Two people can sit through the same meeting and walk out with completely different conclusions — not because one of them is lying, but because they climbed different ladders to get there.

What the Ladder Actually Is

The ladder of inference was developed by the organisational psychologist Chris Argyris in 1970 and popularised by Peter Senge in The Fifth Discipline (1994). It maps the almost-instant mental journey from raw observation to action, and it has seven rungs:

1. Observable data — everything that actually happened, available to anyone in the room.
2. Selected data — the slice you actually noticed, filtered by what you already expect to see.
3. Added meaning — the interpretation you layer on top of that slice.
4. Assumptions — the things you take as given, built on that meaning.
5. Conclusions — the judgement you form from the assumptions.
6. Beliefs — the conclusion, now generalised into “how things are.”
7. Action — what you actually do about it.

Climb it once and it feels like reasoning. Climb it a hundred times and it feels like knowing. That’s the trap.

Climb it once and it feels like reasoning. Climb it a hundred times and it feels like knowing.

Why It Runs So Fast You Don’t Notice

None of this happens in a considered, step-by-step way. It happens in under a second, and it happens on a loop. Your beliefs shape which data you select next time, which reinforces the same conclusions, which hardens the same beliefs. Argyris called this the reflexive loop. In practice, it means the ladder isn’t a one-off climb. It’s the staircase you live on, and you stop noticing the steps.

The uncomfortable part isn’t that people jump to conclusions. It’s that the jump feels like the floor.

Where It Shows Up in Delivery and Leadership

You don’t need a psychology lab to see this. It’s in every status report, every incident review, every stakeholder update. I’ve sat through enough of these across four different platform shifts to know the pattern doesn’t change; only the technology does.

• A RAG status goes red. One person selects “the vendor missed a date” as the relevant data; another selects “the vendor flagged the risk three weeks ago.” Same board, different ladder.
• A postmortem lands on “the engineer should have known better,” a belief several rungs up, when the observable data was a config change reviewed by two other people first.
• An AI-generated meeting summary quietly does rungs two and three for you, selecting and interpreting, before you’ve read a word of the transcript. You inherit someone else’s ladder and don’t know it.
• A stakeholder says “the team isn’t engaged.” Pressed for the observable data, it’s usually one quiet person in one meeting.

None of these people are acting in bad faith. They’re doing what brains do: climbing fast and calling the top rung a fact.

None of these people are acting in bad faith. They’re doing what brains do: climbing fast and calling the top rung a fact.

Climbing Back Down

The fix isn’t to stop inferring. You can’t run an organisation on raw data alone. The fix is to make the climb visible, especially your own.

1. Name the rung you’re on. Before you defend a conclusion, ask whether you’re defending the data or the story you built on it.
2. Separate the data from the story. Write the observable fact in one column and your interpretation in the next. If you can’t fill the first column, you don’t have a fact yet.
3. Say the assumption out loud. “I’m assuming the delay means they’re deprioritising us” invites correction. A silent assumption doesn’t.
4. Ask what would change your mind. If no answer exists, you’re not holding a conclusion — you’re holding a belief, and beliefs don’t respond to evidence the same way conclusions do.
5. Show your working. Senge’s “left-hand column” exercise, writing what you’re actually thinking next to what you actually said, is uncomfortable precisely because it exposes how much of the ladder never gets spoken.

The Questions Worth Keeping

A short set, worth having ready in any review where the room is about to agree too quickly:

• What did we actually observe, versus what did we decide it meant?
• Whose data got selected, and whose got left out?
• What assumption is doing the heavy lifting here?
• Is this a conclusion, or has it already hardened into a belief?
• What evidence would change this view, and have we gone looking for it?

It’s rarely a disagreement about facts. It’s two people standing on different rungs, each certain they’re on the ground floor.

The next time a disagreement feels intractable, it’s rarely a disagreement about facts. It’s two people standing on different rungs, each certain they’re on the ground floor.

Sources used

• Argyris, C. (1970). Intervention Theory and Method.
• Senge, P. (1994). The Fifth Discipline: The Art & Practice of the Learning Organization.
• The Ladder of Inference — MindTools
• How to Use the Ladder of Inference for Better Decisions — Asana
• The Ladder of Inference re-visited: “Don’t jump to confusions!” — PMC