How to build deeper, more robust relationships | Carole Robin (Stanford professor, “Touchy Feely”)
Executive Summary
Carole Robin discusses gradual disclosure, feedback, inquiry and repair in working relationships. Selected third-party transcript coverage focuses on communication models and omits broader personal-development claims. Her examples are her accounts, unverified here. These are conversation tools, not guaranteed outcomes or mental-health guidance. Untimed editorial topics are not video chapters.
Key Takeaways
Stretch a little, not all at once
Robin's 15% rule means stepping modestly outside a conversational comfort zone to learn. It is not a numerical measure or instruction to reveal sensitive information. Notice how a particular person responds, because the same disclosure will not fit every relationship.
Separate your story from the other person’s experience
Her three-realities model distinguishes your internal experience, observable behavior and the other person's experience. You know your own side and what happened, but cannot directly know their motives. The shared behavior is a better starting point than claiming to know what they meant.
Use a feeling rather than an accusation
Robin explains that I feel you do not care still attributes a motive. Her feedback formula names the behavior, your feeling and why you are telling the person. It gives them information about the impact without turning your interpretation into a fact.
Move feedback into joint problem solving
Her example ends with a conversation about how two people can behave in a way that works for both. Constructive feedback is not a demand to change someone's identity. Ask what arrangement would help, and distinguish a request from a judgment about the person.
Ask to learn rather than confirm a verdict
Robin describes inquiry as searching without already knowing the answer. A question that smuggles in an accusation is not real curiosity. Suspend judgment long enough to ask what happened or how the other person sees the situation.
Repair by checking what was heard
When someone reacts unexpectedly, she suggests asking what they heard before arguing about what you meant. Her example then acknowledges why that interpretation produced the response and tries again. Understanding the impact does not require pretending your original words were something else.
Skills improve odds, not guarantee an outcome
Robin explicitly says the approach shifts probabilities rather than guarantees success. Even a well-intended conversation can go wrong, which makes repair part of the skill. Treat it as practice in a particular relationship, not a script that compels cooperation.
The Conversation Breakdown
Stretch a little, not all at once
⏱ Untimed topicRobin's 15% rule means stepping modestly outside a conversational comfort zone to learn. It is not a numerical measure or instruction to reveal sensitive information. Notice how a particular person responds, because the same disclosure will not fit every relationship.
Separate your story from the other person’s experience
⏱ Untimed topicHer three-realities model distinguishes your internal experience, observable behavior and the other person's experience. You know your own side and what happened, but cannot directly know their motives. The shared behavior is a better starting point than claiming to know what they meant.
Use a feeling rather than an accusation
⏱ Untimed topicRobin explains that I feel you do not care still attributes a motive. Her feedback formula names the behavior, your feeling and why you are telling the person. It gives them information about the impact without turning your interpretation into a fact.
Move feedback into joint problem solving
⏱ Untimed topicHer example ends with a conversation about how two people can behave in a way that works for both. Constructive feedback is not a demand to change someone's identity. Ask what arrangement would help, and distinguish a request from a judgment about the person.
Ask to learn rather than confirm a verdict
⏱ Untimed topicRobin describes inquiry as searching without already knowing the answer. A question that smuggles in an accusation is not real curiosity. Suspend judgment long enough to ask what happened or how the other person sees the situation.
Repair by checking what was heard
⏱ Untimed topicWhen someone reacts unexpectedly, she suggests asking what they heard before arguing about what you meant. Her example then acknowledges why that interpretation produced the response and tries again. Understanding the impact does not require pretending your original words were something else.
Skills improve odds, not guarantee an outcome
⏱ Untimed topicRobin explicitly says the approach shifts probabilities rather than guarantees success. Even a well-intended conversation can go wrong, which makes repair part of the skill. Treat it as practice in a particular relationship, not a script that compels cooperation.
Core Frameworks & Ideas
15% rule
Make a modest conversational stretch into a learning zone rather than a leap into a danger zone; the percentage is a metaphor.
Three realities and the net
Distinguish your internal experience, observable behavior and another person's experience. Avoid claiming access to their motives.
Behavior, feeling, purpose
Describe the behavior, name the feeling it produced and explain why you are sharing it, then move into problem solving.
Actionable Next Steps
- ✓Replace one claim about a motive with observed behavior and its impact on you.
- ✓Make one modest, appropriate conversational stretch without revealing sensitive details.
- ✓Ask an open question to learn rather than confirm your first explanation.
- ✓After an unexpected reaction, ask what the person heard before defending your intent.
Notable Quotes
“They're attributions and imputed motives, and that's where we make our biggest mistakes when it comes to feedback.”
“You can always go back to being judgmental, but you got to suspend it long enough to see if there's something for you to learn.”
“First of all, let's make sure that we point out that what we're doing here is we're shifting probabilities of success, we're not guaranteeing anything.”
The One Thing to Remember
Stay with the behavior and impact you know, then ask about the experience you do not know.