Portico_logo

FEATURES
FOR FACTS’ SAKE

FOR FACTS’ SAKE

Think facts are dead? Reckon experts are a busted flush? Stop right there. This is not a time for despair. Knowledge and expertise matter – and here’s why.

Words William Ham Bevan Images James Richardson

IS4_P2

‘The USA is the highest taxed nation in the world.’ An oft-repeated yet even more oft-disputed ‘fact’; in reality, the United States has average or below average taxes compared with similar economies. Whatever metric you employ, using direct or indirect taxation, there still seems to be no way to make the claim work.

Each November, the lexicographers at Oxford Dictionaries come together to choose their word of the year. Reading down the list provides a pithy timeline of recent social history, from ‘sudoku’ via ‘credit crunch’, ‘omnishambles’ and ‘selfie’ to last year’s ‘youthquake’. But of them all, ‘post-truth’ perhaps holds the most significance.

Though first recorded in 1992, the phrase may well have reached its peak last year, when three books entitled Post-Truth were released in the same week. Along with its close relatives ‘alternative facts’ and ‘fake news’, it speaks of a world where facts are of less consequence in shaping opinion than emotions and personal beliefs, where political figures may reject awkward evidence as they please, and where public discourse has fallen out of love with experts and expertise.

But is it really a new phenomenon at all? Professor Tamar Garb, Director of the UCL Institute of Advanced Studies (IAS), says: “It has become almost a kind of banality now for everybody to say, ‘Oh, the world’s full of fake news.’ Lying has a very long history and there are many different ways in which lies have been manifest.

“We need to identify what is specific about our moment. Has something shifted, or are we just in a long continuum of lies? Has politics always been economical with the truth, or is there something distinctive that is going on now, culturally, technologically, that has created a kind of legitimacy for lying?”

Every year, the IAS identifies two themes that address urgent political and social issues, and explores them in forums and events. Professor Garb says: “This time, we’ve chosen ‘lies’ as one of our themes. It’s indicative of the way in which questions around truth, deceit and falsehood in the news have become so visible in public discourse.”

Melanie Smallman is Deputy Director of the Responsible Research and Innovation hub at UCL and a former Communications Adviser to Defra’s Chief Scientist. She is uneasy with the idea that large numbers of people are rejecting science and enlightenment values; rather, she believes, they are finding that their own experience often doesn’t chime with the privileged ‘expert’ narrative.

She says: “The evidence has been building over the last 30 or 40 years that people don’t base their opinion on cold facts, and that facts look very different depending on where in the world you are and who you are. If people’s experience of the world doesn’t resonate with the experts’ description, they’re going to stop believing the experts, stop acting on their advice and turn elsewhere – to voices telling stories that more closely resonate with their experience.”

As an example, Smallman cites the debate around technology and innovation, which is usually about growth, jobs and the other benefits of the ‘knowledge economy’. However, many people in communities surrounding hi-tech hubs may experience only negative effects, such as exclusion from new work opportunities, widened pay inequality and prohibitively expensive housing.

All this feeds into the growing ease with which vested interests can foster doubt and spread deliberately false information. She says: “What has really altered things is social media. Anybody’s voice can be heard; you don’t have to own a powerful channel, or go through editors. Those checks and balances have gone.”

Smallman believes that what’s required is not so much better information about science, but a deeper conversation about how science can be used better and more responsibly. She says: “If there is a problem with trusting expertise, perhaps we need to think about expertise as much as that issue of trust. It’s not necessarily a problem with how the public think about things. It may be how we as experts analyse, talk and present things.”

Away from the scientific and economic spheres, some of the biggest news controversies have concerned the interpretation of history – from revisionist narratives about the Holocaust to atrocity denial in more recent conflicts. Eric Gordy, Senior Lecturer at UCL’s School of Slavonic and East European Studies (SSEES), has conducted high-profile research into the public memory of war in the former Yugoslavia.

He says: “One research project I did was on ways that people account for facts; how they understand the recent past. I found an interesting pattern, because there were moments when facts that were previously unknown would come out, or a new piece of evidence came out that couldn’t really be disputed.

There will always be contestable facts. But you can look at their foundations and consider whether they’re reasonable

WHY FACTS MATTER

Professor Tamar Garb says: “It’s very, very dangerous when politicians refuse to acknowledge the research that goes against their own prejudices and their own ideological stances, and end up discrediting the bodies and experts that produce that knowledge. The problem with political rhetoric is that people choose the facts that suit their argument.

“There are always many different facts out there, and you can make a narrative out of them to produce any argument you like. But true expertise is hugely important: without the respect for expertise within a society, it’s very difficult to have the rule of law and to have a consensus about how you can run things equitably.

“Expertise is hard-won, and it takes a long time to become an expert. We have to build our own knowledge on the expertise of those who have come before us. Knowledge is always shifting and changing, and professionals need to be constantly upskilling, learning afresh and reinvigorating their understanding of their own fields and interrelated areas of inquiry. That’s what universities do – and that’s what you need in a mature democracy.”

IS4_P3

‘Britain is paying the EU £350m a week.’ The now infamous side-of-a-bus claim made by the Vote Leave campaign during the UK’s EU Referendum campaign. In reality, as official figures revealed last year, the UK’s net contribution has fallen to its lowest level for five years, at just £8.1bn in 2016-17 – or about £156m a week.

IS4_P4

‘Migrants are more likely to claim benefits in the UK.’ Evidence suggests that EU migrants are actually more likely to be in work than UK nationals, and therefore not claiming unemployment benefits, though they may claim more in-work benefits.

IS4_P5

‘Vaccines have been shown to cause autism.’ Authored by Andrew Wakefield, the scientific limitations of a paper linking the MMR vaccine and autism were clear when it appeared in 1998. But it took 12 years for the paper to be finally retracted and labelled “fraudulent”, during which time the damage to public health continued, fuelled by unbalanced media reporting and an ineffective response from government, researchers, journals and the medical profession.

My pet solution is that everyone should learn research methodology and design. It’s not extremely complex, yet this is seen as radical, romantic – even idiotic

“In the short term, sometimes this had a strong effect. People who had previously been in denial said, ‘Yes, this is important.’ But generally, these were not permanent effects. The facts don’t seem to have changed people’s minds over the long term. They’d develop a strategy to recontextualise the new fact – not to deny it, but to say that it’s unimportant if you put it in the perspective of some larger facts, or to discredit its source.”

Dating back to the late 50s, this notion of ‘cognitive dissonance’ – that people marginalise facts that apparently go against their entrenched beliefs – features in many discussions of the ‘post-truth’ age. The extent to which it takes place is much contested.

Alan Renwick, Deputy Director of UCL’s Constitution Unit, says: “The research results are mixed. It’s probably more likely to happen on issues that are polarised, such as the EU referendum debate in the UK, where people identify strongly with one side and use it to build up their identity.”

In his current work, Renwick is looking at the ways that governments could improve the quality of public discussion during election and referendum campaigns. And, despite the potential for confirmation bias, he believes that access to fair and impartial facts could help citizens make better sense of the issues at stake.

He considers three types of intervention to raise the standard of debate: prohibiting campaigners from making misleading statements; providing high-quality, impartial information for citizens; and creating spaces such as ‘citizens’ assemblies’ in which representatives of the public could debate the issues and report back to society at large.

“The first approach would bring out obvious concerns about free speech,” he says. “The second is the easiest to implement, and has been done with some measure of success; but there’s the broader issue of whether the establishment or ‘experts’ are trusted. That’s one reason to be interested in the third possibility, where instead of relying on some kind of commission, you involve ordinary citizens in the process – like a jury, where the evidence is set out and they can quiz the people presenting it. It’s a fairly new idea that only very few places, such as Oregon [where ‘citizen initiative reviews’ have been held at each electoral cycle since 2010], have tried.”

All three approaches rely on some form of fact-checking, whether carried out by electoral authorities or citizen panels. “We shouldn’t give up on the idea that there are facts, and things can be true or false,” says Renwick. “In the political realm, there will always be things that are contestable, such as claims about what happens in the future if you take certain policy approaches. But you can look at the foundations of these claims and consider whether they’re reasonable. Are they inferring from what we know, or have they just been made up?”

Fact-checking is a skilled, specialised and time-consuming task, and UCL researchers are looking into whether technology can be harnessed to facilitate the process. Sebastian Riedel, Reader in Natural Language Processing and Machine Learning, says: “Full automation of fact-checking is a long way away, but recent developments in natural language processing could make it a little easier for humans to do it.”

One way is to teach a computer to flag up statistics in text that are of suspect accuracy; but even this is a complex operation. “Take the simple example of proving the statement that the unemployment rate is 42 per cent,” says Riedel. “The machine doesn’t know which country, which year – it needs to learn that context.

“Then a phrase like ‘unemployment rate’ can have many meanings. It could be the number of people who have tried and failed to find work, or the total number of people who don’t have work. Should it include students, homemakers and people who don’t need to work? To be able to get a machine to understand this, and do all these types of reasoning steps, is still relatively far away.”

This strand of research is being brought to market by Factmata, a company founded by alumnus Dhruv Ghulati (MSc Computer Science 2016) and advised by Riedel. It received a €50,000 grant from Google last year, and has since secured further finance to develop its tools to help people fact-check news and media content.

“We’ve recently handed in our prototype to Google,” says Ghulati. “It’s an experiment in showing how we can scan a web page for claims and give readers facts about those claims so they can investigate them. For example, if a reader has a statement saying unemployment is rising, we can show a chart of the actual unemployment rate, so the user can judge whether it’s true, or get additional context.”

His hope is that with further research and investment, Factmata will be able to widen its remit. “We want to build a broader fact-checking and news platform, perhaps focusing on checking fake news or scoring news stories for accuracy. But readers will always be involved in some way. Factmata will always be about giving people tools to help them make that judgment themselves.”

Ultimately, the key to making these judgments – and the greatest corrective to fake news and misinformation – is education. And while universities such as UCL have their part to play, encouraging critical thinking and a respect for facts at an earlier age is a concept that appeals to many, including Eric Gordy of SSEES.

“That’s my pet solution,” he says. “Everybody should learn research methodology and research design. Then it becomes very easy to recognise when information is being gathered selectively or people are misusing statistics. None of this stuff is extremely complex; any person can learn research methods. Yet this is seen as a radical, romantic idea – or even an idiotic one.”

Fact or fiction? Do you think expertise still matters? Give us your views via Twitter @UCLAlumni or on email using alumni@ucl.ac.uk

IS4_P1

‘The number of people in extreme poverty is the rising globally.’ In truth, in the past 30 years, the percentage of people in the world who live in extreme poverty has decreased by more than half.

  • Contents 2017/18Contents 2017/18
  • DeconstructedDeconstructed
  • We Want to Hear From You...We Want to Hear From You...
  • UCL Leads on DementiaUCL Leads on Dementia
  • Free RadicalFree Radical
  • Festival A HitFestival A Hit
  • Jeremy Bentham SpeaksJeremy Bentham Speaks
  • Extra CurricularExtra Curricular
  • Experts And FactsExperts And Facts
  • For Facts’ SakeFor Facts’ Sake
  • A Different Type of Capitalist MarketA Different Type of Capitalist Market
  • The Big CThe Big C
  • Trump In The Age Of Captain AmericaTrump In The Age Of Captain America
  • The Front Line Of JusticeThe Front Line Of Justice
  • Calling All UCL GraduatesCalling All UCL Graduates
  • CloisteredCloistered
  • The Strong RoomThe Strong Room
  • This idea must dieThis idea must die
  • London  vs  WorldLondon vs World
Portico Issue 4. 2017/18
UP FRONT
FEATURES
UCL+