It’s no good Twitter and Facebook serving as an ad hoc international court



“Everyone who thinks at all has noticed that our language is practically useless for describing anything that goes on inside the brain.”
– George Orwell

In January, Twitter banned the account of Donald Trump after tweets allegedly inciting violence at the US Capitol. Facebook suspended the then US president’s account as well, recently announcing that the de-platforming would last a further two years.

This month Twitter deleted a tweet by Nigeria’s president Muhammadu Buhari, deeming it to be in violation of its abusive behaviour policy, which prohibits content that promotes “a desire for death, serious bodily harm or serious disease against an individual or group of people”. In retaliation, the Nigerian government tweeted (without a shred of irony) that it was suspending Twitter.

As has often been pointed out, Twitter and Facebook (and any other for-profit business) have the right to turf people off their platform. Anyone, president or ordinary person, has to agree to Twitter and Facebook’s rules in order to use to their service. Some have drawn an analogy with a restaurant. If a customer starts to kick over the tables or spit at other people, they can be ejected from the restaurant without apology.

But there’s a problem with seeing Twitter and other social media platforms as equivalent to a restaurant. The Financial Times’ Africa editor David Pilling has noted (paywall), “Twitter is less local restaurant and more global commons, the modern-day equivalent of the town square. It likes to see itself as a democratic marketplace of ideas. Yet who is Twitter, a for-profit US company, to determine which elected leaders of sovereign nations are free to speak?”

That’s true.

Add to that the problem with tweets that aren’t in English or use English words that mean something quite different in local situations.

For example, as I found when I was in Kigali, Rwanda, the word “cockroach” in that country means something very different. It’s a reminder of the terminology broadcast on Rwanda radio 30 years ago, inciting genocide of the Tutsis. Mr Pilling rightly notes that “It is not obvious how such companies can determine what crosses the line in foreign countries about which they know little and where no algorithm is equipped to make sensible judgments. People can post in literally thousands of languages and in numerous scripts such as Arabic, Armenian or Ge’ez. Even posts in English have different cultural connotations.”

Clearly, it’s not feasible for Twitter and Facebook to serve as an ad hoc international court, policing speech and intention. They don’t have the credentials, the authority or the skills.