Holding two at once

Give two halves of the group the same situation and two different ways of thinking. Put the results side by side. Where they agree, confidence goes up. Where they disagree, you have found the part nobody understands yet, and that is where the attention should go.

Ways of thinking: the four moves

Move 3 of 4 · Finds where the frames disagree

Run two frames on one problem and study where they split.

Time: 75 min
People: 6–14, split into two groups of 3–7
You need: Two ends of a room, flip chart paper, two colours of pen, the side-by-side grid, the open questions sheet
You leave with: A list of findings both frames support, and a short list of open questions where they split, each with an owner and a way to find out

Needs a confident facilitator

The idea

What the move is, and why it works

What the move is

Holding two at once means running two ways of thinking on the same situation, properly and at the same time, and then comparing what they found. Switching tries four frames for a few minutes each and keeps one. Holding two gives two frames twenty-five minutes each and keeps both, to see where they part company.

The output is a map of the decision’s weak points: the claims both frames support, which you can treat as sound, and the places they split, which become questions to investigate. No decision is taken yet.

The idea behind it

Roger Martin studied leaders who were good at hard decisions and found that many of them could hold two opposed models in mind without rushing to pick one. He called it integrative thinking. The tension between the models was the useful part, because it showed where each was missing something.

Michael Jackson, and Mingers and Brocklesby, make the same case for analysis. Any single method sees part of a situation. Two methods chosen to differ catch each other’s errors, and agreement between them means more than either alone.

Why it works

A neat analysis from one frame hides its own assumptions. Put it beside an analysis from another frame and the assumptions show up as points of disagreement, which are far easier to see than an assumption on its own.

Choose the pair with care. Frames on opposite diagonals of the matrix, such as systems and computational or design and critical, disagree about almost everything, which can swamp a first attempt. Neighbours, such as systems and design, share a concern with wholes and disagree more narrowly, so the splits are easier to pin down.

When

When to reach for it, and when not to

Use it when

An important decision rests on one analysis, and the analysis came out suspiciously neat.
Switching the frame produced two restatements that pull in different directions, and both look right.
Two people or departments keep disagreeing about a proposal, and you suspect they are using different frames without saying so.
The decision is expensive or hard to reverse.

Skip it when

The group is new to the four ways. Run naming and switching a few times first.
Relationships in the room are already tense. This move turns frames into camps very easily.
One frame clearly fits and the other is there for balance. That is a debate format, not this move.
There is no time to act on the open questions. A list of unknowns nobody will investigate does harm.

Before

Before anyone is in the room

Choose the two frames in advance, and write down why those two.
Split the group on purpose. Put the proposal’s champion in the frame most likely to criticise it.
Brief one person in each half to keep it in the frame, with a one-page summary of what that way of thinking asks.
Print the side-by-side grid and the open questions sheet.
Agree with the most senior person beforehand that they will speak last in the comparison.
Book the follow-up date for the open questions before the session, not after.

Steps

How to run it

7 steps, about 75 min in all. The questions in quotation marks are meant to be read out as they are.

01
Name the two frames and why
5 min
Tell the group which two ways of thinking you are using and why. If they come from the switching session, say which restatements they were. Read out the situation in one sentence so both halves start from the same words.
Ask the room
- We are going to run the same situation through two ways of thinking. Here is the situation, word for word.
- Before we split, does anyone expect these two frames to agree?
Watch for Someone suggesting a third frame “for completeness”. Two is the move. Three becomes a carousel.
02
Split the group and brief each half
5 min
Send each half to its own end of the room with its frame’s summary and a sheet of flip chart paper. Tell them they will report findings, not recommendations, and that they must stay inside their frame even when it feels unfair.
Ask the room
- Using only your frame, what is going on here?
- What does your frame predict will happen if the proposal goes ahead as written?
- What would your frame want to know that it does not know yet?
Watch for The proposal’s champion explaining the proposal to their half instead of thinking in the frame.
03
Work the situation in each frame
25 min
Each half works the situation in its frame only. A systems half draws loops and predicts behaviour over a year. A design half sketches a rough version and walks a user through it. A critical half lists the claims and the evidence for each. A computational half writes the steps. Each half ends with five to eight findings written as plain statements.
Ask the room
- What is the most important thing your frame shows that was not in the proposal?
- What does your frame predict will have happened by the end of the first term?
- Can you write each finding as a statement someone could disagree with?
Watch for Halves drifting back into the proposal’s original frame. The person briefed to keep the frame pulls them back.
04
Put the results side by side
10 min
Bring both sheets together and copy the findings onto the side-by-side grid, matching them where they talk about the same thing. Each half reads its findings aloud without comment from the other. Questions of clarification only.
Ask the room
- Read your findings out. The other half: questions of clarification only, no rebuttals yet.
- Which of their findings is about the same thing as one of ours?
Watch for Rebuttals disguised as questions. “Have you considered that…” is a rebuttal.
05
Mark agreement and disagreement
15 min
Go line by line. Mark each pair of findings as agree, split, or seen by one frame only. Do not resolve the splits. The aim is to locate them precisely. A good split fits in one sentence: frame A says this, frame B says that.
Ask the room
- Do these two findings agree, disagree, or talk past each other?
- Can we say the disagreement in one sentence?
- What did one frame see that the other could not see at all?
Watch for The session turning into a contest the more senior half wins. If it starts, stop and restate the rule: we are mapping the disagreement, not settling it.
06
Turn each split into a question
10 min
For each disagreement, write a question that would settle it and a way to get the answer. Good questions can be answered by doing or looking at something within a few weeks. Bad questions can only be argued about.
Ask the room
- What would we need to see to know which frame is right here?
- How could we find that out in the next few weeks?
- What answer would make us change the proposal?
Watch for Opinions dressed as questions. “Is the project worth it?” cannot be investigated. “Does the afternoon survive a week with two staff absent?” can.
07
Decide what to find out first
5 min
Choose the one or two questions that would most change the decision, and give each an owner and a date. Write the agreed findings at the top of the proposal. Those are the parts you can now state with more confidence.
Ask the room
- Which question, once answered, would most change what we decide?
- Who will find out, and by when?
- What can we now say with more confidence than we could this morning?
Watch for A long list of open questions with no owners. Two with owners beat ten without.

Worksheets

Two sheets to print and fill in

Each one prints on its own sheet of A4. The grey line under each is an example from the worked situation.

Ways of thinking · Holding two at once · Worksheet

Side by side

Puts the two frames’ findings next to each other, so the agreements and splits can be seen and marked.

Situation

Frame A

Frame B

Date

AboutWhat the finding concerns	Frame A found	Frame B found	Agree, split or one only	The split in one sentence

Example About: whether a trial proves anything. Systems: a one-off trial gets the keenest staff and no timetable pressure. Design: a trial is how we learn what the afternoon needs. Split.

geraldajam.com/work/toolkit

Ways of thinking · Holding two at once · Worksheet

Open questions

Turns each disagreement into something a named person can find out, with the answer that would change the plan.

Situation

Keeper of this sheet

Review on

Question	How we will find out	Who	By when	The answer that would change the plan

Example Does the afternoon hold in an ordinary week? Run the second trial with the usual relief arrangements. Ms Chen. Before the March holidays. If it falls apart with two staff away, protect the staffing before committing.

geraldajam.com/work/toolkit

Worked example

The squeezed timetable beside the trial afternoon

One afternoon a week for Secondary 1

The switching session left two frames pulling against each other. The design restatement said run a trial afternoon. Mr Hafiz, thinking about his timetable, kept saying a trial would prove nothing. The group gave the two frames 75 minutes side by side. At the facilitator’s suggestion, Ms Chen joined the systems half. Mr Hafiz joined the design half and agreed to speak last.

The systems half They drew the squeeze. The afternoon takes one lesson a week from each of three subjects. Each department protects its content by packing its remaining lessons. The project becomes the soft afternoon, the first to lose staff when relief cover is short, and before each weighted assessment it is used for catch-up. Their prediction: by its second term the afternoon exists on the timetable and not in practice. Ms Chen said the loop out loud and did not enjoy it.
The design half They sketched one trial afternoon on the food waste project, with Mdm Kavitha as the client and Arjun’s class as the users. Students weigh a day’s canteen waste, trace where it goes, and write to the canteen stallholders. Their finding: nobody can know what the afternoon needs until a class has been through one. Mr Hafiz added that a good afternoon builds the demand that protects it.
Where they agreed Both halves said do not commit to a full year now. Both said the afternoon fails if it is made of three departments’ leftovers. Both said nobody yet owns planning time. Those three went to the top of the proposal as findings.
Where they split Two splits. First, whether a trial proves anything. Design said a trial is how you learn. Systems said a one-off trial gets the keenest staff and no timetable pressure, so it says nothing about week fourteen. Second, what success means. Design measured it by what students made and said. Systems measured it by whether the time survived.
The awkward part Mr Hafiz, in the design half, kept sliding into the systems case, because he knew the timetable better than anyone. The facilitator held him to speaking last, and the design half’s findings stood on their own. Afterwards he said the systems half had been gloomier than he would have been. It was the first time anyone had heard him defend the project.
The questions The first split became: does the afternoon hold in an ordinary week, with the usual relief arrangements? The test was to run the second trial in a week with no special staffing. The second became: what would tell us, well before its second term, that the afternoon is being squeezed? The answer was a one-line weekly record, kept by the Secondary 1 Year Head, of whether the afternoon was used for the project.

What came out of it

The proposal went to the school leaders with three agreed findings, two trial afternoons instead of one, and a weekly record that would show the squeeze if it came. Neither frame won. Design got its trial. Systems got the condition that the second trial runs in an ordinary week, which is the version that would tell the school something.

Signs

How to tell it is working, and when it has gone hollow

Working

At least one finding is supported by both frames and goes into the proposal with more confidence than before.
Each split can be stated in one sentence without naming a person.
Every open question has an owner and a date, and someone checks on them.
The proposal’s champion ends up arguing the other frame’s case.

Gone hollow

The two halves become two camps, and the more senior camp wins.
Every finding is marked “agree”, because nobody wanted the awkwardness of a split.
The open questions are written up and nobody investigates them.
One frame was picked as a straw man, and the result was decided before the split.

Variations

Other ways to run it

In 45 minutes

Cut the working time to 15 minutes per half and the comparison to 20. Skip the formal grid. Use two columns on one flip chart, with a tick for agreement and a cross for a split. You will find fewer splits, but the big ones survive.

On your own

Write the analysis twice, once in each frame, on separate days so the first does not colour the second. Then lay them side by side. It is weaker than the group version, because you will be kinder to your favourite frame, so ask someone else to mark the splits.

With students

Give two halves of a class the same question, such as whether the school should ban phones at recess, one half as systems thinkers and one as critical thinkers. Each half presents, and the class marks agreements and splits. Secondary 1 students tend to find the splits more interesting than the answer, which is the point.

What to do with what you made

Answer the open questions before the decision, not after it. Put the agreed findings at the top of the proposal so the people deciding can see what both frames support. If the move changed how the group works, turn the side-by-side into a routine for large decisions.

Making it routine Shrink the move to a question the team can ask without a facilitator.Systems thinking The frame that found the squeeze, with what it notices and what it cannot see.Holding the frame lightly The complexity move that goes a step further: how each frame shapes what it sees, and who it leaves out.

Tools

Tools, templates and guides worth having

Published elsewhere, and worth the time.

Method
Integrative thinking
Martin’s method from The Opposable Mind (2007). Name the two opposing models, find what each gets right, and look for a third that keeps both. A natural next step after the splits are marked.
Method
Analysis of competing hypotheses
From Heuer’s Psychology of Intelligence Analysis (1999). A matrix that tests the evidence against several explanations at once. Useful when a split is about what is causing something.
Method
The pre-mortem
Klein’s method (Harvard Business Review, 2007). Imagine the plan has failed and explain why. A systems half can use it to find the squeeze quickly.
Guide
Decision quality checklist
Kahneman, Lovallo and Sibony’s twelve questions (Harvard Business Review, 2011) for spotting a recommendation built from a single frame.
Method
1-2-4-All
A Liberating Structure (Lipmanowicz and McCandless, 2013). Use it inside each half so the quiet members are heard before the loud ones.

Reading

Where it comes from

Start with the first on the list.

Martin, R. (2007). The Opposable Mind. Harvard Business Review Press.
Integrative thinking: holding two opposed models and building a better third.
Jackson, M. C. (2003). Systems Thinking: Creative Holism for Managers. Wiley.
The case for using several methods on one problem, and how to choose them.
Mingers, J., & Brocklesby, J. (1997). Multimethodology: towards a framework for mixing methodologies. Omega, 25(5), 489–509.
The academic case for combining frames.
Kahneman, D., Lovallo, D., & Sibony, O. (2011). Before you make that big decision. Harvard Business Review, 89(6), 50–60.
A checklist for spotting a single-frame recommendation.

Singapore