sends a C&D matter (of questionable legal basis), and the anonymous publishers remove additional posts. The anonymous publishers occasionally find themselves redacting the woman’s name from the comments. (forthcoming — I have spoken with both the women and the publishers)
There are several common patterns to these. All cases, the forum (even the one sponsored by the newspaper) were set up in a way to allow anonymously (and sometimes, untraceable) postings which contributed to a “rowdy” atmosphere. The aggravators used the cloak of anonymity to say things they wouldn’t have said in responsible public forums. Nonetheless, in many of the cases, either the posters or the publishers ultimately regretted what they posted when they heard the claims of the aggrieved parties. The exception is the Cahill case, where the Delaware newspaper saw that Cahill was already “fighting back” on his own in the forum.
The remedy suggested by the Delaware Supreme Court was that more speech (by the aggrieved party) can remedy harmful speech. On an interview for On the Media, Jane Kirtley, Director of the Silha Center for the Study of Media, Ethics and Law at the University of Minnesota, suggested that the proper solution for Cahill was to: “Do your own blog. Make your own posting. Take the person on in his own environment, in cyberspace.” In another context, Susan Crawford, an ICANN Board member, blogged, “In an era in which anyone can be a publisher, libel law seems much less relevant — rather than sue, you can just write back.”
These opinions fail to recognize that the Internet forums where the insults originated were never level playing fields. The aggrieved party often finds herself besieged by anonymous commenters, posting at multiple places, and at multiple times around the clock. This leads to a couple of devastating effects: one, the general fear of another aggravator lurking around the corner; second, that any outsider is susceptible to hearing only part of the story, and will likely prefer to believe the most outlandish of the accusations. Furthermore, in the cases besides Cahill, the targets had never participated in the insulting forums, to begin with. In two of the cases, the victims felt that they were further subject to what we’ll call the “synthetic libel” of Internet search engines, which can paint an immutable character picture. The victim must resort to search engine optimization tactics to “restore” their good name. It is highly disconcerting for a judicial system to suggest more publicity for an aggrieved party seeking more privacy.
Instead, it is the point of the judicial system is to provide a level playing field. Both parties present their cases and do as public testimony. The court can suggest remedies such as ordering the removal of harassing web pages or award damages.
Kirtley and Crawford are likely articulating the point that it is impossible for all cases of harassment to be handled by the judicial system, and thus should be handled by extralegal means. Two of the above four cases have led to civil lawsuits, which gave them the added notoriety for me to hear about them. Most cases don’t.
Granted, the courts can provide only as much as a level playing field as the law allows. According to U.S. law, a plaintiff may have to prove the falseness of allegedly slanderous statements– devastating if the statements regard sexual morality, as they do in many of the above cases. This has not always been the law for every society. In 19th-century England, sexual slander was governed by the ecclesiastical courts, as it was seen as an offense against the church. The defendant could not prove his innocence by claiming that the statements were true (it could only lessen his punishment), and thus the plaintiff (a woman 90% of the time) need not be called to defend her own sexual behavior. After the ecclesiastical courts were abolished in 1855, the protections against sexual slander did not return to Britain until the 1891 Slander of Women Act. As with common law, the truth could be a defense, so the special protections were lost. (see Sexual Slander in Nineteenth-Century England: Defamation in the Ecclesiastical Courts by Stephen Waddams, 2000.)
This is another common thread in the above cases. None of the cases began as manifestations of criminal behavior (just as in real life: not every hurling of insults leads to a criminal action). In the MeanKids and AutoAdmit cases, sexual threats (stalking, rape) were posted at some point, but neither led a criminal indictment. The victim’s aid organization Working to Halt Online Abuse reported that just 4% of abuse cases they received in 2006 were handed off to law enforcement (The numbers do not tell on how many led to indictments).
The point of handing off to law enforcement is a process and a protocol. What we aim to establish with PONAR is to develop a protocol that can be used to handle the communications and evidence gathering either in advance of or in lieu of, legal means. This will be discussed in the coming sections.
There’s another pattern that needs to be highlighted. In two cases (MeanKids and AutoAdmit), the aggrieved parties were disappointed by the tardiness of the responses from publishers, and thus resorted to “publicity outreaches,” which succeeded (garnering national press), only to be viewed cynically as “PR campaigns.” This, unfortunately, colors the perception of the cases in the public eye. The point of PONAR is to be expeditious enough so that a PR campaign would not be fully necessary.