The influence on science and the arts of Johann Wolfgang von Goethe — poet, playwright, novelist, proto-scientist, philosopher and general all-round egghead — is profound. His views about the physiological nature of colours, however, have never really caught on, in part because he proposed that colours are more an invention of the mind than a physical reality. One thing, however, rings true: the appearance of objects is not objective, but a conversation between the observer and observed.

Neuroscientists have long recognized that the perception of colour and shade depends strongly on context. Illusions exist, for example, in which one can be utterly convinced that black is white, depending on the surrounding patterns or the conditions in which an object is lit. But it is also true that all other things being equal, the perception of colour differs between people.

One editor of this journal, for example, once owned a car that was, in his opinion, quite clearly green. It remained green in all conceivable circumstances of context, shade and illumination other than complete darkness. Except, however, that everyone else was equally convinced that it was blue — including the vehicle-licensing authority. The car was not only blue — it was officially blue.

Last week, the Internet was deluged with strongly held opinion about colour, specifically of a dress. The dress was advertised as being blue and black. But if illuminated in a certain way, the dress appeared white and gold. People were absolutely convinced of its colour combination, one way or the other. The web exploded with chromatic debate after various celebrities bruited their opinions on Twitter. A straw poll of Nature’s editors (including the owner of the blue car) was roughly split down the middle, and convictions were strong — one way or the other.

The explanation for the illusion lies in the colour of the light in which the dress was photographed. The brains of people who read the overall ambience as too blue will overcompensate, seeing the dress as white and gold. Others, whose visual systems read that the lighting was not blue enough, saw the dress as blue and black.

Wired magazine hosted a full discussion on the effect (see go.nature.com/uqf7bo), and the consternation in that publication’s office seemed to reflect the brouhaha that briefly reigned in the otherwise serene halls of Nature. (The wheels of this international weekly journal of science briefly ground to a halt as so: “I can’t read any more manuscripts until I find out WHY?!”)

On being told of the illusion, some people — but not all — could just about force themselves to see the dress as black and blue rather than white and gold. The picture is a clear demonstration that colour perception varies between individuals, and according to the conditions of illumination. Such perception is distinct from the genetic conditions that predispose people to the various syndromes known as colour blindness.

Had the ghost of Goethe been watching ‘dressgate’, he might have allowed himself a rueful smile, given the brickbats thrown in his direction by his scientific critics even in his own time, who, he said, “forgot that science arose from poetry, and did not see that when times change the two can meet again on a higher level as friends.”