Thinking Tools1.3 Framing the SituationAlso known as Ladder of Inference with Five Whys
Combines Chris Argyris's ladder of inference (Overcoming Organizational Defenses, 1990), as set out in The Fifth Discipline Fieldbook (Peter M. Senge, Art Kleiner, Charlotte Roberts, Richard B. Ross, and Bryan J. Smith, 1994), with the Five Whys of Sakichi Toyoda and Taiichi Ohno (Toyota Production System, 1978).
Walks a conclusion back down the ladder of inference and asks why at each rung, producing a list of the assumptions a position rests on.
One conclusion about a new manager, taken back down the ladder it climbed
A director decided his new regional manager was disengaged. Each tag is a step of his reasoning. Place it on its rung, from what anyone could have observed up to the conclusion, and for each assumption ask whether the conclusion survives if it is false. The first step is placed as an example. Illustrative.
Items (select one, then a box)
Not placedPlacedThe conclusionLoad-bearing: test it firstMinor: note itMeaning addedObserved
Not placedPlacedThe conclusionLoad-bearing: test it firstMinor: note itMeaning addedObserved
Not placedPlacedThe conclusionLoad-bearing: test it firstMinor: note itMeaning addedObserved
Not placedPlacedThe conclusionLoad-bearing: test it firstMinor: note itMeaning addedObserved
Not placedPlacedThe conclusionLoad-bearing: test it firstMinor: note itMeaning addedObserved
Not placedPlacedThe conclusionLoad-bearing: test it firstMinor: note itMeaning addedObserved
Not placedPlacedThe conclusionLoad-bearing: test it firstMinor: note itMeaning addedObserved
Not placedPlacedThe conclusionLoad-bearing: test it firstMinor: note itMeaning addedObserved
Not placedPlacedThe conclusionLoad-bearing: test it firstMinor: note itMeaning addedObserved
Rung of the ladder
ConcludedAssumedMeaning addedObserved
The conclusion
Minor: note it
Load-bearing: test it first
Meaning added
Observed
The conclusion standsThe conclusion falls
If this proved false
Camera off, 3 weeks. She joined the weekly call with her camera off three weeks running.Not placed yet.Camera off, 3 weeks. She joined the weekly call with her camera off three weeks running.The conclusion: The top of the ladder: what he now believes and is about to act on. It is only as good as the rungs under it.Camera off, 3 weeks. She joined the weekly call with her camera off three weeks running.Load-bearing: test it first: An assumption the conclusion cannot survive without. Check it with the person or the record before acting.Camera off, 3 weeks. She joined the weekly call with her camera off three weeks running.Minor: note it: An assumption, but the conclusion stands even if it is wrong. Write it down and move on.Camera off, 3 weeks. She joined the weekly call with her camera off three weeks running.Meaning added: A reading of the data. Another reading fits the same facts, so ask why this one was picked.Camera off, 3 weeks. She joined the weekly call with her camera off three weeks running.Observed: Data a camera would have recorded. Ask also what else was there that did not make the list.A fact anyone on the call could confirm. What it means is a separate step.No budget comments. She sent nothing back on the draft budget before the deadline.Not placed yet.No budget comments. She sent nothing back on the draft budget before the deadline.The conclusion: The top of the ladder: what he now believes and is about to act on. It is only as good as the rungs under it.No budget comments. She sent nothing back on the draft budget before the deadline.Load-bearing: test it first: An assumption the conclusion cannot survive without. Check it with the person or the record before acting.No budget comments. She sent nothing back on the draft budget before the deadline.Minor: note it: An assumption, but the conclusion stands even if it is wrong. Write it down and move on.No budget comments. She sent nothing back on the draft budget before the deadline.Meaning added: A reading of the data. Another reading fits the same facts, so ask why this one was picked.No budget comments. She sent nothing back on the draft budget before the deadline.Observed: Data a camera would have recorded. Ask also what else was there that did not make the list.Observed, and checkable in the shared folder.Filled 4 open roles. In her first month she closed four hires the region had carried for a quarter.Not placed yet.Filled 4 open roles. In her first month she closed four hires the region had carried for a quarter.The conclusion: The top of the ladder: what he now believes and is about to act on. It is only as good as the rungs under it.Filled 4 open roles. In her first month she closed four hires the region had carried for a quarter.Load-bearing: test it first: An assumption the conclusion cannot survive without. Check it with the person or the record before acting.Filled 4 open roles. In her first month she closed four hires the region had carried for a quarter.Minor: note it: An assumption, but the conclusion stands even if it is wrong. Write it down and move on.Filled 4 open roles. In her first month she closed four hires the region had carried for a quarter.Meaning added: A reading of the data. Another reading fits the same facts, so ask why this one was picked.Filled 4 open roles. In her first month she closed four hires the region had carried for a quarter.Observed: Data a camera would have recorded. Ask also what else was there that did not make the list.Observed, and left out: it never made his list. The audit is the first time anyone set it beside the rest.Camera off: tuned out. He read the dark camera as a sign she was not paying attention.Not placed yet.Camera off: tuned out. He read the dark camera as a sign she was not paying attention.The conclusion: The top of the ladder: what he now believes and is about to act on. It is only as good as the rungs under it.Camera off: tuned out. He read the dark camera as a sign she was not paying attention.Load-bearing: test it first: An assumption the conclusion cannot survive without. Check it with the person or the record before acting.Camera off: tuned out. He read the dark camera as a sign she was not paying attention.Minor: note it: An assumption, but the conclusion stands even if it is wrong. Write it down and move on.Camera off: tuned out. He read the dark camera as a sign she was not paying attention.Meaning added: A reading of the data. Another reading fits the same facts, so ask why this one was picked.One reading. She could be on a shop floor, on a weak connection, or taking notes.Camera off: tuned out. He read the dark camera as a sign she was not paying attention.Observed: Data a camera would have recorded. Ask also what else was there that did not make the list.Silence: did not read. He read her silence on the budget as not having read it.Not placed yet.Silence: did not read. He read her silence on the budget as not having read it.The conclusion: The top of the ladder: what he now believes and is about to act on. It is only as good as the rungs under it.Silence: did not read. He read her silence on the budget as not having read it.Load-bearing: test it first: An assumption the conclusion cannot survive without. Check it with the person or the record before acting.Silence: did not read. He read her silence on the budget as not having read it.Minor: note it: An assumption, but the conclusion stands even if it is wrong. Write it down and move on.Silence: did not read. He read her silence on the budget as not having read it.Meaning added: A reading of the data. Another reading fits the same facts, so ask why this one was picked.One reading. She may have had no objections, or no time.Silence: did not read. He read her silence on the budget as not having read it.Observed: Data a camera would have recorded. Ask also what else was there that did not make the list.Good hires speak up. He assumed an engaged manager speaks up in her first month.Not placed yet.Good hires speak up. He assumed an engaged manager speaks up in her first month.The conclusion: The top of the ladder: what he now believes and is about to act on. It is only as good as the rungs under it.Good hires speak up. He assumed an engaged manager speaks up in her first month.Load-bearing: test it first: An assumption the conclusion cannot survive without. Check it with the person or the record before acting.If false, and some good managers watch first, the whole conclusion goes.Good hires speak up. He assumed an engaged manager speaks up in her first month.Minor: note it: An assumption, but the conclusion stands even if it is wrong. Write it down and move on.Good hires speak up. He assumed an engaged manager speaks up in her first month.Meaning added: A reading of the data. Another reading fits the same facts, so ask why this one was picked.Good hires speak up. He assumed an engaged manager speaks up in her first month.Observed: Data a camera would have recorded. Ask also what else was there that did not make the list.Same load as others. He assumed she carries the same work as the other managers.Not placed yet.Same load as others. He assumed she carries the same work as the other managers.The conclusion: The top of the ladder: what he now believes and is about to act on. It is only as good as the rungs under it.Same load as others. He assumed she carries the same work as the other managers.Load-bearing: test it first: An assumption the conclusion cannot survive without. Check it with the person or the record before acting.Load-bearing, and false: she has covered a second site since a resignation, and dials in from its floor.Same load as others. He assumed she carries the same work as the other managers.Minor: note it: An assumption, but the conclusion stands even if it is wrong. Write it down and move on.Same load as others. He assumed she carries the same work as the other managers.Meaning added: A reading of the data. Another reading fits the same facts, so ask why this one was picked.Same load as others. He assumed she carries the same work as the other managers.Observed: Data a camera would have recorded. Ask also what else was there that did not make the list.The team noticed too. He assumed her team had seen the same signs.Not placed yet.The team noticed too. He assumed her team had seen the same signs.The conclusion: The top of the ladder: what he now believes and is about to act on. It is only as good as the rungs under it.The team noticed too. He assumed her team had seen the same signs.Load-bearing: test it first: An assumption the conclusion cannot survive without. Check it with the person or the record before acting.The team noticed too. He assumed her team had seen the same signs.Minor: note it: An assumption, but the conclusion stands even if it is wrong. Write it down and move on.Whether the team noticed changes how he talks to her, not whether the conclusion is true.The team noticed too. He assumed her team had seen the same signs.Meaning added: A reading of the data. Another reading fits the same facts, so ask why this one was picked.The team noticed too. He assumed her team had seen the same signs.Observed: Data a camera would have recorded. Ask also what else was there that did not make the list.Will not last a year. His conclusion: she is disengaged and will not last the year.Not placed yet.Will not last a year. His conclusion: she is disengaged and will not last the year.The conclusion: The top of the ladder: what he now believes and is about to act on. It is only as good as the rungs under it.With one load-bearing assumption false and a fact left out, the conclusion does not hold. He asks her about the second site instead.Will not last a year. His conclusion: she is disengaged and will not last the year.Load-bearing: test it first: An assumption the conclusion cannot survive without. Check it with the person or the record before acting.Will not last a year. His conclusion: she is disengaged and will not last the year.Minor: note it: An assumption, but the conclusion stands even if it is wrong. Write it down and move on.Will not last a year. His conclusion: she is disengaged and will not last the year.Meaning added: A reading of the data. Another reading fits the same facts, so ask why this one was picked.Will not last a year. His conclusion: she is disengaged and will not last the year.Observed: Data a camera would have recorded. Ask also what else was there that did not make the list.
Your sort matched 0 of 9.Your sort matched 1 of 9.Your sort matched 2 of 9.Your sort matched 3 of 9.Your sort matched 4 of 9.Your sort matched 5 of 9.Your sort matched 6 of 9.Your sort matched 7 of 9.Your sort matched 8 of 9.Your sort matched 9 of 9.
The conclusion
The top of the ladder: what he now believes and is about to act on. It is only as good as the rungs under it.
Load-bearing: test it first
An assumption the conclusion cannot survive without. Check it with the person or the record before acting.
Minor: note it
An assumption, but the conclusion stands even if it is wrong. Write it down and move on.
Meaning added
A reading of the data. Another reading fits the same facts, so ask why this one was picked.
Observed
Data a camera would have recorded. Ask also what else was there that did not make the list.
The sort
Camera off, 3 weeks: Observed. A fact anyone on the call could confirm. What it means is a separate step.
No budget comments: Observed. Observed, and checkable in the shared folder.
Filled 4 open roles: Observed. Observed, and left out: it never made his list. The audit is the first time anyone set it beside the rest.
Camera off: tuned out: Meaning added. One reading. She could be on a shop floor, on a weak connection, or taking notes.
Silence: did not read: Meaning added. One reading. She may have had no objections, or no time.
Good hires speak up: Load-bearing: test it first. If false, and some good managers watch first, the whole conclusion goes.
Same load as others: Load-bearing: test it first. Load-bearing, and false: she has covered a second site since a resignation, and dials in from its floor.
The team noticed too: Minor: note it. Whether the team noticed changes how he talks to her, not whether the conclusion is true.
Will not last a year: The conclusion. With one load-bearing assumption false and a fact left out, the conclusion does not hold. He asks her about the second site instead.
An example
Six weeks after hiring a regional manager, a sales director told HR he had made a mistake: she was disengaged and would not last the year. Before anyone started a performance plan, the HR partner asked him to walk the conclusion back down and say why at each step.
The facts at the bottom were thin. Her camera had been off on three Monday calls, and she had sent no comments on the draft budget. Everything above that was reading: a dark camera meant she was not listening, silence meant she had not read the draft. Two assumptions carried the weight. He took for granted that an engaged manager speaks up in her first month, and that her workload matched the other managers'. The second was false. Since a resignation she had been covering a second site, and she dialed in from its floor. A fact he had never put on his list also surfaced: she had filled four roles the region had carried open for a quarter. The plan became a conversation about the second site.
Why it works
People reach conclusions fast, by a climb they do not notice. From everything there is to see, they pick a few facts, give those facts a meaning, fill the gaps with assumptions, and arrive at a belief that then decides which facts they notice next. Only the bottom and the top are visible. The middle rungs, where most errors sit, are skipped in conversation, so two people argue about conclusions when they really differ about an assumption.
Walking the ladder down makes the middle rungs visible, and asking why at each one stops the walk from ending at the first comfortable answer. Sorting the assumptions by whether the conclusion survives without them turns a vague doubt into a short list of things to check.
The thinking behind it
Chris Argyris Chris Argyris, a Harvard scholar of how organizations learn, is credited with the ladder of inference, set out in Overcoming Organizational Defenses (1990): the steps by which people climb from selected data to meanings, assumptions, conclusions, and beliefs, usually without noticing the climb. His wider work with Donald Schön on double-loop learning asked people to question the governing assumptions behind what they do, which is what this audit does with one conclusion.
Richard B. Ross Richard B. Ross wrote the chapter of The Fifth Discipline Fieldbook (1994), with Peter M. Senge, Art Kleiner, Charlotte Roberts, and Bryan J. Smith, that turned the ladder into a practical tool. It names the rungs, from observable data through selected data, meanings, assumptions, conclusions, and beliefs to actions, and shows how to walk them aloud in a meeting so others can test each step.
Taiichi Ohno Taiichi Ohno, the Toyota executive behind the Toyota Production System, wrote in his book of that name (1978) that asking why five times was the basis of Toyota's scientific approach, a habit he traced to Sakichi Toyoda. The audit borrows the repeated why and points it at reasoning instead of a machine: each rung is asked why until the answer can no longer be justified.
How to run it
Write the conclusion in one sentence. Use the words the person would use in the hallway, not the polished version for a memo.
Find the facts at the bottom. List only what a camera or a record would show, with dates, and nothing about intent.
Ask what else was there. Look for facts that were available and left out, since the selection is itself a step.
Separate meanings from facts. For each fact, write the meaning that was given to it, and one other meaning that fits the same fact.
Name the assumptions that bridge the gaps. Ask why at each rung until the answer is a general belief about people, the business, or the situation.
Sort the assumptions by weight. Mark each one the conclusion cannot survive if it is false; those are the ones to test.
Test before acting. Check each load-bearing assumption with the person involved or the record, then decide whether the conclusion still stands.
Where it goes wrong
The audit becomes a cross-examination. Run on someone else's conclusion in front of others, it reads as an accusation, and the person defends instead of looking. Run it first on a conclusion of your own, or in private, and ask permission before walking someone else's.
Every assumption gets equal attention. A list of fifteen assumptions with no weights produces paralysis. Push for the two or three that carry the conclusion, and stop.
The bottom rung is not really data. Statements like she seemed checked out sit at the bottom of the list as if observed. If a camera could not have recorded it, move it up a rung.
When to use something else
Five Whys. When the chain to trace is a run of causes in a process, not a run of inferences in someone's head.
Double-Loop Learning. When the assumption in question governs how a whole team works, and the goal is to change the rule, not one decision.
Premortem. When the conclusion is a plan, and the useful question is how it would fail.