Collaborative Empiricism
Diagnosis, vision, strategy, smallest step — and the person has to choose every one of them
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 94%
Peterson names the behavioural clinician's method explicitly: collaborative empiricism. The person lays out every card they have about what is wrong, you discard roughly 90% as symptoms, then you ask what better would look like and they lay out their cards again. That gives a problem statement and a hypothetical cure; the gap between them is the strategy, and you break the strategy down experimentally until you find a step the person will actually take. The non-negotiable condition is voluntariness — the person must own the diagnosis, the vision, and the strategy, because advice they merely act out is not theirs.
Origin
Peterson's standard clinical negotiation with clients, described as what you do 'if you're sensible', and demonstrated on the client who came to him saying he had mucked up his life and did not want to destroy his unborn son.
Core principles
- 01The person must discover the problem themselves; you cannot enforce a diagnosis.
- 02Around 90% of what someone first lays out is symptom, not the central issue.
- 03Diagnosis, cure, and strategy are three separate negotiated artefacts.
- 04Solving someone's problem for them steals the problem from them.
- 05The strategy is only valid if the person feels it is right for them — voluntariness is 100% crucial.
How to run it
- 1
Let them put all the cards on the table
Listen at length while the person unfolds everything they think might be wrong, without interrupting or sorting as they go.
Pro tip You have to listen and listen and listen while the whole set comes out — the ordering happens afterwards.
Watch out Cutting this short leaves you diagnosing from the screen concerns rather than the real one.
- 2
Discard the symptoms
Sort the cards and get rid of roughly 90% — the items that resolve just by being spoken, or that the person no longer finds interesting or central.
Pro tip Watch for the ones the person themselves dismisses on hearing out loud; those were symptoms.
- 3
Get the problem statement
Converge with the person on a single specified problem. This is the diagnosis and it is the hardest move in the sequence.
Pro tip Once the problem is specified, roughly 95% of the problem is solved.
Watch out Do not assume you or they already know what the problem is at the start.
- 4
Ask what better would look like
Have them lay out their cards again, this time about the desired end state, producing a hypothetical cure they actually want.
Pro tip This is described as a great general problem-solving strategy, not just a clinical one.
- 5
Derive the strategy and find the first executable step
The steps between the problem and the destination are the strategy; break them down experimentally until you locate the one the person will actually take.
Pro tip Set a single agreed action, a return date, and a report-back — 'go implement it, come back next week'.
Watch out If the person is not genuinely on board with the strategy, it will not be executed, whatever they say.
In the wild
A thirty-year-old still living at home under his mother's thumb got a girlfriend pregnant and came to Peterson saying he had mucked up his life but did not want to destroy the kid. They negotiated the problem, the desired state, and a strategy that bottomed out at vacuuming his bedroom carpet.
→ He failed the first task, they renegotiated a smaller one, and the collaborative process continued rather than collapsing.
Bartlett spent thirty minutes only asking questions of a colleague who said she was unhappy in her role and gave a list of reasons. He kept asking rather than proposing fixes, including when she had been happiest in the business.
→ She concluded that none of her stated reasons were the issue; the real problem was distance from him and from meaningful work, and moving her back toward the hard problems fixed it.
Common mistakes
Giving good advice too early
An experienced person can simply hand over the answer, but then the client acts out someone else's conclusion. It is not theirs, so it does not hold.
Taking on the other person's problem
If you treat their problems as your problems, you steal the problem from them and remove the capability-building that solving it would have produced.
Skipping the vision step
Moving from diagnosis straight to tactics leaves the destination unspecified, so no coherent strategy can be derived and the steps have nothing to point at.
Is it for you?
Best for
Managers, coaches, therapists, and partners trying to help someone change without doing the changing for them.
Not ideal for
Acute emergencies requiring immediate directive action, or people who genuinely want an instruction rather than a process.
From the transcript
“this is called technically this is called collaborative empiricism it's a behavioral approach for clinicians”
“then you break down the steps until you find the step that they that the person will take”
“it's absolutely 100% crucial that it's voluntary”
“first of all they're not your problems. You have to understand that.”
From the episode
Doctor Jordan Peterson: "The NUMBER ONE Reason For Divorce!" & "The One Small Step You Have To Take To Turn Your Whole Life Around!"
Jordan Peterson