Epistemic Humility Protocol
Replace false certainty with transparent, structured inquiry
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 95%
The protocol turns uncertainty from a perceived weakness into a disciplined reasoning process. Begin by stating what you do not know or where your confidence is limited. Then define the problem, disclose whether it falls inside your expertise, and lay out several competing ways to interpret the available evidence. Work through those possibilities rather than asserting a polished answer prematurely. The output is not indecision; it is a calibrated conclusion whose assumptions and limits are visible to others. This approach improves trust because audiences can distinguish evidence from speculation, while the speaker remains free to revise the conclusion as new information arrives. It is particularly useful in media, research, management, and other environments where social pressure rewards confidence more strongly than accuracy.
Origin
Extracted from Masters in Business during Rex Sorgatz's discussion of writing, media certainty, and his research into misinformation.
Core principles
- 01Admit uncertainty before making unsupported claims
- 02Separate established knowledge from tentative ideas
- 03Consider opposing explanations before settling on one
- 04Decline to speculate outside your expertise
How to run it
- 1
Declare uncertainty
Open by identifying what you do not know or where your confidence is weak. This prevents a tentative judgment from sounding like an established fact.
Pro tip Use direct language such as "I don't know" rather than burying uncertainty in qualifications.
Watch out Do not use uncertainty as an excuse to avoid investigating the question.
- 2
Define the problem
Restate the question in a form that can be examined. Separate the factual issue from the emotional or political pressure surrounding it.
Pro tip Name which part of the question is answerable and which part remains ambiguous.
Watch out A poorly framed question can make every subsequent answer misleading.
- 3
Check your competence boundary
Determine whether the subject falls within your expertise and say so openly. If it does not, limit your contribution or defer to a qualified source.
Pro tip Distinguish direct knowledge from an informed guess.
Watch out Familiarity with a topic is not the same as expertise.
- 4
Compare opposing interpretations
Lay out multiple plausible explanations and test each against the available evidence. Make the reasoning process visible instead of presenting only the conclusion.
Pro tip Include the strongest credible alternative, not a deliberately weak counterargument.
Watch out Do not create false balance between evidence-based and baseless positions.
- 5
Conclude provisionally
State the best-supported current judgment with an honest confidence level. Identify what evidence could cause you to revise it.
Pro tip Pair the conclusion with its most important assumption.
Watch out Do not let provisional language conceal a conclusion that the evidence supports strongly.
In the wild
A television guest is asked about a policy outside her specialty. Instead of improvising a categorical answer, she states that the implementation details are outside her expertise, identifies two plausible economic effects, and explains what data would distinguish them.
→ The audience receives a useful analytical map without mistaking speculation for expert knowledge.
Common mistakes
Performing uncertainty without analysis
Merely repeating "I don't know" adds no value. The protocol requires a good-faith attempt to frame and investigate the problem when possible.
Treating every claim as equally uncertain
Calibrated doubt should reflect the evidence; well-established facts should not be presented as arbitrary opinions.
Is it for you?
Best for
It is best for researchers, leaders, writers, and commentators addressing ambiguous or unfamiliar questions.
Not ideal for
It is not ideal when an immediate emergency requires decisive action based on already-established procedures.
From the transcript
“I always start with I don't know or I am not sure. Let me try to work this problem out. I've got a few ideas.…”
“oh, that's outside of my expertise. I can I really have nothing to add on that.”
“I don't trust writing that asserts itself in some sort of like overly um unbiased intellectual way that just so firm and has no sense…”
From the episode
Rex Sorgatz Discusses the Mechanics of Misinformation (Podcast)
Rex Sorgatz