Assume Unknown

Input: $ARGUMENTS

Core Move

Take something treated as settled, known, or obvious and assume it’s actually unknown. We don’t know this. Maybe we can’t know it. What changes?

Powerful for exposing false certainty and finding where confidence outpaces evidence.

Procedure

Step 1: State What’s “Known”

What are we assuming is actually unknown?

Step 2: Force the Assumption

“We don’t actually know this. Our confidence is unjustified. This is an open question.”

Step 3: Trace Implications

If we don’t actually know this:

1. What becomes uncertain? — What downstream beliefs, plans, or decisions lose their foundation?
2. What should we stop asserting? — What claims can we no longer make?
3. What investigation is needed? — If truly unknown, what would it take to actually know?
4. What have we been doing wrong? — What actions were based on false certainty?
5. Is it unknowable? — Unknown-but-findable vs. fundamentally unknowable. Which is it?
6. How do we act under this uncertainty? — What’s the right strategy when this is genuinely unknown?

Step 4: Test the Assumption

• What’s the actual evidence base for the “known” claim?
• Could the evidence be wrong? How robust is it?
• Are we confusing familiarity with knowledge?
• Are we confusing consensus with truth?

Step 5: Synthesize

CLAIM: [what was assumed known]
ASSUMING UNKNOWN:
  Becomes uncertain: [downstream effects]
  Investigation needed: [what it would take to know]
  Unknowable?: [yes/no — if yes, strategy for acting anyway]
ACTUAL KNOWLEDGE STATUS: [known/uncertain/unknown/unknowable]
NEXT MOVE: [investigate / act under uncertainty / redesign to not depend on it]

When to Use

• Suspect false certainty
• Building on “facts” that might not be facts
• Need to identify what you’d need to actually establish knowledge

Integration

• Pair with /aknown for the opposite stance
• Follow with /nsa for decision-making under uncertainty
• Use /aex to surface other potentially false “knowns”