This past February a photograph of a dress nearly broke the Internet. It all started when a proud mother-in-law-to-be snapped a picture of the dress she planned to wear to her daughter's wedding. When she shared her picture with her daughter and almost-son-in-law, the couple could not agree on the color: she saw white and gold, but he saw blue and black. A friend of the bride posted the confusing photo on Tumblr. Followers then reposted it to Twitter, and the image went viral. “The Dress” pitted the opinions of superstar celebrities against one another (Kanye and Kim disagreed, for instance) and attracted millions of views on social media. The public at large was split into white-and-gold and blue-and-black camps. So much attention was drawn, you would have thought the garment was conjured by a fairy godmother and accessorized with glass slippers.
To sort out the conundrum, the media tapped dozens of neuroscientists and psychologists for comment. Pride in our heightened relevance to society gave way to embarrassment as we realized that our scientific explanations for the color wars were not only diverse but also incomplete. Especially perplexing was the fact that people saw it differently on the same device under the same viewing conditions. This curious inconsistency suggests that The Dress is a new type of perceptual phenomenon, previously unknown to scientists.
Although some early explanations for the illusion focused on individual differences in the ocular structure of the eye, such as the patterning and function of rod and cone photoreceptor cells or the light-filtering properties internal to the eye, the most important culprit may be the brain's color-processing mechanisms. These might vary from one person to the next and can depend on prior experiences and beliefs.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
For example, people may have different assumptions about color constancy—the phenomenon that enables us to see an object's color as constant despite changing illumination sources [see “Color Contrast and Constancy,” on opposite page]. Light in the natural world typically comes either from direct golden sunlight or from the blue sky, and our perception thus assumes that most illumination has these colors. It follows that people looking at The Dress might assume the fabric is lit by either blue sky or sunlight. If the observers conclude—even unconsciously—that the source of illumination is the blue sky, their brain will helpfully subtract the blue from their perception of the image, and The Dress will appear white and gold. The brain of observers who assume The Dress is sunlit will subtract gold from the image and consequently see it as blue and black.
Before the discovery of The Dress, vision scientists had thought that people with normal vision experienced color illusions similarly. Earlier examples of ambiguous images were constrained to shape effects, such as when people see a vase or two faces in the same picture. Ambiguous shape illusions differ from The Dress in one fundamental aspect, however: whereas observers can usually flip their perception from vase to faces with little trouble, people often appear stuck in either the white-and-gold or the blue-and-black camp. Strangely, these differences can be irreconcilable. It is as if—in addition to clichéd dichotomies such as glass-half-full versus glass-half-empty or cat versus dog people—The Dress has now presented us with a new divide for humanity. Could it be that different people have different prejudices about the color of the light source? Or maybe about the type of fabric (shiny or matte) The Dress is made of?
Perhaps more than any previous perceptual observation, The Dress demonstrates that we can see the world in strikingly different ways depending on what our individual brain brings to the table. Although by and large the reason for the various interpretations of The Dress remains a scientific mystery, vision laboratories all around the world are conducting dozens of experiments to investigate the enigma. Here we offer a roundup of some of the most promising theories to date—and curious readers can visit our blog, Illusion Chasers, (https://blogs.scientificamerican.com/illusion-chasers) for new developments.
[break]
PICK A COLOR
In one of the first experiments with The Dress, Bosco Tjan of the University of Southern California and his colleagues asked students to match the blues and golds in the photograph above to color patches along a continuum of yellows and blues without naming the colors. People's perceptions fell on multiple spots along the continuum, indicating a wide variability in the way the general population sees color. Yet when the team asked the same students to name the colors, they got two main responses: blue/black and white/gold. This finding suggests that the dramatic split may have to do with how the question is posed. A different color-matching study conducted at the Massachusetts Institute of Technology by Rosa Lafer-Sousa, Katherine Hermann and one of us (Conway), involving more than 1,000 subjects, showed that perceptions fell into three main camps: white/gold, blue/black and blue/brown. CREDIT: TUMBLR/SWIKED (dress); BOSCO TJAN (color swatch frame)
[break]
DOES THIS LOOK BLACK AND BLUE TO YOU?
Famed illusion creator Akiyoshi Kitaoka of Ritsumeikan University in Kyoto summed up the problem succinctly with this simple pattern (left). If you take the colors directly out of the original picture, this is what you get. Keep in mind that The Dress actually is blue and black in real life, so those people who see it that way are not wrong. If you cannot see blue and black here, squint your eyes, view this image in dim lighting or turn down the brightness of your screen. Viewing conditions matter to how we see color. CREDIT: AKIYOSHI KITAOKA Ritsumeikan University
[break]
DRESSING UP THE ILLUSION
Demonstrations isolating the various factors that contribute to perceptions of The Dress are pouring in from vision labs everywhere. The image at the right shows that two identical pictures of The Dress look different in the context of shade versus direct sunlight. The garment's actual colors—blue and black—come across more clearly in the outdoor light. As Conway had previously predicted, our brain's color-processing mechanisms are especially powerful when we assume the sun and sky are the primary sources of light for a scene. CREDIT: ISTOCKPHOTO (woman); GETTY IMAGES (light and shade background); ROMAN ORIGINALS
[break]
ILLUMINATING THE DRESS
Neuroscientists Anya Hurlbert and her colleagues Bradley Pearce and Stacey Aston of Newcastle University in England and independently Rosa Lafer-Sousa of M.I.T. combined two light sources to photograph The Dress, as shown at the right: a diffuse cool blue light and a direct warm gold light. In the resulting image, the black regions of The Dress look golden-brown to some people, with the blue regions white or very pale blue. Others see black lace, nonetheless, along with blue cloth. This peculiar combination of light sources promotes the ambiguity that viewers experience. CREDIT: BRADLEY PEARCE AND ANYA HURLBERT Newcastle University
[break]
COLOR CONTRAST AND CONSTANCY
Visual neuroscientists R. Beau Lotto of University College London and Dale Purves of Duke University have created perhaps the most dramatic demonstration to date of color effects in the brain.
See the blue chips on the top of the left cube and the yellow chips on the top of the right cube? They are identical and appear as plain gray when the surrounding colors are removed. This phenomenon, called color contrast, causes red apples to appear redder against a background of green leaves. More generally, it makes equal colors look different because of context.
The two cubes also demonstrate color constancy, which Conway has predicted plays a role in The Dress phenomenon. Take a look at the red chips on both cubes. They are actually orange on the left and purple on the right, if you look at them in isolation. They appear more or less equally red across the images because your brain is interpreting them as red chips lit by either yellow or blue light. CREDIT: FROM WHY WE SEE WHAT WE DO: AN EMPIRICAL THEORY OF VISION, BY DALE PURVES AND R. BEAU LOT TO. SINAUER ASSOCIATES, 2003