CFAR rationality technique for identifying where you've entangled questions that should be separate, causing you to flinch away from evidence. Use when the user: (1) feels defensive about acknowledging something true, (2) conflates unrelated beliefs ("Am I a good writer?" with "Did I spell this correctly?"), (3) has trouble accepting feedback, (4) wants to practice detecting question substitution, or (5) notices they avoid thinking about certain topics. Related to Kahneman's attribute substitution. Triggers: "bucket error", "conflating", "flinching away", "defensive about", "question substitution", "entangled beliefs", "CFAR", "can't accept feedback".
A CFAR technique for identifying where multiple concepts have been incorrectly lumped into one mental "bucket," causing defensive reactions and distorted thinking. The flinch away from evidence is the symptom; the misbucketing is the disease.
Your mind categorizes experiences into buckets: "Am I a good writer?", "Is this a good idea?", "Am I competent?" When unrelated evidence gets dropped into the same bucket, accepting one piece of evidence threatens to force updates across all beliefs in that bucket.
Classic example: Sally believes "I am a good writer." When she makes a spelling mistake, she flinches away from the evidence — because her mind has bucketed spelling ability together with overall writing quality. Admitting the mistake threatens the whole bucket.
The fix: Separate the buckets. "I am a good writer" and "I sometimes misspell words" can both be true simultaneously.
Signs of a bucket error:
Ask: "What would be bad about accepting this information?" Then: "Does that consequence necessarily follow, or have I bucketed things together?"
Name the distinct questions that have been conflated:
Route evidence to the appropriate bucket only. A spelling error informs the spelling bucket, not the writing-quality bucket.
"Does X actually imply Y? Or only weakly? Can I think of cases where X is true but Y is false?"
Detection: "Is there something you resist acknowledging even though it might be true?" / "What would be bad about accepting this?"
Untangling: "What are the separate questions here?" / "Does [evidence] really tell you about [broader belief], or is it about something more specific?"
Testing implications: "Can you think of a case where X is true but Y is false?" / "How strongly does X actually predict Y?"
Reframing: "What if these were separate questions with separate answers?" / "Can both be true at the same time?"
Exercise 1: Flinch Inventory
Exercise 2: Question Substitution Detection
Exercise 3: Implication Audit