Communication · a life skill
Difficult Conversations
start from the story you'd both agree with
The hard part of a hard conversation is usually the first sentence. We rehearse it for weeks, and by then it has turned into a verdict: you never, you don't care. The other person hears an accusation and defends themselves. Start instead from the difference between your two views, the way a fair observer would put it. Give one example, ask for their side, and save your one request for after you've listened.
- Challenging
- High impact
- Learn it: 2-4 weeks practice
- One example, one request
Our own illustration of the “third story” from Stone, Patton & Heen's Difficult Conversations (1999). The lines in red are verdicts: each one's explanation of the other.
The method
Open with the difference, not the verdict.
-
Raise it while it's small
One sink of dishes is a conversation. Two months of them is a case you've been building, and it tends to come out as a closing speech. Pick a private, unhurried moment.
-
Sort the three stories first
Write down your story, your best guess at theirs, and the third story: how someone fair, who likes you both, would describe the difference.
-
Open with the third story
“I think we have different ideas about …” puts you on the same side of the problem, looking at it together, rather than across a table from each other.
-
One behaviour, one example
Name one moment you both saw: “On Thursday the sink was still full from Monday.” Leave out “always”, and leave their motives to them.
-
Ask, listen, then make one request
“How do you see it?” Hear their story out before you tell yours again. Then ask for one change, and agree when you'll check how it's going.
Sort the stories, open with the third, one example, ask, one request.
Example: Jo has been quietly cross about the washing-up for two months. Instead of the speech they've rehearsed, they say: “I think we have different ideas of when washing-up counts as done. On Thursday the sink was still full from Monday. How do you see it?” Ari explains the Sunday routine. Then Jo makes the request: “Could we try clearing the sink by bedtime for two weeks, and see how it feels?”
The opener bench
Sort the stories, then write the first sentence.
1 · Pick a situation
2 · Whose story is each line?
Yours, theirs (your best guess), or the third story, which both of you could agree with.
3 · The three stories
Your story
Their story, as you guess it
The third story
All sorted. Your first sentence comes from the green box.
4 · Write the opener
5 · How it opens
The checks are our own rules of thumb, not a test from the research. They read words, not tone, and they can't tell whether your example is fair.
Nothing you type is saved or sent anywhere.
The research behind it
It goes better than we fear.
Emma Levine and Taya Cohen wanted to know why people avoid honest conversations. In a preregistered lab study published in 2018, 106 pairs of people who knew each other well, about two thirds of them friends, came in together. One of each pair wrote down a piece of critical feedback for the other: one thing they thought the other person should do differently.
Before reading it aloud, they predicted how the conversation would go. Then they read it out, the two talked it over, and both rated what had actually happened.
The conversation was more enjoyable than expected, by almost a full point on a five-point scale (2.91 expected, 3.87 felt). It felt more connecting and more meaningful than they'd guessed, and it did less harm to the relationship than they'd feared. The people who heard the feedback rated it about the same, or a little better.
The biggest miss was about the other person. Before speaking, people expected their partner to be offended, defensive or judgemental. Afterwards they thought it had gone much better than that, and the partners' own ratings were milder still.
An earlier study in the same paper found the same pattern outside the lab: people asked to be completely honest with everyone for three days found it more pleasant and more connecting than others predicted.
These were people talking to someone close, about one piece of feedback, in a lab, and the authors note that some of their measures were not very reliable. The paper doesn't test the third story or any way of opening; those come from the book and from us. What it does suggest is that the dread is a poor forecast, and a poor reason to wait.
The idea
Why it works.
An opener that starts with a verdict asks the other person to defend who they are. They will, and the conversation becomes about their character rather than the sink. An opener that starts with the difference asks them to add what they know, which is usually the part of the story you were missing.
The example and the request keep it small. One moment you both saw is something to talk about; “always” is something to argue with. One request, made after you've listened, can be agreed to. A list of grievances can only be survived.
When it goes wrong
Habits that help.
- They get defensive. Slow down and ask what they heard. They may be answering a verdict they think is coming, not the words you said.
- You learn something new. Let it change your story. The third story you opened with was a draft.
- They raise a complaint of their own. Take it seriously, and agree to come back to it. Two problems at once tend to blur into one row.
- It goes nowhere. End it kindly and pick a time to try again. One conversation doesn't have to settle everything.
Go deeper
Keep going.
Practise this in HourLife.
In Journal: Guided templates. One page per day, written a note at a time, with guided templates, photos, tags, search, a cover and writing insights.
