Surface and examine the assumptions you are making
Name the assumptions filling the gap between the data you selected and the interpretation you reached.
Why it works
Assumptions are the invisible links in the inferential chain — the beliefs you did not consciously choose but that are required for the interpretation to follow from the data. They are often so obvious to the holder that they are not recognized as assumptions at all. Making them explicit converts them from invisible architecture to examinable hypotheses, which is the minimum condition for checking whether they are warranted.
How to do it
- Ask: "What would have to be true about people (or the world) for my interpretation to follow from the data?"
- Write those statements as "I am assuming that…"
- Check each assumption: is it based on evidence, or is it a standing belief imported from elsewhere?
- Invite a person with a different background to name what assumptions they think you are making.
Evidence
Argyris’s own research on organizational double-loop learning documents how unexamined assumptions sustain defensive routines that prevent effective problem-solving — directly motivating the practice of surfacing assumptions. The general principle (implicit beliefs drive behavior and must be surfaced to change it) is foundational in cognitive therapy as well. (clinical)
The research is primarily organizational and case-study-based; RCT evidence on the practice of surfacing assumptions as an individual cognitive intervention is limited.
Sources
- Argyris & Schön (1978), Organizational Learning: A Theory of Action Perspective
Common mistake
Identifying your assumptions in words that are so abstract they cannot be checked ("I assume people have good intentions") rather than specific enough to be examined ("I assume she knew about the meeting").
Practice this with IX Coach
7 days free, then $40/month (~$1.30/day).
More practices for The Ladder of Inference
- Distinguish observable data from interpretation
Separate what a camera would record from what you made of it.
- Notice which data you selected — and which you ignored
Your conclusions are built on a sample of the available data — ask what the sample excluded.
- Test conclusions before acting on them
Treat your conclusion as a hypothesis and find one piece of evidence that would confirm or disconfirm it.
- Walk down the ladder together in disagreements
In conflict, stop arguing about conclusions and descend together to the data level.
- Slow the reflexive loop
Your current beliefs shape which data you select, reinforcing themselves — interrupt this loop deliberately.
Related concepts
- Thinking, Fast and Slow, Made Usable
Two systems, the biases they create, and when to slow down
- Hanlon's Razor: Never Attribute to Malice What Stupidity Can Explain
The charitable-attribution heuristic and why it usually serves you better
- The Map Is Not the Territory
Mental models are tools, not truths — how to use them without being captured by them
- Cognitive Reframing, Made Practical
How reappraisal changes emotion, the common distortions, and how to do it without toxic positivity