Agree What You're Talking For
Name which of the six conversation types you're in before you start talking
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 6
- Confidence
- 78%
Harkness's Rule One is to agree what you are talking for. He maps conversations into distinct types, each needing a different mindset. A listening conversation is pure information transfer. An emotional conversation adds shared feeling with no endpoint of resolution. A values conversation evaluates what is important. A fairness conversation asks what is just, and quietly rests on one of three bases: equality, desert (what someone deserves), or need. A prediction conversation is where both people want the same outcome but disagree on the route. Sitting above all of these is the meta-conversation, where you talk about the talking itself. Naming the type stops mismatched agendas, and gives you an escape hatch: when any conversation breaks down, bail out and start talking about how you are talking.
Origin
Drawn from Tim Harkness's book '10 Rules for Talking', where Rule One is to agree what you are talking for. He illustrates it with everything from asking a partner about their day to Brexit and household chore negotiations.
Core principles
- 01Most conversational conflict comes from two people running different agendas at once
- 02Conversations have distinct objectives, and the type dictates the right skillset
- 03Fairness disputes hinge on an unspoken basis: equality, desert, or need
- 04When a conversation stalls, you can bail up a level and talk about the talking
How to run it
- 1
Recognise the conversation has a purpose
Before engaging, ask what this exchange is actually for. Most breakdowns come from skipping this and assuming the other person shares your goal.
Pro tip If you can't state the purpose in a sentence, that's your first clue the conversation will drift.
- 2
Classify the type
Decide whether it's a listening, emotional, values, fairness, prediction, or meta conversation. Each demands a different posture: listening transfers information, emotional conversations have no 'now I understand' endpoint.
Watch out Don't treat an emotional conversation like a listening one and wait for a resolution point that never comes.
- 3
Check for agenda alignment
Confirm both people want the same thing from the exchange. Conflict is far easier when the goals match and gets tangled the moment each party pursues a private agenda.
Pro tip Ego is the classic hidden agenda: the moment advice feels like an insult, people stop listening.
- 4
For fairness disputes, surface the basis
When arguing what's fair, name whether each side is reasoning from equality (equal effort), desert (I had a hard day, I deserve a rest), or need (I don't need the carpet as clean as you do). The clash is usually a clash of bases, not of facts.
Pro tip Making the basis explicit often dissolves the argument, because people realise they're measuring fairness differently.
- 5
For prediction disputes, find the shared goal
In a prediction conversation you may both want the same end (a happy country, a clean house) and only disagree on the method. Recognising this reframes an argument as a technical problem to solve together.
- 6
Bail to the meta-conversation when stuck
When you get jammed in any of the other types, stop and talk about the talking: 'this isn't working, let's talk about how we're talking.' This is the master conversation that rescues the others.
Pro tip Treat the meta-conversation as a fire escape you can reach for at any moment, not a last resort.
In the wild
Harkness describes the domestic bargaining over who cleans what. His own household splits the work equally by effort and time, which resolves it. But he notes not everyone uses equality as the basis for fairness: some argue from desert ('I had a stressful day, I deserve time off') and some from need ('I don't need the carpet as clean as you need it, so why equal effort?'). Naming which basis each partner is using turns a recurring, energy-sapping row into a solvable disagreement about the standard being applied.
→ A perpetual chore argument becomes tractable once the hidden basis of 'fair' is made explicit.
Harkness reframes Brexit not as a values clash but as a prediction conversation: 'we all want the same thing, we all want a happy country, we all want people to have jobs, we all want community.' The genuine disagreement is technical, about which course of action delivers those shared goals. Seeing it this way reveals the two sides are less different than they appear.
→ A polarising political fight is recast as a solvable question of method rather than of fundamentally opposed values.
Common mistakes
Diving in without agreeing the purpose
Starting to talk before establishing what the conversation is for guarantees mismatched agendas and talking past each other.
Treating fairness as one thing
Assuming everyone measures fairness by equality misses that others reason from desert or need, so you argue endlessly without noticing you're using different yardsticks.
Never escaping to the meta-level
Staying trapped inside a failing conversation instead of stepping up to talk about the talking wastes both people's energy.
Is it for you?
Best for
Anyone entering an emotionally loaded or high-stakes discussion at home, work, or in politics.
Not ideal for
Pure banter or casual chat where imposing structure would kill the spontaneity.
From the transcript
“that's rule one is agree what you're talking for”
“we run into problems when two people each have their own agenda in a conversation it's much easier when we both want the same thing”
“when you get stuck in one of the previous five conversations you need to be able to bail out of that conversation”
From the episode
An Expert's Guide To Mastering Difficult Conversations - Tim Harkness - #198
Tim Harkness