Order of European Inventors and Innovators
by Ileana Vlad, clinical psychologist and psychotherapist
In 1926, in The Art of Thought, Graham Wallas described four stages an idea passes through: preparation, incubation, illumination, verification. The model has lasted a century not because it is complete — it is not — but because it is useful in a very practical way: it tells you which stage you are in and, above all, which mistake is characteristic of each.
1. Preparation
You gather what is known, delimit the problem, formulate it. It is the most underrated stage, because it resembles background reading and produces nothing visible.
The characteristic mistake: starting to look for solutions before the problem has been formulated. It is easy to spot — if you cannot write the problem in a single sentence, you are not yet in the second stage, however much you are drawing.
The reverse is also worth saying: a well-formulated problem is already half solved, and reformulating it is itself a creative operation. Many blockages are not blockages of solution, but of statement.
2. Incubation
The stage that, seen from outside, looks exactly like idleness. Nothing happens. Nothing is produced. The person appears to have given up.
The characteristic mistake: reading the silence as failure and stopping. Or, just as often, not allowing yourself the stage at all — because in an organisation that reports weekly, a week with no visible result is hard to defend.
About what actually happens in that interval I have written separately: incubation has been measured, and it is not rest.
3. Illumination
The short stage. The solution appears, usually whole, often at the least convenient moment, and almost always accompanied by a very strong sense of certainty.
The characteristic mistake is bound up with that very sensation: the feeling of being right is not evidence that you are right. Illumination arrives with the same intensity whether the solution is good or wrong. From the inside there is no difference between them. That is why the fourth stage exists.
4. Engineering verification
The longest, the least glamorous, and the stage in which inventions actually die. Here things are measured, calculated, built and broken.
The characteristic mistake: having fallen in love with the idea and therefore testing it gently. A gentle test is not a verification, it is a ceremony. A good test is one built to demolish the idea; if it fails to, only then have you learned something.
This is also where the link with explanatory style shows: verification produces failures in series, by construction. Whoever cannot cross this stage never arrives at an applied invention.
What the model does not say
It does not say the stages come in order, once. In reality they loop: a failed verification sends you back into preparation with the problem reformulated; incubation can occur dozens of times within a project, at different scales.
And it does not explain the mechanism of illumination. It is a map, not an engine. It tells you where you are, not how it works.
Why it is useful anyway
Because it changes what you do. If you are in preparation, you need information, not ideas. If you are in incubation, you need to be left alone — and that is hard to grant yourself and nearly impossible to grant someone else, unless you know the stage exists. If you are in illumination, you need paper. If you are in verification, you need someone who will try to contradict you.
The most practical use of the model is perhaps this: it protects the second stage. In a culture that measures visible activity, incubation has nobody to defend it unless it at least has a name.