Oxford psychologists to help Twitter to combat hate speech
Fake news and targeted messages are splitting us into partisan cliques where no debate is possible
TWITTER has enlisted the help of top Oxford psychologists to tackle the spread of fake news and hate speech.
The company is working with Prof Miles Hewstone and John Gallacher from Oxford and Dr Marc Heerdink from the University of Amsterdam.
The move is part of Twitter’s search for algorithms to better distinguish between hate speech and conversations that break the “norms of politeness”.
Meanwhile academics from the US, the Netherlands and Italy will look at the effects of “echo chambers”, online groups of like-minded individuals that perpetuate hostility against outsiders.
If you had to use a single word to sum up what is happening to politics all over the democratic world, that word would almost certainly be polarisation. From an America torn between hating Trump and loving him to a Europe increasingly split between antiimmigrant and mainstream parties, and a Britain where opinions on Brexit become ever more entrenched, people who have long accepted being governed together are becoming more ferociously divided about what governments should do.
Constant debate and a variety of viewpoints are healthy, but deep and permanent polarisation is not. Human nature means that once we have entered into a bitter argument, we are less willing to climb down, accept a different view or see any merit in the arguments of the other side. As a result, the ability of the US Congress to pass sensible cross-party laws has evaporated and the chances of the EU even holding together for the long term are sinking, while the prospects of any Brexit deal getting through Parliament continually diminish.
Such events will in turn lead to even greater disillusionment and bitterness, fundamentally weakening democracy itself. Part of the answer is open debate, and listening to contrasting points of view. Yet it is now commonly accepted that the movement of so much discussion, news and political communication into the digital world through Facebook, Twitter and other sites is adding to the polarisation.
People are increasingly living in an echo chamber of their own views, being sent more opinions that they already like, more news – true or false – that they are inclined to believe, and more adverts that target their particular worries or prejudices. The internet was meant to be the greatest ever expansion of human knowledge and understanding. In politics, however, it has also become a cause of escalating narrow-mindedness, extremism and intolerance, both on the Left and Right.
A functioning democracy requires open debate, which means assertions that voters are hearing can be challenged by others. Yet at the moment targeted messages can be sent to individuals without anyone else knowing about them, or knowing where they’ve come from, or having any chance to show they are false. Such messages might disappear once you’ve read them. You might end up deciding how to vote because of “information” that was designed specifically to mislead or enrage you. Or you might not vote because of a message you were sent. The indictment of Russian interference in the last US election refers to many ethnic minority voters – inclined to vote for Hillary Clinton – being quietly and skilfully influenced not to vote at all.
Democracy will not survive being turned into a system of hidden falsehoods or “fake news”, transmitted from unknown sources to selected people to inflame or reinforce their bitterness or their apathy. The learning, challenging and debating in the open that are at the heart of free and collective decisions are being lost.
Polarisation is becoming so severe that people on the winning side of recent elections, like President Trump, deny that all this is a problem. But to many others it is blindingly obvious that something has to change, and fast. Technology is developing so quickly that you can now be shown totally convincing footage of someone saying words they have never actually uttered. Any free society needs rules of behaviour designed to preserve it from such abuse.
Facebook and Twitter have announced new rules to try to prevent foreigners buying advertisements during US elections, and to identify paid political ads. But laws are now needed to get ahead of these issues and to tackle reinforced polarisation more broadly. The interim report of the Commons digital, culture, media and sport committee, published on Sunday, is a very good start.
Such reports from MPS come out so often they are widely neglected. This one, however, should be taken seriously and much of its thinking incorporated into the Government review of digital advertising and inaccurate news also underway.
It contains good ideas about how to resolve the argument over whether a company like Facebook is a “platform” – not responsible for how they are used, or a “publisher” – held to account for what they allow to appear. Create a new category that is neither, says the committee, but establishes clear legal liability for tech companies to act against harmful and illegal content on their platforms.
I would encourage this committee and ministers to think even more radically in some respects. For instance, it recommends that the algorithms used to determine what news to show to each user should be audited by a regulator. What about requiring such algorithms to be published? And saying that it is necessary to provide news and comment from some alternative way of thinking, so that people are not forever living on a diet of views and advertisements that confirm everything they already think?
It goes on to advocate new rules on political advertising, “so that it is obvious at a glance who has sponsored that campaigning material”. But again, it is worth thinking about going further. In Britain we have always banned paid political advertising on television, even when TV viewing was by far the main medium for news and discussion. It has helped save British politics from being as expensive and divisive as it can be elsewhere.
In a recent book, the Labour activist Tom Baldwin has made the case for extending that ban to social media, arguing that anything short of that will inevitably be open to manipulation, abuse and being quickly out of date. Such a ban would not stop parties and candidates producing videos and messages that were shared widely if they were powerful or interesting, but it would stop the quiet breaking of spending limits or exploitation of data about individuals.
To agree with Tom Baldwin, who has campaigned for almost everything I disagree with, I have to overcome my own predisposition to be opposed to whatever he supports. Yet in the interests of not having a closed and polarised mind myself, I think I can manage it. And it might just be in the interests of our wider democracy to adopt such a British solution to a threat undermining our long-held attachment to open and fair debate.