Photo by Dan K
Was talking to Sanjana about a number of things, including flaming and hate speech online. By chance I saw a very relevant article in the Times today. As I said to Sanjana, “I think part of the problem is anonymity. I donâ€™t tell people to fuck off in real life cause my identity and my place in this culture means something to me. A lot of perfectly sane people are complete sociopaths on the road, when theyâ€™re hidden behind a windshield. No one walks like that. I think what we call civility is heavily dependent on facial and non-verbal communications, which simply doesnâ€™t exist on the Internet.”
As the article says,
In a 2004 article in the journal CyberPsychology & Behavior, John Suler, a psychologist at Rider University in Lawrenceville, N.J., suggested that several psychological factors lead to online disinhibition: the anonymity of a Web pseudonym; invisibility to others; the time lag between sending an e-mail message and getting feedback; the exaggerated sense of self from being alone; and the lack of any online authority figure. Dr. Suler notes that disinhibition can be either benign â€” when a shy person feels free to open up online â€” or toxic, as in flaming.
The emerging field of social neuroscience, the study of what goes on in the brains and bodies of two interacting people, offers clues into the neural mechanics behind flaming.
This work points to a design flaw inherent in the interface between the brainâ€™s social circuitry and the online world. In face-to-face interaction, the brain reads a continual cascade of emotional signs and social cues, instantaneously using them to guide our next move so that the encounter goes well. Much of this social guidance occurs in circuitry centered on the orbitofrontal cortex, a center for empathy. This cortex uses that social scan to help make sure that what we do next will keep the interaction on track.
Research by Jennifer Beer, a psychologist at the University of California, Davis, finds that this face-to-face guidance system inhibits impulses for actions that would upset the other person or otherwise throw the interaction off. Neurological patients with a damaged orbitofrontal cortex lose the ability to modulate the amygdala, a source of unruly impulses; like small children, they commit mortifying social gaffes like kissing a complete stranger, blithely unaware that they are doing anything untoward.
Socially artful responses emerge largely in the neural chatter between the orbitofrontal cortex and emotional centers like the amygdala that generate impulsivity. But the cortex needs social information â€” a change in tone of voice, say â€” to know how to select and channel our impulses. And in e-mail there are no channels for voice, facial expression or other cues from the person who will receive what we say.
As the article mentions our social throttles are designed for face-to-face constant feedback situations. Even voice has a high emotive content. Text, on the other hand has a great flatness of affect. That’s why I prefer SMS, because it’s more shielded and safe in a way. Online commenting lacks everything that leads to flaming – feedback, identity, authority.
Towards that end, I’ve been trying to figure out ways to engender civil discussion without excessive comment moderation. Moderation is necessary for large blogs, both for spam and flames, but it’s time consuming and annoying. Plus, especially vituperative flamers will post again and again and again. As the article says, ‘Consider an experiment, reported in 2002 in The Journal of Language and Social Psychology, in which pairs of college students â€” strangers â€” were put in separate booths to get to know each other better by exchanging messages in a simulated online chat room. While coming and going into the lab, the students were well behaved. But the experimenter was stunned to see the messages many of the students sent. About 20 percent of the e-mail conversations immediately became outrageously lewd or simply rude.’
What I’d like is to get at least 10 percent to behave with normal politeness. One way is to give commenters more identity. In all cases I’d like to make this as automatic as possible. That can be accomplished through
- Avatars: Avatars are icons to represent people. Gravatar was a system where people choose their own avatar, but it went down for a long time, and people here never really adopted. I just saw a new plugin which generates a cute monster icon for each email. May try that out.
- Location: Another thing is that expats can be extra crazy. I’d say that 99% of the people that tell me to get out of the country are not in the country. There’s a plugin for that. IP to Nation displays little flags based on their IP address. May try that but it’s a pain in the butt to install
Some more radical stuff would be using some Ajax to dynamically moderate the comment, maybe turning the box red and displaying a warning if they kept typing obscene or flagged words. As some sort of feedback. You could count on other people to moderate flamers, but it doesn’t really happen that way.
All in all, I think the way towards civility is attaching some identity, and identity costs to commenting. I don’t think the problem is that flamers are necessarily bad people, I just think that online discourse is outside of the normal social environment that we’re psychologically equipped to deal with. Perhaps understanding the psychology more will lead to some solutions.