Disinformation, trolls, and #pman: more serious than Skittles (DRAFT)

DRAFT!  Work in progress!  Feedback welcome …

Final version to appear on The Seminal

If you’re joining the series already in progress:

  • Lessons from Skittles for poets and activists introduced the series, briefly described how Skittles’ Twitter-centric viral marketing campaign caught fire, and concluded that Activists and poets — and anybody else who wants media attention without spending a lot of money — should consider including Twitter in their plans.
  • Mr. President, do you like Skittles? and Activism at the speed of Skittles, looked at how a small group of activists used Twitter to highlight a question about homeless vets and it was answered on the White House blog less than 48 hours later. The conclusion: Things happen very quickly in the Twitterverse
  • What rhymes with Skittles? shifted attention to poets, looking at my brother Greg’s 30 Poets/30 Days project and the #kidlit #poetry hashtags, and concluded that Everybody knows: fun rules.

This week, we’ll turn our attention to a more serious matter: the protests in reaction to Moldova’s disputed election results. As a “flash mob” congregated in Chisinau’s central square, the Piata Marii Adunari Nationalethe, and dramatically hoisted the EU and Romanian flags on Parliament, the #pman hashtag became a key communication channel for supporters and media around the world.

There’s been a lot of discussion about Twitter and other social networks roles in the protest.  Evgeny Morozov’s Moldova’s Twitter revolution is NOT a myth on Foreign Policy and Ethan Zuckerman’s Unpacking the “Twitter Revolution” in Moldova on My Heart’s in Accra are excellent starting points, and the Moldova and pman Diigo group has a lot more links.  Here, I’d like to focus on something Dumitru Minzarari brought up on Political Moldova: an apparent trolling and disinformation campaign on Twitter.

Dumitru notes that in preparation for April 10, a day when another protest seemed likely, several messages were repeatedly being posted: warnings not to go to the main square because bloodshed was likely, statements that there were no leaders and so the protests were futile, and most interestingly

3. do not take your ID tomorrow with you, if you are going to join the protests. While the first two messages are basically used to scare, intimidate and discourage the potential protesters, the third one is really the interesting one. Anyone without an ID, can be detained by the police, until their identity is established, so not having an ID will offer the police a legitimate pretext to arrest you, even if you protest peacefully.

On top of that, from time to time anti-protester Twitter profiles like SuperGodiva and mixman2009 generated so much traffic on #pman that it threatened to overwhelm more useful discussions.  As cerberus_md tweeted:


Can Skittles help?

The challenges of open channels

Twitter hashtags are an “open channel”: anybody with an account can send a message.  The Skittles tweets from bluewiggirl and baratunde were just the tip of the iceberg.  By the second day of Skittlemania, hate speech and flaming far outweighed the good clean fun.

So Skittles yanked the Twitter feed off their home page, and redirected to Facebook instead.  A lot of people saw this as a retreat.  As Charlene Li pointed out at the time, though, this was probably the plan all along.  I imagine a conversation something like this in the pitch meeting where Agency.com was trying to convince Skittles to take the risk on a Twitter-centric strategy:

Skittles exec: I dunno, is there a chance that people will start posting garbage about Skittles to Twitter?

Agency.com: oh yeah, in fact we’re counting on that happening to help build traffic.  So we’ll temporarily make the home page adults-only, and then when things get too nasty we’ll just switch things to Facebook.

Or something like that.  So this week’s lesson from Skittles:

Expect interference — and have a plan to deal with it.

What are the options?

Skittles’ tactic of abandoning the hashtag when things got nasty can be useful for activists as well — sometimes it’s better to shift communication to a channel that can be better protected.  It’s far from the only option, though.

There are a lot of excellent examples on #pman of the straighforward approach of alerting people to the possibility of trolls and disinformation — and identifying specific trolls.  Of course, this leads to credibility questions: who to trust?  Often, it’s pretty clear from the discussion on the channel and a quick look at users’ profiles.  By the second or third time SuperGodiva described the protests as a zionist-rothschild coup, I think most people had made up their minds.  By contrast, Cezar Maroti had started up a Facebook group supporting the protestors, and had been quoted in several articles.  His statements had a lot more credibility.

Other activism campaigns illustrate a few other possibilities:

  • #dontgo responded to a progressive “counter-insurgency” by introducing a filtered feed hosted on a separate site
  • TweetLeft includes some basic “troll filtering” that works surprisingly well
  • Twitter Vote Report developed a custom “sweeper” user interface allowing volunteers to weed out malicious or badly-formatted posts
  • Swift River, a Ushahidi project, is a general crowdsourced filtering mechanism, currently being used for India Vote Report

Which ones to apply in a particular situation depend a lot on the people involved and how much up-front planning you can do.  Over time, I expect we’ll see a lot more tools to make this easier.  For now, though, it’s something activists need to keep in mind — and be ready to react.

Coming attractions

Our next post in the series will shift attention back to the US and an upcoming Congressional race.  Here’s a teaser:

(image coming soon)