Civil Beat is trying something new with Civil Comments, including offering readers the option to use a pseudonym when making comments beneath articles.

I have argued that anonymity should be reserved for only the most special cases, a stance supported by Civil Beat’s own policy on using anonymous sources in news stories. The policy says Civil Beat only uses anonymous sources “for substantive, factual information; not for opinion, personal criticism or incidental information. Sources need to have direct knowledge of the subject involved.”

Reader comments generally are not held to such standards for a variety of reasons, including:

• Comments are not considered part of the journalistic discourse. But they are, of course, since they represent an emerging form of citizen journalism in the digital age. They are also directly juxtaposed as part of the journalistic conversation and, practically speaking, provide the last word on any journalistic piece.

• News organizations do not have the resources to make sure everyone commenting on a piece in an open digital system – freely accessible to everyone around the world – will be forthright and use an authentic name. So they shouldn’t do it that way.

In a slightly regulated system with some boundaries, like, for example, what the Hawaii State Public Library System uses for the distribution of cards, users acknowledge some of their personal identity in exchange for participation rights. If, after an editor’s review, such a user warrants the protection of a pseudonym, they should get it. But if not, they should use their authentic name.

• Journalistic organizations generally like to foster civic debate, as a way to encourage participation in their channels (and increase engagement and loyalty to their news brands). This philosophy includes opening space on the channel for responses of all kinds about the journalistic content.

This is a noble idea, in theory, but just as our First Amendment-protected “free” speech has exceptions and journalists vet “letters to the editor,” reader comments have bounds that are often crossed. That includes the insertion of unrelated marketing links, intentional derailing of productive conversations and all sorts of random abuse, threats and bullying.

Civil Beat and Civil Comments should be commended for trying to foster better reader conversations, but offering blanket anonymity to do so is a questionable strategy, which prompted me to ask Civil Comments why they thought it would work. The answer I received was not convincing.

In the original argument for the anonymous comments, a spokesperson identified only as christa_m wrote: “Studies show that ‘real-name’ policies do more harm than good online.” Since I was skeptical of these – anonymous – “studies” and the definitive nature of that statement, which seemed counterintuitive, I asked christa_m for the citations of her evidence, which she provided as:

• “Digital Social Norm Enforcement: Online Firestorms in Social Media,” by academic and industry researchers from Zurich. To begin with, Plos One, the online journal of the original article, requires authors to pay for publication, as a type of vanity press. Even if that scholarship turned out to be solid, this particular study actually analyzed social media posts related to online petitions, not reader comments about journalistic content (and the research was conducted in Germany, not the United States, creating questions about cultural crossover). The piece contradicted the central claim made by christa_m in many ways, including noting research showing the use of real names helps a participant gain recognition, increase persuasive power and mobilize followers. Those citations were among multiple references to additional academic research that argued for authentic names in public discourse. The authors also acknowledged their research design did not allow their conclusions to be drawn as cause-and-effect interpretations. One of this article’s primary findings was that named individuals were “more aggressive” in their comments, in this sample, compared to anonymous individuals. Yet, the tone softened in their discussion section, and the strongest argument made in regards to the harmful-or-not policy assertion was that “real-name policies do not necessarily prevent online aggression in social media.”
• “Surprisingly Good Evidence That Real Name Policies Fail To Improve Comments,” a 2012 TechCrunch article, actually referenced a 2007 policy implemented in Korea that appeared to have mixed results, over its one year of study, for a variety of reasons. But the switch to authentic names for commenters on large Korean websites, according to this report, also “reduced swearing and ‘anti-normative’ behavior at the aggregate level by as much as 30 percent.” The story included a comment by former Facebook marketing director Randi Zuckerberg: “I think anonymity on the internet has to go away. … People behave a lot better when they have their real names down.” Therefore, this article appears to have a dramatically misleading headline.
• The final source, “Who is harmed by a ‘Real Names’ policy?” on the Geek Feminism Wiki, primarily is a list of potentially vulnerable audiences online, such as women, children and people with disabilities, with about half of the references in this piece coming from Wikipedia and the other half from popular media.

So I suppose I can see why christa_m initially shielded the identity of her sources, because they seem much more convincing in the anonymized label of “studies” than they do in their actual forms.

I also quickly searched Google Scholar for academic articles on this topic and found about 12,000 related scholarly papers, including many directly studying online news, such as those that address how commenter anonymity has complex implications for reader perceptions, how comment quality improved when anonymity was reduced, that specific policies on online news websites “regarding user registration, moderation of comments, and reputation management systems are effective facilitators of civil discussion,” and that non-anonymous comments were significantly more likely to be civil as anonymous comments.”

These scholarly articles acknowledged gray areas in this research, including anomalies, exceptions, and the variances of perspectives, such as how anonymity might offer some positive effects on the system – increased participation, for example – tempered by significant negative implications, such as an increase in drive-by flaming.

So if your goal is simply more comments, regardless of quality, then anonymity might be a good strategy. But if your goal is a higher-quality conversation, then anonymity is a questionable choice. A recurring example referenced in this type of research was the “broken windows” theory, in community policing, in which a concerted focus on small indiscretions also eliminated the larger problems.

In other words, disorder, as in not knowing who is going to spring from the shadows and attack you anonymously, creates fear in a community (like with the Civil Beat reader commenters). That fear compels people to stop commenting, and with less participation, the social-control system of commenters peer-policing one another weakens. The trolls gain more power, which cyclically leads to worse and worse behavior.

So I encourage you to think again about anonymity in reader comments and why you might support it. The research, if anything, points us away from it. I think it’s evident – at least to anyone who has participated in such forums – that in most situations the benefits of anonymity are not worth the costs.