Skip to content
Mental Models

Lateral Thinking: Walk Around the Problem

Provocation & Random Entry

De Bono's proactive lateral tools: use PO to license absurd provocations and random words to jolt yourself out of the groove on purpose, then harvest the workable idea by movement, not judgment.

12 min Updated Jul 6, 2026

In lesson 3 you learned to drop hidden assumptions — to notice the wall you’d built around a problem and quietly remove it. That’s powerful, but it’s reactive: you can only drop an assumption once you’ve caught yourself making it. What if you’re stuck and no assumption is jumping out? You can’t just sit there waiting for inspiration to knock.

This lesson is about the opposite move: manufacturing new starting points on demand. De Bono built a small kit of tools that deliberately shove your thinking off its usual track — not because the jolt is smart, but because a self-organising pattern system will never leave its own groove voluntarily. You have to push it.

Before you read — take a guess

A colleague, brainstorming ways to cut delivery times, blurts out: 'What if the customer delivers the package to us?' Everyone laughs. What's the most *productive* thing to do with that idea?

Movement, not judgment: the switch under everything

Here’s an analogy. In ordinary thinking, an idea is a defendant: it walks in, you cross-examine it — Is this true? Does it fit the facts? Would it survive a meeting? — and you convict or acquit. That’s judgment, and it’s exactly what vertical thinking is for. You need it. Bridges stay up because engineers judge ideas ruthlessly.

But judgment has a side effect: it kills ideas at the door, before they’ve had a chance to lead anywhere. And in the generative phase, an idea’s value often isn’t in the idea itself — it’s in the direction it points.

So de Bono’s tools all run on a different mental setting he called movement. Instead of asking “is this idea right?” you ask “where does this idea lead? What does it suggest? What follows from it? What’s interesting about it?” You treat the idea not as a verdict to reach but as a stepping stone to step off. You’re not trying to stand on it — you’re using it to get somewhere you couldn’t have walked to directly.

This is the single most important idea in the lesson, so let it land: a deliberately “wrong” idea is useless when judged and priceless when moved-from. Judged, “cars have square wheels” is idiotic — case closed. Moved-from — where does that lead? — it points to a suspension that rises to meet each bump before it hits, which is roughly how modern active suspension actually works. Same statement. The only thing that changed was which mental setting you fed it into.

Tip:

The one-word test

When an idea shows up, notice your first reflex. If it’s “yes/no”, you’re judging. If it’s “and therefore…?”, you’re moving. In the generative phase, catch the yes/no reflex and swap it for and-therefore.

Why not just “think positive”?

This isn’t a plea to be nice to bad ideas. Judgment stays essential — you’ll bring it back with full force in the next lesson. The point is timing. Movement and judgment are two different jobs, and doing them at the same instant means judgment always wins (it’s faster and it feels responsible). The tools below are ways to force a window where movement runs first and judgment waits its turn.

Provocation and PO: licensed nonsense

A provocation is a statement you make knowing it’s unreasonable, purely to knock your thinking out of its channel. The problem is social and psychological: say something absurd in a normal register and either people argue with it (judgment kicks in) or you feel like a fool and retreat. So de Bono coined a word to carry the absurdity safely: PO.

PO is a signal — like a flag — that means “this is a provocation, not a proposal.” Putting PO in front of a statement is a license: it announces “I am not claiming this is true or good or even sane; I’m using it as a stepping stone.” It buys the statement immunity from judgment just long enough to move from it. Think of it as the lab coat that lets you handle a dangerous idea without it exploding in your face.

There are a few reliable ways to manufacture a provocation on purpose:

TechniqueWhat you doExample
ReversalTake the normal direction and flip it”PO: the customer pays us to leave a review”
ExaggerationPush a number to an absurd extreme”PO: one support agent handles a million tickets a day”
Wishful thinkingState a fantasy outcome as fact”PO: the factory is downstream of itself”
DistortionWarp a normal relationship or sequence”PO: you eat dessert before you’re hungry”

None of these is meant to be believed. Each is a diving board.

Worked example: the factory downstream of itself

De Bono’s own classic. A factory sits on a river and dumps its wastewater downstream, where it becomes the next town’s problem. Judged normally, that’s just how factories work — regrettable, but standard.

Now the provocation:

PO: the factory is downstream of itself.

Judged, this is meaningless — a factory can’t be below its own position on a river. So we don’t judge it. We move from it. Where does that lead?

  1. “Downstream of itself” would mean the factory takes in its own output as its input.
  2. So the factory would drink the water it just used.
  3. And therefore — the workable idea — build the water intake downstream of the water outflow.

Now the factory pumps in the very water it dumped out. The pollution isn’t the next town’s problem anymore; it’s the factory’s own problem, first thing every morning. So the factory is forced to clean its discharge — not by regulation, not by conscience, but by plumbing. That’s a real, adoptable pollution-control principle, and it fell out of a sentence that was literally impossible.

Trace the chain: absurd provocation → movement (where does this lead?) → workable idea. The absurdity was never a proposal. It was the launch, and movement was the flight.

Follow the chain: 'PO: a restaurant has no menu.' Which response uses *movement* to reach a workable idea, rather than judging the provocation?

Random entry: letting an unrelated word open the lock

Provocation jumps your pattern from the inside — you distort the problem itself. Random entry jumps it from the outside: you feed in a word that has nothing to do with the problem and force yourself to build associations from it.

The analogy: your problem is a house, and your thinking always comes in through the front door — the same framing, the same hallway, the same three rooms you always end up in. A random word is a window on the far side of the house. Climb in there and you’re suddenly standing in a room the front door never leads to. The point isn’t the window; it’s that you arrive somewhere your habitual path can’t reach.

Why does this actually work, mechanically? Recall the self-organising pattern system from lesson 2: your brain files ideas into grooves, and later thinking flows down those grooves automatically. When you enter through the problem’s own framing, you get dumped into the same well-worn channels every time — that’s why you’re stuck. A random word activates a completely different set of channels, and where those unrelated channels happen to cross the problem, you get a connection your normal route would never have made. The randomness isn’t decoration; it’s the whole mechanism. It guarantees a fresh entry point.

Worked example: cutting office energy use

Problem: how do we reduce energy use in the office? Your normal grooves offer the usual suspects — LED bulbs, turn off monitors, a stern email. Fine, but tired.

Now grab a random word. (Genuinely random — more on that in a second.) Say the book falls open on cinema.

Force associations from “cinema”, ignoring whether they seem relevant:

  • It’s dark — dim lighting, only what you need to see.
  • Everyone leaves together when it ends.
  • It runs on a schedule — sessions start and end at set times.
  • There’s a clear “the show’s over” signal — lights up, credits roll.

Now cross those back into the problem:

Scheduled “lights-out” moments — like a screening ending — where the office dims and non-essential systems power down on a set timetable, and everyone knows it’s coming.

That’s a genuinely different idea from “remember to turn things off”. It reframes energy-saving as a scheduled event rather than a nag, and it came straight out of a word about movies. The problem’s own framing would never have handed you “cinema”.

Warning:

The word has to be genuinely random

This is the rule people break, and breaking it wrecks the tool. If you choose a word — “hmm, let me pick something related to energy… a lightbulb!” — you’re just re-entering through the front door. You’ll only confirm the pattern you already have. The word must arrive without your judgment: point at a random noun on a list, open a book and stab the page, roll for it. The discomfort of getting a word with no obvious connection is exactly the signal that it’s doing its job.

Why must the random word be picked at random (from a list, a book, a dice roll) rather than chosen because it 'feels related' to the problem?

Why these feel silly — and work anyway

Let’s be honest: saying “PO: the factory is downstream of itself” out loud, or building an office-energy plan out of the word cinema, feels ridiculous. That feeling is not a bug. The discomfort is the point.

Here’s why. Your brain resists leaving a groove — that’s what a self-organising pattern system does, it’s the reason patterns are efficient. Willpower alone (“okay, be more creative now”) doesn’t help, because willpower still runs inside the existing pattern; you can strain as hard as you like and you’re still on the same track. What a technique gives you that willpower can’t is an artificial jolt — an external, arbitrary shove that doesn’t come from your framing and therefore isn’t captured by it. The silliness is simply the sensation of leaving the groove. If it felt natural, you wouldn’t have gone anywhere new.

Info:

Silly is a symptom, not a strategy

Feeling silly means the tool moved you off your usual track — good. But feeling silly is not itself the goal. Plenty of weird statements go nowhere. The jolt is only half the tool.

The failure mode: provocation without the harvest

Here’s the discipline that separates lateral thinking from a comedy improv game. A provocation you don’t harvest is just noise. If you say “cars have square wheels”, laugh, and stop — you’ve generated nothing. The tool has two halves and both are non-negotiable:

  1. The jolt — the provocation or random word that gets you off the track.
  2. The harvest — the movement that follows the jolt to a workable idea.

Skip the harvest and you’re just being weird for its own sake. Being weird is not the goal; getting unstuck is. Every PO statement, every random word, carries an obligation: you must eventually apply movement until it pays out in something you could actually use. If nothing workable emerges after honest movement, fine — drop it and jolt again. But the provocation is never the destination. It’s the diving board, and a diving board with no pool under it is just a plank you fall off.

Match each of de Bono's terms to its plain meaning:

Pick a term, then click its definition.

Recap: provocation & random entry

Question 1 of 40 correct

In the generative phase, what is the key difference between *movement* and *judgment*?

Check your answer to continue.

Big picture

Provocation & random entry at a glance

  • Generating starting points on purpose
    • Movement, not judgment
      • Judgment: is this right?
      • Movement: where does it lead?
      • Wrong idea = stepping stone, not verdict
    • Provocation (PO)
      • PO = 'jolt, not proposal' flag
      • Reversal / exaggeration / wishful / distortion
      • Factory downstream of itself → intake below outflow
    • Random entry
      • Unrelated word = window on the far side
      • Opens grooves the framing keeps you from
      • Must be GENUINELY random, not chosen
    • The discipline
      • Feeling silly = you left the groove
      • Jolt + harvest — both required
      • No harvest = just noise

Where this goes next

You now have two proactive tools for breaking out of the groove on command: provocation (jolt from the inside with PO) and random entry (jolt from the outside with an unrelated word) — both run through movement and both obliged to end in a harvest. Between dropping assumptions (lesson 3) and these, you can now generate wild, unreasonable starting points at will.

But generation is only half the machine. Produce a pile of provocations and half-formed leads and you’ve created a new danger: strangling good ideas by judging them too early — the exact reflex these tools exist to suspend. In lesson 5, Diverge, Then Converge, you’ll learn how to organise the two modes so they don’t sabotage each other: when to open the tap wide and generate without mercy, and when — separately, deliberately, and only then — to bring judgment roaring back to sort the gold from the noise.

Mark lesson as complete