Nearly three months after the infamous blue and black dress (or was it white and gold?) tore the Internet apart, three teams of scientists have provided a closer look at the science behind the viral phenomenon. In their papers, published Thursday in the journal Current Biology, the teams have proposed reasons that different people saw different colors, and what the whole thing means for our understanding of visual perception.
In one study, Michael Webster, a psychologist from the University of Nevada, Reno, places blame for Dressgate on the ambiguity of the color blue, and people’s inability to reliably discern blue objects from blue lighting.
In the second study, Karl Gegenfurtner, a psychologist from Giessen University in Germany, had 15 volunteers use a customizable color wheel to show what color they saw on the dress. He found that the pixels of the dress matched with the natural spectrum of blues and yellows we see from sunup to sundown, making it more difficult for people looking at it to tell how the color of the lighting might affect perception.
In the third study, Bevil Conway, a neuroscientist from Wellesley College, asked more than 1,400 people what colors they saw when they looked at the dress. He found that people fell into not only the two warring groups, “blue and black” and “white and gold,” but also a third group: blue and brown. He posits that this is an example of individual variety in internal models — the set of assumptions that the brain applies to processing visual cues.
In the case of the dress, Dr. Conway said, the poor quality of the image is what sets off the brain’s internal model. The ambiguous conditions and lack of context are important “because your brain doesn’t have enough information to discern it,” he said.
“So the brain has to turn to the internal model and say, ‘Hey, guru, what do you think is going on out there?’ ” he added.
Each person’s internal model reacts differently. He said that people who saw the dress as white and gold did so because their internal model presumed they were observing the dress under a blue sky. They discount the color blue. For people who saw blue and black, their internal models primed them to think they were viewing the dress under orange incandescent light.
David Brainard, a neuroscientist from the University of Pennsylvania who wrote a dispatch that accompanied the three papers in the journal, said that each paper contributed to our understanding of the color constancy hypothesis, which is the ability to perceive an object’s color regardless of the color of the light source being shone on it.
“These papers are particularly valuable from the scientific view,” Dr. Brainard said, “because they move the discussion of the dress into a realm where we have actual data about the phenomenon beyond the tens of thousands of tweets.”