Conflicted Page 19
When you’re submitting a question, CMV asks you not just to state your view baldly but to give a summary of why you believe it. That might include your reasoning, although the most valuable information, says Turnbull, usually has to do with how you came to believe it. Does your view have something to do with your personal life, the home you grew up in, a specific experience? A philosopher might argue that how someone came to a view is irrelevant information – that whether or not a belief is wrong or right can be established purely on logical or factual grounds. But Turnbull and CMV’s users have found that embedding the opinion in someone’s life has a way of improving the conversation.
Knowing something about where the submitter is coming from helps commenters shape their questions and arguments. The users who are most successful at changing views tend not to post off-the-peg arguments that could be used on anyone, but tailor their comments to the person behind the view. The more that commenters feel they’re talking to a real person, the less likely they are to denigrate the view and the more likely they are to show that they are truly listening – and people who feel they’ve been listened to are more open to having their mind changed. ‘What does listening look like in a text-based forum?’ asks Turnbull. ‘I think it’s about referring to someone’s reasons for their view and the path they took to get there.’
In most social media forums, he told me, people usually jump straight to the point of disagreement without taking time to understand each other first. The result is that they scrabble around on the surface of an issue without digging beneath. I think this! Well, I think this! ‘Sometimes it’s better to take your time getting to the point where you disagree,’ Turnbull said. ‘Start at the beginning, take a walk around the question, then see where you get to.’ Another benefit of asking submitters to spend time articulating their view is that it begins the process of opening their mind to flaws in their own position. As people discuss things in more detail, they often discover that they are less certain than they thought they were.
Commenters are required to disagree with the submitted view. Turnbull didn’t want this to be a place where people can come simply to have their views affirmed – after all, there are plenty of places for that. Crucially, however, the disagreement must be expressed courteously and respectfully. The first rule to be agreed on was the simplest: don’t be rude or hostile. Turnbull’s team of moderators are vigilant for any signs of rudeness or aggression. The prohibition of hostility, he says, is fundamental: ‘It’s very important people don’t break this rule because if they do it just throws off the conversation and makes it hard for anyone else to get through to the author of the post after that point. It changes the tone.’
This is worth dwelling on. Tone is sometimes discussed as if it’s a secondary characteristic of human interaction. (‘Why are you worrying about tone? Focus on the substance.’) But tone is more important than content. It goes deeper than words. It is the medium through which we express the relationship we expect to have with who we’re talking to. Your tone efficiently communicates how you see yourself in relation to me: as more intelligent or less; as dominant or deferential; as somebody with whom I can be serious or playful. And as we’ve seen throughout, until people establish a mutually agreeable relationship, a disagreement is bound to go badly.
Turnbull notes that when people break the rule against hostility on CMV, they often do so without realising and immediately apologise. ‘They want to change the view of the other person. So by being rude, it’s going against their own interests, because they’re less likely to get through to them.’ This should be obvious, but people often forget it in the heat of the moment. CMV’s most ingenious rule is one that keeps people focused on why they are there in the first place: submitters give awards to commenters who have shown them they are wrong. If and when a submitter decides that their view has been modified by a CMV conversation, they bestow, on the commenter who made the difference, a delta: Δ (in maths and physics, the delta sign signifies change).
A person with a high delta score gains status within the community, since they have a proven ability to change other people’s views. ‘The first time I got a delta,’ one participant told a researcher who has studied CMV, ‘it felt like a big deal.’ Commenters have a strong incentive to maintain good relations with submitters, because otherwise they won’t be able to persuade them of anything.
* * *
In the early days of the site, Turnbull and his fellow moderators noticed a problem: people who had their views swayed by a discussion had a tendency not to admit it. Despite adopting a different position to the one they started with, they would pretend, perhaps as much to themselves as anyone else, that they had pretty much held the same opinion all along. Turnbull wanted people to admit that they had been wrong. Not because he wanted them to feel humiliated – quite the opposite. He wanted to de-stigmatise wrongness. ‘Nobody likes to be wrong. It’s not a great feeling. But if you approach it positively, it’s a chance to learn, gain insight and lose some of your ignorance. It doesn’t have to feel like an attack.’ Over 2,000 years after Socrates tried to reassure Athenians, the message still needed sending.
When we talk about ‘changing my mind’ we tend to think in terms of a 180-degree reversal – a Damascene conversion. The trouble with this is that it puts a high price on the act of changing your mind. At CMV, participants are encouraged to award deltas merely when they feel that the discussion has led to a deeper understanding of the topic, or when they have learnt something from their challengers. ‘It helps to literally think in terms of view,’ says Turnbull. ‘If you’ve moved even slightly around, then you’ve got a new perspective.’
Turnbull has noticed that the people who are really good at changing minds – users with high delta scores – are skilled at asking productive questions. They don’t use questions as a way to flay the other person (‘What on earth makes you think . . . ?’) but to understand their view more fully. Sometimes their questions expose a contradiction at the heart of the submitter’s position, prompting them to rethink (personally, I know that the moment I hear someone summarising my point of view fairly is when I get the fear). But this only works when the questions spring from genuine curiosity – those who ask questions in a forensic, prosecutorial style are less effective at changing minds. As Turnbull puts it, ‘It might seem to contradict the whole idea of CMV, but starting with the intention of changing someone’s view can actually ruin the process a little.’
We’ve seen how the righting reflex can backfire. A better approach to changing minds is to become someone’s partner in an exploration. Successful delta-accumulators often exhibit a little uncertainty about their own position. The act of having your mind changed and the act of seeking to change someone else’s mind start in the same place, with questions that point simultaneously at the questioned and the questioner – don’t you think?
* * *
CMV is a fertile source of data for academic researchers interested in how and why people change their minds. A team of computer scientists at Cornell University analysed over two years of postings on the platform. They found that about a third of submitters to the site had their mind changed. That may not sound high, but in context of previous research on persuasion and attitude-change, it’s exceptional. The Cornell researchers ran numbers on which kinds of conversation resulted in deltas and which did not. Their findings echo and augment Turnbull’s observations.
For instance, the factor most associated with successful persuasion was using different words to those used in the original post. That’s intriguing, because it implies that to change a view, you need somehow to reframe the argument in different terms, putting it in a fresh context. It links to another finding the researchers made: the use of specific examples really helps to change minds, as does the use of facts and statistics; the best recipe is a combination of storytelling and hard evidence.
Longer replies tended to perform better than shorter ones (although not if the commenter is seen to
be soapboxing). A person’s view was more likely to be changed over the course of a conversation with a commenter, rather than by one devastating point. But if it wasn’t changed after five rounds of back-and-forth exchanges, it was unlikely to be changed at all (this has been useful to me in knowing when to bail out of Twitter arguments).
The researchers also found that ‘hedging’ helps. Arguments that included phrases like ‘it could be the case’ tended to be more persuasive than those that projected certainty. When a commenter signals with his tone that he’s not entirely sure of himself, the submitter lets their guard down. Weakness is power.
* * *
Experts on traffic know it only takes two or three drivers on a crowded highway to do something that disturbs the traffic flow for hundreds of cars to coalesce into a jam. The same principle applies to online debate. A small number of deliberately disruptive or bad-tempered participants can force their way to the forefront of everyone’s attention and before you know it, a new normal has been set, resulting in bad-tempered gridlock.
Most people don’t make a decision to be rude, any more than they are implacably determined to remain polite. Whatever environment we’re communicating in, we instinctively take cues from how others are behaving. That’s true at the basic level of a dyad: when two people are talking and one injects even a hint of hostility into the discussion, the other notices immediately and feels an urge to follow suit. Online, that dynamic plays out at scale. Without thinking about it, participants look around and ask, is this the kind of place where I’m expected to let fly with sick burns, or to engage in respectful debate?
This places a certain responsibility on each individual. These days we’re used to the idea that what we buy or how we travel has environmental effects. How we communicate has environmental effects, too. With each remark we make, we can choose either to improve or to pollute the discourse. In fact, what you say may be less important than how you say it. After all, you cannot know for sure that you are right. What you can be sure of – what you can control – is the example you set.
Improving the quality of online disagreement isn’t just down to individuals, though: well-designed rules make a difference too, as Change My View has shown us. Even the simplest of rules can help, providing people are made aware of them. Some worry that strict rules can have a chilling effect on participation and free expression. In 2016, Nathan Matias, a professor of communication at Cornell University, looked into this. He conducted a study with Reddit’s science discussion community, which had 13.5 million subscribers at the time. Matias arranged things so that users in some threads saw an announcement of the community’s rules, which include a prohibition on hostile language. Compared with discussions in which users were not shown the announcement, first-time commenters were significantly more likely to abide by the rules. Crucially, the participation rate of newcomers increased by an average of 70 per cent. They weren’t being put off by the rules. Online, or in the workplace, simply making rules of interaction visible improves the conversation for everyone, most of all for those who haven’t been around long enough to absorb the group’s norms through osmosis. Shared constraints create space for livelier disagreements.
Rules are important not so much because people need telling what to do or not do, but because they feel better about expressing themselves within a structure to which everyone is expected to adhere. Arguments can always get out of hand, either because people knowingly break rules or because they didn’t know the rules in the first place. But when people do get into a messy argument, they usually want to find a way out, and a clearly articulated set of rules can lead them there.
This is a principle that hostage negotiators understand. They often have to de-escalate situations that are on the verge of chaos. In order to do that, Professor William Donohue told me, they try and provide a structure through which the hostage-taker can express himself. ‘The negotiator enters this highly contentious, identity driven, emotionally fraught moshpit of confusion, and they have to impose order on it. They feed back what they’re hearing in a structured manner: “The first thing you’re concerned about is this, the second thing is this . . .”’ Experts in tough conversations are trained to do this, but as Donohue pointed out to me, it’s something skilled communicators do instinctively: ‘A good friend listens to your emotional diatribe and helps you find a shape for it. They turn your mess into something you can get a handle on – so that you can do something about it.’
Donohue likened this process to a political system. ‘It’s where democracy starts. The idea of Magna Carta was to take back rights from a capricious king who was making decisions on the spur of the moment. A democratic society needs due process, so that every demand or grievance isn’t settled by the whim of the king. In my research, I see the same forces at work. Negotiators and mediators create a map of the issues and propose a process for working them through. When that structure falls apart, it opens up the need for people to grab power and impose their will.’
13. Only Get Mad on Purpose
No amount of theorising can fully prepare us for the emotional experience of a disagreement. Sometimes your worst adversary is yourself.
Back to where we started: an anonymous hotel meeting room, somewhere deep in the English countryside, on a winter’s night. This is what I know: a young woman was raped in an underpass last Sunday. A video of the incident had been uploaded to a website from a phone that belonged to a local delivery firm with seventeen male drivers. The last text message sent from the phone pinged off a mast close to the home of this Frank Barnet, a man with a history of domestic violence.
The man I was looking at wasn’t really Frank Barnet, and he wasn’t really a rape suspect. He was an actor. Laurence Alison had offered me the chance to take part in the kind of role play that forms the backbone of his training for police interviewers. It was Laurence who briefed me on the case, addressing me as ‘Detective Inspector Leslie’.
So why was my stomach turning over? Because even the simulation of a high-pressure conversation has the power to trigger your nervous system, regardless of how much your rational brain might tell you it’s not real. Alison had warned me that role-play interviews often feel all too authentic. Hardened police interviewers can be driven to the edge in training simulations. In the real world, professional interrogators sometimes need therapy in the aftermath of an intensive period of interviewing.
What I took away from this brief but intense simulation is that in order to make a disagreement productive, you need to influence the other person, but the first person you need to influence is you. Mastering your own emotions, your own reactions, is the hardest skill of all.
* * *
FB: I don’t give a fuck about people, why are you talking to me? Frank Barnet. Why me?
The man playing Barnet was Lloyd Smith, a regular participant in Alison’s courses. Lloyd has been interviewed by more police officers than almost anyone else alive. He has been a Rwandan warlord, an Islamist terrorist and a Brazilian gangster. He has been a murderer, a paedophile, a rapist. He is sometimes aggressive and borderline violent, sometimes charming and evasive, sometimes determinedly silent. For each case, he absorbs a character profile written by Alison, together with information about the crime or crimes the character has committed. It is his job to give the interviewer as hard a time as possible, and he does so with tremendous skill and guile. Lloyd has not studied interrogation tactics in an academic context, but he has taken part in so many interviews, with so many different interviewers, that he has a profound understanding of their dynamics.
He is also just an extremely good actor. To me, in that room, he wasn’t Lloyd. He was Frank Barnet. I explained to him, in a voice that I tried and failed to make self-assured, that we were interviewing several drivers who worked for the delivery company.
FB: So you believe I might be a rapist?
I didn’t know how to answer. How do you answer that? I just repeated myself, telling him, in a more convoluted way
than last time, that he was one of a number of people we were talking to. I sounded like a politician evading a straightforward question.
FB: It’s a yes or no question. Do you believe I could be a rapist?
Finally, I said yes.
FB: You should know I don’t have no love for you people.
IL: Which people?
FB: The police. You’ve been pestering me and beating me since I was a kid.
I tried to return to the question of what he was doing that Sunday afternoon, but Barnet had other ideas.
FB: Do you think I’m a cunt?
IL: No, I don’t.
FB: Don’t you think a rapist is a cunt?
I ignored the question. In an officious-sounding voice, I asked again about that Sunday. Barnet peered at the ring on my left hand.
FB: You’re married, ain’t you?
IL: Yes.
FB: If I raped your wife, would you consider me a cunt?
IL: [stunned silence] Yes, probably.
I already felt hot, nervous and itchy with discomfort. Now, I felt furious. Why the hell did this guy think it was OK to talk about raping my wife? Why was he asking me questions? I wasn’t the one suspected of a crime. Laurence, who had been observing, paused the interview, and asked me how I felt. I told him. He nodded. The challenge, he said, was to avoid getting sucked into Barnet’s force field – into the conversation he wanted to have. ‘You’re thinking, God this guy is being a real jerk to me. But a really good interviewer is able to step back and think, “Ah, that’s interesting – this guy is behaving like a jerk. What is that about?”’