The bee dance that won a Nobel Prize, and the man who spent his career saying it meant nothing. · published 25 September 2026
Everyone assumes they’d simply smell their way to the flowers — instead they flew the exact direction and distance the dance had specified, landing precisely as wrong as the displacement, which proved they were carrying an instruction rather than following a trail, and vindicated half of the dissenter who’d spent thirty years insisting scent was the whole story.
These are the working notes, unedited. Everything the video
states as fact comes from here, and nothing else may be stated as fact.
The sections on limits and on what the finding does not show are
part of the brief, not an afterthought — see
the method.
The finding. Karl von Frisch worked out that a returning honeybee performs a figure-of-eight on the comb whose straight middle run encodes two things: the angle of the run against vertical gives the direction of the food relative to the sun, and the duration of the waggle gives the distance. He was awarded the Nobel Prize in Physiology or Medicine in 1973, shared with Konrad Lorenz and Niko Tinbergen. It is one of the few cases where an animal was shown to be communicating something abstract - a place that is not here - to another animal.
The challenge, which was serious. From 1967 Adrian Wenner, later with Patrick Wells, argued that the dance was a delusion of the observer. His case was not that bees do not dance; it was that the other bees do not use the information. Bees, he said, find food by smell - the odour the discoverer carries on her body and leaves at the site. And when Wenner ran the experiments with tighter controls against odour cues, recruits could not find the food from the dance alone. That is a real result, not a complaint. He and Wells published in Nature in 1973, the same year von Frisch took the prize.
What happened to Wenner. He was largely frozen out: for years the editors of major journals would not take his results, and he withdrew from active bee research. Whatever one concludes about the science, that is what happened to the dissenter while the consensus collected a Nobel.
How it was settled, thirty-eight years later. Riley, Greggers, Smith, Reynolds and Menzel, “The flight paths of honeybees recruited by the waggle dance”, Nature, 2005. They glued harmonic radar transponders to bees that had attended a dance and tracked the actual flight paths - the thing nobody had ever been able to see.
The manipulation that decided it. Some recruits were caught as they left the hive and released up to 250 metres away. If they were following a smell, they would have gone to the food. They did not. They flew the vector the dance had given them - the direction and distance that would have been correct from the hive - and ended up in the wrong place, displaced by exactly the amount the experimenters had moved them. They were carrying an instruction, not following a trail.
Where it actually landed, which is not a clean win. The accepted picture now is that both mechanisms operate: the dance carries recruits to the right area, and odour does the last part, homing them onto the flowers. Von Frisch was right that the dance conveys location. Wenner was right that odour matters and that the early controls for it were not good enough. The textbook version drops the second half.
Limits
The 2005 study is one study, and radar tracking constrains how far and where you can work. It settled the specific question of whether recruits use the vector, not every question about what bees attend to. And “language” is doing a lot of work as a word here - conveying a direction and a distance is not the same thing as a grammar.
What this does not show
Not that Wenner was a crank; his odour experiments were competent and his objection forced a level of control the field had not applied to itself. Not that von Frisch fabricated anything. And not that consensus is usually wrong - it was substantially right here. The uncomfortable part is narrower: the objection was reasonable, it took nearly four decades and a technology nobody had in 1967 to answer it properly, and in the meantime the person raising it could not get published.