Don’t give in to the Twitter mob: Social media is too often just an echo chamber
Greater Manchester Police staged a simulated terror attack in the Trafford Centre last week. As with many real life atrocities, the carnage began with the mock-bomber crying “Allahu Akbar!” Following a storm of protest on Twitter, the police felt forced to apologise.
Almost at the same time, a frenzied chorus rose up demanding the resignation of the BBC’s political editor, Laura Kuenssberg, for having had the temerity to suggest that the local election results were something less than a complete triumph for the Great Leader and Teacher, Jezza. This campaign was halted by the extremely sexist nature of many of the comments posted by left wing Twitterati.
The way the Kuenssberg saga ended is in fact very unusual. Following a storm of outrage on social media about a statement or an action, the “guilty” party almost invariably confesses his or her crime and issues a heartfelt apology to the raving crowd.
Social media is a new and radically disruptive technology. It is hardly surprising that traditional institutions and social norms have not yet adapted to the challenges which are raised. Many thousands of voices were raised against the police for their allegedly racist opening cry, with virtually no one springing to their defence. It seemed that public opinion was firmly against them, and so they bowed to pressure and apologised.
But Twitter, along with other social media outlets, is in many circumstances simply an echo chamber. When the polling booths closed in the Scottish referendum in 2014, many in the SNP leadership were convinced they had won. Their researchers had carried out seemingly sophisticated analysis of social media, and concluded the Yes campaign was ahead. The actual result gave rise in turn to all sorts of conspiracy theories, bouncing backwards and forwards between die-hard pro-independence Scots.
Similarly, late last summer, the US Army carried out a routine training exercise called Jade Helm 15. This sparked a torrent of concerns on social media, a prominent one being that the federal government was planning to invade Texas and civil war was imminent.
Guido Caldarelli at Lucca and Gene Stanley at Boston published a paper in January this year in the prestigious Proceedings of the National Academy of Science. They find that such problems as described above are widespread in social media, with users frequently forming communities of interest which foster confirmation bias, segregation and polarisation. Biased narratives fomented by unsubstantiated rumours, mistrust and paranoia proliferate.
How do we know whether to take a trend on social media seriously, or whether to just dismiss it as a bunch of fruitcakes egging each other on? Santa Fe-based scientists Rich Colbaugh and Kristin Glass (I am currently working with them) have found that a topic which has only a small number of mentions in each of several different social media communities is potentially far more significant than one which has a huge number in just one. Public bodies need to learn how to differentiate between social media topics, and not to just routinely capitulate to the mob.