Reframing primitives change the angle from which a problem is viewed without changing the problem itself. They are the fastest way to escape a cognitive rut.
Trigger: When stuck, when planning, when evaluating risk
Input: A goal or desired outcome
Operation: Ask “how would I guarantee this fails?” then negate the answers
Output: Hidden risks, necessary conditions, blind spots
Inversion is one of the most universally applicable primitives. Its power is that most people approach problems by thinking about what they should do to succeed — a mode that activates motivated reasoning and confirmation bias. Inversion switches the question entirely.
Rather than “how do I build a successful product?” you ask “what are all the ways this product certainly fails?” This forces an honest inventory of failure modes that optimistic forward thinking systematically misses.
The mathematician Carl Jacobi made “invert, always invert” a professional maxim. Charlie Munger applied it to investing with results documented over decades.
Known limits: Less useful in genuinely novel domains with no precedent, where failure modes are themselves unknown.
Trigger: When stuck on tactical details; when a solution feels forced
Input: Any problem statement
Operation: Zoom out to purpose (“why are we doing this?”) or zoom in to mechanism (“what exactly happens step by step?”)
Output: Either a redefined problem or a more granular understanding
Most problems get stuck at a single level of abstraction. A team arguing about button colors may be having a conversation that should be happening at the level of brand identity. An executive debating market strategy may need to zoom in to the specific customer interaction that is failing.
Level shifting breaks abstraction lock (the tendency to keep working at the same altitude even when the answer is at a different one).
Protocol: When stuck, first ask “what is this actually in service of?” (zoom out), then ask “what precisely happens here, step by step?” (zoom in). Often the answer becomes obvious at one of these levels.
Known limits: Can be used to avoid commitment (i.e., endless zooming in or out without ever acting).
Trigger: Emotional decisions; decisions with long-horizon consequences
Input: A decision or situation under evaluation
Operation: Ask: how will I feel about this in 10 minutes / 10 months / 10 years?
Output: Temporal perspective; reveals short-termism and long-term regret
Human beings systematically overweight the present moment — hyperbolic discounting. The 10-minute window surfaces immediate emotional reactions without allowing them to dominate. The 10-month window asks whether this will still matter by next year. The 10-year window, the most important, asks whether this will register at all against the background of a life.
The 10-year frame radically deflates the apparent importance of most immediate stressors, and equally flags genuine long-term consequences that were being discounted.
Known limits: Can induce false detachment from situations that genuinely do matter in the short term.
Trigger: Multi-stakeholder situations; conflict; negotiation
Input: A situation with multiple parties
Operation: Systematically view the situation from each stakeholder’s perspective as completely as possible
Output: A multi-perspective map; reveals hidden interests and constraints
The default cognitive mode is egocentric. Role shifting is the deliberate discipline of inhabiting another perspective fully enough that you could articulate their interests, fears, and constraints better than they might themselves.
This is not mere empathy (feeling what they feel) but perspective-taking: understanding what they believe, need, and stand to gain or lose. In negotiation, the party with the best model of all parties’ interests consistently outperforms.
Test: Could a sophisticated version of the other party read your role-shift and say “yes, that’s an accurate representation of my position”?
Known limits: Difficult under high emotional activation; requires genuine effort not to collapse back into your own frame.
Trigger: When a problem is handed to you rather than defined by you
Input: A presented problem statement
Operation: Ask: “Is this the right problem, or the presented problem?”
Output: Either confirmation or a redefined problem at higher accuracy
One of the most costly reasoning errors is solving the wrong problem brilliantly. Problems are usually framed by whoever identifies them, and that framing embeds assumptions about causes, constraints, and solutions that may be wrong.
Classic example: “we need a faster horse” vs. “we need faster transportation.”
Category questioning requires the discipline to pause before solution-generation begins, the moment when most people are already committed to a problem frame.
Known limits: Can become an excuse to never engage with the stated problem; must be applied quickly and decisively.