Disagreeing Well

What was measured

Yeomans, Minson, Collins, Chen and Gino built a computer program that reads what someone wrote in an argument. It picks out the words and phrases that show you are actually willing to hear the other side. They called this being 'receptive'.

They tested it on government executives arguing about hot policy questions. Some people came across as more open, both to the program and to the person they were arguing with. Those people were rated better teammates, better advisers and better representatives.

Here is the catch. People's own ratings of how open they were did not match. You could feel open and still not sound open.

The person you were arguing with was the one who decided. If they felt heard, they liked you more, trusted your judgement more, and were more likely to change their mind.

The move

Aim to sound open, not just to feel open

Feeling fair-minded inside your own head is worth nothing to the other person. They can only go on what you say. Assume your good intentions are invisible until you say something that shows them.

Executives who came across as more open to the person they were arguing with were seen as better teammates and advisers, and were more persuasive. How open they thought they were did not predict any of that.

The field test

The same team looked at Wikipedia editors. Editors argue in public, in writing, all day, about what stays on the page.

Editors who wrote in a more open way got fewer personal attacks back. Sounding open early in a thread meant the thread was less likely to turn nasty later.

That is the part to keep. It is not just that open people are liked more. Opening in an open way changes what comes back at you.

The move

Your first message sets the ceiling

How you open predicts how it ends. Most of your influence over an argument is spent before you have made your own point.

Wikipedia editors who wrote in a more open way got fewer personal attacks from editors who disagreed with them.

Limits — The researchers watched what happened on Wikipedia. They did not run an experiment. So we cannot be sure the open writing caused the calmer replies, rather than the other way round.

Move 01 — say it back

The things that made someone sound open were boring. Name the other person's view out loud. Say what you agree with before you say what you don't. Say 'I think' and 'maybe' instead of 'obviously'. Say what you want, not what you don't want.

Saying their side back to them does several of those at once. And unlike nodding, it can be checked. They can tell you whether you got it right.

The move

You have not earned the right to disagree yet

Say their position in your own words, well enough that they would agree with your version, before you say a word of yours. If they correct you, good. You just found out you were about to argue with the wrong thing.

Naming the other person's view, saying out loud what you agree with, and softening your own claims were among the things that made people sound open. And sounding open went with being liked, being trusted, and changing minds.

Move 02 — ask how

Fernbach, Rogers, Fox and Sloman asked people to explain, step by step, how a policy they backed would actually work. People then felt they understood it less than they had thought. And their position got less extreme.

Only the 'how' question did this. Asking people to list their reasons for their view changed nothing. In a third experiment, explaining how a policy works also made people give less money to groups campaigning for it.

'Why' asks for a reason. Everyone has reasons ready. 'How' asks you to describe the machine. Most people find out they can't.

The move

How, not why

Ask it as a real question. Then answer it about your own side too. The point is not to catch them out. If you can't explain how your own idea works either, that is the lesson.

Asking people to explain step by step how a policy would work made them feel they understood it less, and made their position less extreme. Asking them to list reasons did neither.

Limits — See the next section. A 2022 study found this only works for arguments about what a policy would do, not for arguments about right and wrong.

Where it stops working

Sloman and Vives went back to that finding in 2022 and found its edges. Explaining still made people feel they understood less. That part held up.

But people's views only softened when the issue was about what would happen. Some views are about right and wrong, the kind people will not trade. On those, explaining changed nothing.

This is the most useful thing in the lesson. It tells you which arguments the tool is for. And if someone will not move when you ask 'how', you are probably not arguing about what would happen. You are arguing about what is right.

The move

Work out which kind of argument you are in

If asking 'how would that work?' moves them, it was about what would happen. If it does nothing, it is about their values. Asking again is no longer curiosity. It is pressure. Stop asking and listen.

Explaining still made people feel they understood less, as first reported. But it only softened positions on issues about what would happen, not on issues people held as a matter of right and wrong.

Move 03 — stop trying to win

Broockman and Kalla sent 56 canvassers to knock on the doors of 501 voters. Each conversation lasted about ten minutes. The canvassers did not judge, and they asked voters to imagine being in someone else's shoes.

Voters were picked at random to get the conversation or not. Those who got it became a lot less prejudiced against trans people, and the change was still there three months later. The authors point out that this drop was bigger than the drop in anti-gay prejudice across the whole of America between 1998 and 2012.

It is worth knowing why this study is careful. An earlier study on a similar question turned out to be faked. These researchers had helped expose that. So they wrote down their plan before they started and made all their data public. This is the honest version of a finding that once was not.

The conversations had no argument in them. Canvassers asked, listened, and told their own stories. Nobody was trying to win.

The move

Trade winning for being someone they would talk to again

You will almost never change a mind in one go. What you can do is end the conversation as someone worth coming back to. That is the only way the change ever happens.

One door-to-door conversation of about ten minutes, built on asking people to imagine being in someone else's shoes, made voters a lot less prejudiced against trans people. The change lasted at least three months. Voters were assigned at random, so this was a real experiment.

Limits — One issue, one city, one canvassing organisation. Later studies found the effect is real but harder to get when you try it on a bigger scale.

If you use this to get something, it stops working. The doorstep conversations worked because the listening was real and the canvasser had no trick waiting behind it. Acting open in order to win is just a slower argument. People can feel the difference. The Yeomans study found the person across from you is a better judge of how open you are than you are.

So the way this fails is not by being too soft. It fails when you perform being open while keeping score, and then wonder why nobody moved.

None of this means you have to take every view seriously. Deciding a position is not worth your time is a fair answer. It is far more honest than pretending to weigh it.