Back in the day, I knew a smart, handsome fellow student — let’s call him John. John was a talented musician and great company: the sort of person who’d always make a night at the pub more cheerful. We bonded over our mutual undergrad-philosophy-seminar disdain for religion, but other than his jaundiced view of the Catholic Church, John held resolutely mild-mannered, moderate views on most topics — including politics.
So I remember being surprised when, after falling out of touch, I found he had turned into a hardcore libertarian, of the requiring-driving-licences-is-infringing-my-freedom type. But people go through phases. I was sure it wouldn’t last.
Indeed, John’s views did change. He popped up on Twitter in around 2016, suddenly a huge fan (and emulator) of Milo Yiannopoulos and his flamboyant brand of Right-wing trolling. Grimly fascinated, I checked his page every so often, and over the years watched him become more and more extreme, until there was no further he could go. He was a slavish fan of Donald Trump, before getting disillusioned and attacking the President for “betraying” his supporters. Having previously flirted with the alt-Right, he went all the way, making and posting anti-Islam jokes, then anti-Semitic memes, then straight-up White Nationalist propaganda. Eventually, Twitter banned his account. He seems now to have disappeared (though could easily still be posting under an anonymous account).
John’s journey from milquetoast liberal to foaming-at-the-mouth Neo-Nazi genuinely disturbed me. His profile didn’t fit that of the stereotypical online troll: he wasn’t some asocial basement-dweller who couldn’t make it in the real world. He came to mind as I was reading a new book by Cardiff University’s Professor of Criminology, Matthew Williams. The subtitle of The Science of Hate is “how prejudice becomes hate and what we can do to stop it”. Was there something inherent to John’s brain or psyche that made him a prime candidate for the slide into extremism? Or would the kind of internet radicalisation he clearly received have the same effect on any of us?
After describing his own awful experience of being assaulted by homophobic thugs in London, Williams presents other horrific hate crimes, such as the 2016 stabbing attack at a disability care home in Japan which left 19 dead and many more injured (the perpetrator faces the death penalty). He then digs into the research from psychology, neuroscience, and elsewhere to try to explain the fear-anger-hate-suffering conveyor belt.
The trouble is, a lot of that research just isn’t very good. Although he is at times sceptical of results from neuroscience — the brain is an absurdly complex organ and our best attempts to understand it are as yet pretty primitive — Williams lets his guard down for a whole host of dodgy-looking or now-debunked studies. We hear, for instance, about a study where white people’s brain “fear centres” lit up in a scanner when they were looking at a black person’s face — but only while they were listening to N.W.A. and not Slipknot. This is a silly study for several reasons, not least that it had a mere 23 participants and has never been replicated.
We also hear about the “Macbeth Effect”, where people made to remember one of their past immoral deeds were more likely to think of cleaning-related words on a verbal test and were more likely to wash their hands as they left the lab (just as the eponymous queen washed out the damned spot). It hasn’t held up well in follow-up research, and might not really exist — but you wouldn’t know that from Williams’s treatment of it.
And we hear, similarly uncritically, about the Implicit Association Test, which is widely used to identify unconscious biases in the lab and at work. Williams informs us that people’s culturally-shaped automatic preferences for one race or another on this computer-based test can predict prejudiced behaviour (which might include, he says, voting for McCain rather than Obama in 2008). But a review of all the data from 2019 showed that changing people’s implicit attitudes appears to have little-to-no overall effect on their explicit views or behaviours.
And so on. Williams does cover some better studies on prejudice and hate, but The Science of Hate — and perhaps by extension, the science of hate — is shot through with veins of low-quality, unconvincing research. If so many of our attempts to apply scientific methods to understand prejudice fail so badly — and yet are still written up by one of the UK’s foremost authorities on hate as if they explain a great deal about our behaviour — we’re in trouble.
In the past year, we’ve seen with greater clarity than ever before that we can’t always rely on published scientific research. The problem is not confined to medical studies, like those of Covid-19. The patchwork of highly-variable studies cited by Williams reminds us that, wherever we turn, scientific literature fails to provide us with a sure footing.
This is partly a result of the perverse incentives that pervade science: the pressure that so many scientists feel to publish more and more research, with flashier and flashier results, regardless of how solid or robust that research is. Williams could’ve shown a few more doubts about the studies he cites, but the problem goes beyond him: we should be able to rely on the contents of the scientific literature to help us understand the big questions — it has, after all, been through the peer-review filter. But alas, peer-review fails so often that it’s no longer a reliable signal of quality — if it ever was one.
Interestingly, by the end of his book, Williams appears, at least partially, to agree. “There is no getting away from the reality”, he writes, “that predicting the behaviour of individuals is an imprecise science, full of known unknowns, and unknown unknowns”. Quite. But he then assures us that in the book he’s relying on solid, “state-of-the-art” science — a statement that, given the studies I mentioned above, rings rather hollow.
Having given those caveats, Williams goes on to offer advice on how we might prevent and deal with hate before it gets out of hand. One of the more specific ideas is based on so-called “Terror Management Theory”, a concept from psychology stating that reminders of mortality will tend to make people want to defend their worldview — and maybe lash out at those who are different. On this basis, Williams suggests that to reduce hate we should educate children to better cope with death. But (you guessed it), Terror Management Theory has come under serious fire in recent years when large-scale attempts to replicate its basic tenets failed.
So much for that. What about Williams’s other solutions? They include: questioning our instinctive, stereotyped thoughts about other people; putting ourselves in other people’s shoes; bursting our filter bubbles; and always reporting incidents of hate that we encounter. These are good ideas. They’re also common sense. A scientific study is hardly going to conclude that reinforcing our stereotypes is the best way to reduce prejudice. But, as Williams notes, most of these solutions will only work on “suitably motivated people”. Try telling the aforementioned John to put himself in the shoes of the Muslims he’s berating online. Try telling him to emerge from his filter bubble when he believes that a worldwide Jewish conspiracy controls the internet.
Of course, a list of common-sense ideas for dealing with hate doesn’t hurt, but one does worry that the “scientific” framing ultimately devalues the science itself. If you were always going to come up with those same recommendations, but pretend (or convince yourself) that you arrived at them through scientific investigation, it raises some fairly serious questions about the point of all that research. It risks reducing science to a mere prop: something that can be used to buttress pre-existing beliefs and commitments, but which you’d never really think of using to question them. This “conclusions first, research later” approach is, of course, a long way from the ultra-sceptical, disinterested ideal of science.
There’s a clichéd way to title an academic paper that begins with “Towards…”. “Towards a dynamic theory of strategy”. “Towards controlling of a pandemic”. It’s intended, I think, to emphasise that the research is far from final, and is merely a stepping-stone on the way to a larger theory or a more solid answer. Perhaps Williams’s book should have been titled Towards a Science of Hate, since — on the basis of the research he cites — there is not, yet, any solid scientific explanation for what goes so tragically wrong for people like John.