The bus step
The best-known account is Henri Poincaré’s. In “Mathematical Creation”, published in Science and Method in 1908, he described working for fifteen days on a class of functions without result. One sleepless night, after black coffee he did not usually drink, he made his first breakthrough. Then he left Caen for a geological excursion and, by his account, forgot about mathematics. In Coutances, as he put his foot on the step of an omnibus, the idea came to him that the transformations he had used were identical with those of non-Euclidean geometry. He did not check it. He went on with his conversation and verified the result after returning to Caen.
He reported similar moments on a seaside walk and in the street during military service, and called this sudden illumination “a manifest sign of long, unconscious prior work”. It is one person’s recollection. It is also the shape of the story people tell about showers.
What the experiments say
Psychologists call the effect of setting a problem aside incubation. In 2009 Sio and Ormerod of Lancaster University pooled the experimental studies in a meta-analysis for Psychological Bulletin. They found a positive incubation effect overall. Open-ended tasks, such as thinking of many uses for an object, benefited more than insight puzzles. Longer work on the problem before the break gave a larger effect. Filling the break with a demanding task gave a smaller one, and for word-based insight problems a low-demand task did better than rest.
A shower fits the profile of a low-demand task: occupied hands, little to decide. None of these studies tested showers, so that part is an inference.
The mind-wandering explanation
In 2012 a team at the University of California, Santa Barbara, led by Baird, offered a mechanism. Participants worked on unusual-uses problems, then took a break doing a demanding task, an undemanding task, resting, or no break at all. Those given the undemanding task improved substantially on the problems they had already seen, and they reported more mind wandering during the break. The authors suggested that simple tasks that let the mind wander may help creative problem solving.
That explanation has not held up well. In 2016 Smeekens and Kane at the University of North Carolina at Greensboro reported three studies in which mind wandering measured during the break did not predict later creativity, which they called “especially problematic” for the idea. A 2025 study in Brain Sciences reported that it also failed to replicate Baird’s result, and noted that two other attempts had found no advantage either.
Where that leaves the shower
Incubation itself has reasonable support: stepping away from a problem you have worked on properly tends to help, and a light activity seems a good use of the break. Why it helps is unsettled. Wandering thoughts are one proposal, and the evidence for that one is weak. The ideas, when they do arrive in the shower, are mostly drawing on work done before it.