Yes, No, Maybe: The Third Answer Binary Thinking Can't See
Ask most decision-making frameworks to model a choice and they will draw you a fork: two branches, one arrow. Yes or no. Left or right. Go or no-go. It is a clean model, and it is almost always wrong. The most useful answer to a hard question is rarely yes and rarely no – it is maybe, followed by a decision to go and look. That third answer is not indecision. It is the entire discipline of lateral thinking, and it is also, whether project teams realise it or not, the entire justification for working in iterations rather than committing to a single upfront plan.
The Fork in the Road
The psychologist Edward de Bono, who coined the term lateral thinking in 1967, drew a sharp distinction between it and what he called vertical thinking – the traditional, step-by-step logic where each move is judged right or wrong before the next one is taken. Vertical thinking digs the same hole deeper. It is precise, and it is often exactly what a well-understood problem needs. But it has a blind spot: it assumes the right direction is already knowable, and that the only job left is to judge each step against it as true or false, yes or no.
Yes/no logic is vertical thinking's native language. It works beautifully when the information needed to answer already exists. It fails quietly – and then loudly – when the honest answer requires information nobody in the room currently has. Forced into a corner, a team under pressure to answer yes or no too early does one of two things: it stalls in analysis until it feels safe enough to be certain, or it gives a confident answer that is simply wrong. Lateral thinking offers a way out of that trap. Instead of judging a direction before moving, it moves first, deliberately entertains the less obvious paths, and defers judgement until there is something real to judge.
What You Don't Know You Don't Know
The clearest map of this problem is not new. It is the four-box grid popularised by Donald Rumsfeld in 2002 – what you know, what you know you don't know, and the far more dangerous territory of what you don't know you don't know. The first two boxes are manageable. A known fact is simply used. A known gap becomes a risk register entry, a contingency plan, a question someone is assigned to answer. Both can be met with a straightforward yes or no once the work is done.
The fourth box cannot. You cannot ask a yes/no question about a risk you have not yet discovered exists. That is precisely what makes it dangerous, and precisely why no amount of upfront planning – however rigorous – can fully protect against it.
The danger was never in the boxes with labels. It sits in the untitled corner – the one that can hurt you precisely because nobody drew a line around it yet.
Four Kinds of Not Knowing
It helps to break that fourth box down further, because “unknown” is not one thing. Project risk thinking generally separates not-knowing into four distinct flavours, each demanding a different response.
| Quadrant | Uncertainty Type | What It Looks Like | Where It Bites a Project |
|---|---|---|---|
| Known Knowns | Stochastic Uncertainty | Predictable future states, independently verifiable project data | The comfortable zone – historical velocity, agreed requirements, settled estimates. Rarely where projects fail. |
| Known Unknowns | Aleatory Uncertainty | Possible states identified, ambiguous outcomes, quantifiable variables | The risk register. You know the storm might come, so you plan the shelter and the buffer in advance. |
| Unknown Knowns | Epistemic Uncertainty | Researchable facts, unshared skills and information, untapped resources | Knowledge that already exists somewhere in the organisation but has not reached the team – a spec sitting in someone else's inbox, an expert nobody thought to ask. |
| Unknown Unknowns | Ontological Uncertainty | Hidden relationships between variables, unpredictable events, “bolts from the blue” | The ravine hidden in the forest. It does not appear on any risk register, because nobody knew to look for it – it can only be found by moving toward it. |
Four different kinds of not-knowing, each needing a different response – and only one of them, ontological uncertainty, cannot be planned for in advance. It has to be discovered.
“Maybe” is routinely mistaken for fence-sitting – a failure of nerve dressed up as nuance. It is nothing of the sort. A genuine maybe is a decision: a decision to go and find out, rather than a decision to guess. You cannot give a truthful yes or no about an unknown unknown, because the information a yes or no requires does not exist yet. The only honest move is to say maybe, and then immediately follow it with a concrete, small, reversible step that will turn the maybe into a fact.
The Forest, the Panga, and the Ravine
Picture a forest that has to be crossed to reach the other side. From the tree line, the canopy is too dense to see through – not because anyone concealed anything, but because you are standing at ground level looking sideways into leaves, not overhead with a survey map. No one commissioned a drone. So the crossing begins the only way it can: with a panga, cutting open one metre of path, then the next.
Twenty metres in, the ground disappears. Not a gentle slope – a ravine, wide and deep enough that momentum will not carry anyone across it. It was there the entire time. It simply could not be seen from the tree line, and no plan drawn up before the first cut would have shown it either, because from where the plan was written it was an unknown unknown. It only became a known fact by doing the work of cutting toward the unknown.
This is precisely where binary thinking fails hardest. “Should we cross this forest – yes or no?” was answered twenty metres ago, before the ravine existed as a fact anyone could act on. “Can we get across – yes or no?” is not yet answerable either, not because the question is badly framed but because answering it truthfully requires information nobody currently holds: how wide the ravine is, whether there is a narrower crossing nearby, what materials are within reach. The only intellectually honest answer is maybe – paired with a resolve to find out, using whatever is at hand: the vines overhead, the trees on either bank, and the patience to test a knot, watch it fail, and adjust it until a bridge holds a person's weight.
That vine bridge was never in the plan. It could not have been, because the ravine that made it necessary was not in the plan either. It exists only because someone kept cutting past the point where the original plan stopped being useful – and was willing to build the solution out of whatever the newly revealed problem left lying around.
Agile as Lateral Thinking in Motion
Iterative, incremental delivery formalises exactly this posture. Nobody plans the entire crossing before the first cut; they plan the next small, reversible step, take it, see what it reveals, and re-plan from there. A sprint is one deliberate swing of the panga, not a guess at the whole route. A reprioritised backlog is the team discovering the ravine and routing around it – not a failure to have predicted it, and not a reason to pretend the fork in the road was decided correctly the first time. A retrospective is the moment spent examining which swing of the blade worked and which did not, before the next one is taken.
Notice what made the vine bridge possible: the first knot was allowed to fail. It was tied, tested, and it slipped – cheaply, safely, before anyone trusted it with their weight. That is the whole logic of failing early. A knot that fails on the first small test costs a few minutes and a length of vine. The same knot, untested and trusted on the strength of a confident plan alone, costs a person their footing over a ravine. Agile's obsession with short iterations is not really about speed – it is about making sure that when something is wrong, it fails while it is still small, cheap, and easy to learn from, rather than late, expensive, and disguised as success right up until it collapses under real weight. A maybe that is never tested is just a guess wearing a more comfortable name. A maybe that is deliberately put in a position to fail quickly is how it turns into a fact.
| Lateral Thinking Move | What It Looks Like | Its Agile Equivalent |
|---|---|---|
| Provocation | Deliberately entertaining an option that sounds unreasonable, purely to see what it reveals | A spike or proof-of-concept – not built to ship, built to find out |
| Movement Over Judgement | Exploring an idea before ruling it right or wrong | A sprint review that shows rough, unfinished work rather than waiting for certainty |
| Random Entry | Starting from an unrelated point to escape a well-worn track of thought | Cross-functional pairing or rotating roles – a fresh angle notices what a specialist in-track would not |
| Escaping the Dominant Idea | Treating the first, most obvious thought as a starting point, not the answer | Backlog refinement that re-examines “settled” stories as more is learned |
| The Third Point | Maybe, not yes or no – a commitment to investigate rather than to decide | A Definition of Ready that treats a story as unproven until a small increment has tested it |
| Failing Early, Cheaply | Testing the knot before trusting it with real weight | Iterations sized so a wrong idea fails in days, not quarters – and costs a rework, not a launch |
Signs You Are Stuck at the Fork – and Signs You Are Cutting Forward
What Binary Thinking Looks Like – and What Lateral, Incremental Thinking Looks Like
- A decision is forced to yes or no before enough of the forest has been cut open to see what is actually in it
- A plan is treated as complete simply because no one has found the ravine yet
- Retrospectives ask “did we follow the plan” rather than “what did we learn that should change the plan”
- The first idea that arrives is treated as the only idea worth acting on
- “Maybe” is treated as a failure to decide, rather than a decision to investigate
- A wrong idea is only discovered after months of investment, when it is far too costly to change course
- Small, reversible increments are used specifically to reveal what a single big upfront decision never could
- Ideas are deliberately tested small and early, so that failure – when it happens – is cheap and informative rather than expensive and hidden
- Discovering an obstacle triggers re-planning, not blame for having failed to predict it
- Provocative “what if” questions are welcomed even when they sound unreasonable at first
- Retrospectives genuinely change the next iteration's plan based on what was actually found
- “Maybe” is always followed immediately by a concrete next step small enough to resolve it
The destination does not change – the far side of the forest is still the far side of the forest. What changes is the belief that the route to it could have been fully known before anyone set out. It could not. The ravine only existed, for practical purposes, once the panga reached it. The vine bridge only existed once the ravine demanded one. Yes and no are the language of a route already known. Maybe, followed by the next small cut, is the only honest language for a route still being found – and it is very often the only route that was ever going to get anyone across.
Further Reading
The ideas behind this article draw on lateral thinking theory, the well-known known/unknown risk framework, and the sense-making models that formalise why some problems can be planned and others can only be discovered.