If you are interested, I am running an experiment in conjunction with this post. Both color blind and normal vision people can participate. Click here to be taken to the survey. I really appreciate your participation and will post the results when they come in.
Update November 8, 2015: You can see the results of the survey here. We discovered some pretty interesting things! I will leave the survey open and may update results if a significant number of people take the survey.
Strap in, we’re about to delve into a question science and philosophy have found deeply fascinating for centuries. What is color?
I am in no way affiliated with EnChroma Inc. But hey EnChroma, if you want to sponsor this article just let me know!
EnChroma Inc. has developed special lenses that help the color blind people see color. We’ll get more into how they work a bit later. YouTube videos are beginning to pop up around the internet of people (mainly of men; 8% of men are color blind, compared to 0.5% of women) trying on the EnChroma glasses. And, for the first time in their lives, being able to perceive all the colors that bring our world to life. A great example is the video below of a father for the first time seeing the true color of his children’s eyes (skip to about 1:20).
“It’s like a new world”
The emotional reaction in these videos is quite astounding but not all that surprising for someone seeing a range of colors they had never perceived before. As a thought experiment try to imagine a color you have never seen. It’s impossible. But those using Enchroma glasses are experiencing entirely new colors. A once muted palette is now a literal rainbow of colors painting their experience of the universe.
Color blindness – Brief overview
The most common type of color blindness is red/green. This is further broken down by cause, to red (no red = protanopia, loss of red = protanomaly) or green (no green = deuteranopia, loss of green = deuteranomaly). Blue color blindness (no blue = tritanopia, loss of blue = tritanomaly) is more rare and has different genetic mechanisms than red/green. There’s is no way to know exactly what color blind people are seeing, but we can get a rough guess based on what we know from research and basic science. Take for example this box of Crayons.
To most humans, there are a massive range of colors to pick from. But, to color blind people, this box looks very different.
These images (above) are at the extreme ends of color blindness. More moderate and common color blindness levels would look more like the below renditions.
These examples can give us a small glimpse into what it may be like to go from color-blind to full color vision. Notice in both severe cases above there is very little purple. Speculating, I would assume the gentleman, Ethan, in the video below has a more severe case of red/green color blindness because his reaction to a simple purple container at 2:14 in this video is amazing (warning, language advisory for this video).
Color blindness commonly has a genetic cause, but it can be acquired by disease (age-related macuar degeneration, diabetes) or damage to the visual system (eye or brain trauma, UV light exposure). Color blindness is usually caused by mutations on the X chromosome. Women have two X chromosomes, one from each parent. If one chromosome has a defect the other can prevent symptoms from manifesting. Females would need to have both X chromosomes containing the mutation, greatly reducing their chance of color blindness. However, men only have one X chromosome, along with one Y chromosome. Therefore, if a defect mutation occurs on their single X chromosome, men do not have a second X chromosome to compensate, and thus men have a much higher likelihood of being color blind. As with many genetic disease, there is no single mutation for all cases of color blindness. 56 different genes have been attributed to color blindness across 19 chromosomes (the X chromosome is still the most common culprit). Before we dive into why normal seeing people can see color and why color blind people cannot see some colors, we first need to know what color is and how organisms perceive it.
So how does color and vision work?
What is color?
The color we see is due to electromagnetic radiation, or EM. Electromagnetic radiation can be thought of as a wave or particle, often called wave-particle duality. For biological systems like vision, electromagnetic radiation is generally thought of in it’s wave form (quantum physics on the other hand often uses the particle nomenclature). An electromagnetic wave consists of energy propagating through space at the speed of light as a transverse wave (seen below). A transverse wave has two perpendicular waves traveling in-synch. Electromagnetic radiation, as the name suggests, has an electric field (E, red oscillation below) and perpendicular to this, a magnetic field (B, blue oscillation below). These fields carry energy and momentum and can transfer this energy onto matter they interact with. To simplify, these fields are very similar to waves in water. If you’re in the ocean and a large wave hits you, the energy is transferred from the wave to your body, and you get knocked over.
We can measure these electromagnetic waves by how fast they oscillate, either in Hertz (called Frequency, or oscillations per second) or in wavelength (Meters, the length from the peak of one wave to the next peak). We can parse out different types of electromagnetic radiation on the entire electromagnetic spectrum by their wavelengths. For example, when the distance from one wave peak to the next is very small, 10-16 meters, we have gamma rays, the smallest wavelength of EM radiation. Slightly longer, 10-10 meters is the well known X-ray. Longer still is UV light, then infrared, microwaves, FM and AM radio waves, and finally long radio waves.
A very small segment of the EM spectrum, termed the visible spectrum, roughly 390 (violet) – 710 (red) nanometers (nm), is what humans are able to perceive. Other organisms can see other EM waves. Our cats, dogs, and hedgehogs can see UV light, and insects use their UV vision to spot flowers. On the other end of the spectrum, snakes can see in the infrared range, which is akin to seeing heat. Warm things, such as a mouse or human or stove top, give off infrared electromagnetic radiation. There are also some ridiculously complex eyes, such as those of the mantis shrimp, which can see 11 wavelengths from UV, visible, and infrared light as well as polarized light.
How do we detect colors?
Humans, and other organisms, are able to detect electromagnetic radiation due to the specialized cells in their eyes. Focused light coming through the pupil falls on the retina (depicted in yellow on the left diagram). The retina is made of a different types of cells and neurons. The cells that detect light are called photoreceptors and consist of two main categories: rods (brightness detectors) and cones (color detectors). Rods do not detect color per se, rather they detect how bright the world is in a sort of green-blue hue. Rod cells are also very sensitive to light and are mainly responsible for our night vision. Rods can be found throughout the retina, but very few of them are in the fovea of the retina. This is because the fovea is densely packed with our color detecting cones. The fovea is where light is primarily focused and is responsible for our sharp central vision. When you focus on these letters, they are falling on the fovea and come up very sharp. However, light does still fall onto other parts of our eyes but these parts lack the visual acuity of the fovea. Focus on these letters and try to read the title of this section. Its nearly impossible! But you can still see there is some bold black text against a white background. The acuity of the fovea is due to how densely packed the cone photoreceptors are. You can think of this as a higher resolution camera. The sharp detection from the fovea is similar to a very high megapixel camera which produces crisp pictures. In contrast, the detection capabilities outside the fovea is like a grainy webcam, or an old pin hole camera.
The densely packed cone cells in the fovea give us detailed color vision due to specialized proteins within the cone part of the cell. These proteins are called Photopsins or just Opsins. Humans posses three different opsins for color vision, refereed to as S, M, and L for the wavelengths they detect. Short wavelengths (blues) are detected by the S-opsin which respond most to electromagnetic radiation with a wavelength around 430nm. The Medium and Long opsins detect green (530nm) and red (560nm) light, respectively. One cone only has one type of opsin protein. When the energy carried by the electromagnetic wave is transferred to the opsin protein in a cone cell (like the ocean wave hitting you), the protein changes shape (called a Conformational change). This change in shape ultimately results in ion channels of the cell opening and charged ions rushing into the cell and increasing the voltage of the cell. This electrical signal is converted to a chemical signal of neurotransmitters at synapses. These neurotransmitters convey to neurons connected to the cone cell that colored light has been detected. This signal eventually goes through the retina, travels down the optic nerve, and ends up in our brain.
Color detection and color blindness
In normal color vision the three opsin proteins of our cone cells work by detecting blue, red, and green wavelengths of light. If we are looking at a purely blue screen, only our blue cones are responding. This is because of the specific amino acids and the order they are put in make up the distinct opsin proteins. The unique shape of each opsin protein is what allows it to respond to a unique wavelength. Amino acids are what build proteins and they are coded by our DNA; you can think of them like different building materials. By changing just a few amino acids, the shape of an opsin protein can change a little too. This change in the protein will change what wavelengths of light it can respond to. So, in theory, each opsin will be similar (similar enough to all detect color) but each a little different in shape and composition from each other; allowing them to respond to different wavelengths of light.
To better understand this, think back to our ocean analogy. Small ocean waves (a foot or so high) mean very little to an adult, they can just walk through them unimpeded. But to a child, a different sized human, these waves are meaningful and can result in endless fun. Medium sized ocean waves (2-3 feet high) are great for adults and adults are very responsive to them, body surfing or jumping around in. Children are rarely effected by these medium waves because they are not allowed to go out into the ocean and experience them. Large ocean waves (6 feet+) are ignored by children and most adults, as they’re just too dangerous. But to a subset of humans, like skilled surfers, these are the waves they love the most. So we have children responding to Small waves, adults responding to Medium waves, and surfers responding to Large waves. They’re all still humans, but subtle differences makes them respond to waves uniquely. Just like the three S, M, and L opsins are all still opsin proteins and are similar in amino acid composition and shape, little differences make them respond to different wavelengths of light. Below you can see the wave lengths of light that the three opsins respond most to. The higher the Y-axis value is, the more responsive a cone is to that wave length.
Cones and color blindness
However, in color blind individuals, these proteins are misshapen due to mutations in the DNA that contains the blueprints for these opsin proteins. In severe cases of color blindness, one (or more) of the cone cell types is totally dysfunctional or missing. This type of color blindness is called Dichromacy, meaning two colors. Dichromats (people with dichromacy) can only match colors of the world with variations of the two colors they can detect. Trichromats (normal color vision) can mix all three cone responsive colors (blue-green-red) when perceiving electromagnetic radiation. The “severe” crayon box examples given at the top of the page would be similar to what dichromats see. By being unable to perceive and mix one out of the three normal colors, the spectrum color blind people see is severely constrained. Roughly 2% of the male population is lacking the L-opsin (red) or M-opsin (green), resulting in severe red-green color blindness. Less than 1% lacks the S-opsin (blue), severe blue color blindness.
The more common form of color blindness is called Anomalous Trichromacy (meaning deviation from the standard trichromacy). Rather than having a completely dysfunctional or missing opsin protein, people with anomalous trichromacy can detect all three colors but the affected color is distorted, to varying degrees. For example, Deuteranomaly, or green-weak, is the most common type of color blindness (6% of males, 0.4% females) and results from a defect in the M-opsin (green). Deuteranomaly is carried on the X chromosome hence why males are much more likely to be affected. Green weak vision is likely similar to the “red/green color blind” example at the top of the page. People affected by green weak color blindness have a hard time discriminating between long wavelengths of light (green-yellow-orange-red). This is due to their M-opsin (green) being shifted in sensitivity to the longer, more red end of the color spectrum due to its misshapen form. Rather than responding a lot to green, it responds more to orange. This would be like an average adult trying to learn to surf on huge waves. They would do alright and probably not drown because they can swim, but they would likely be very unsuccessful at actually surfing. Below is an example of the normal color spectrum and the spectrum from the most severe cases of red/green color blindness.
Other, more rare types of anomalous trichromacy also exist. People with protanomaly have a mutated L-opsin (red) protein that is more sensitive to a greener, shorter wavelength. People with protanomaly have a hard time seeing reds and discriminating colors across the entire normal visual spectrum. Finally, tritanomaly is an extremely rare (0.01% male and female) form of color blindness where the S-opsin (blue) is more sensitive to longer wavelengths (closer to green). Tritanomaly is due to a mutation on chromosome 7 so males and females are equally affected.
Therefore, people with color blindness have difficulty seeing colors because the color spectrum they can perceive is constrained and lacks distinct separation of wavelength detection. Their green sensitive (M) protein is a little bit different so that it responds more to longer, redder wavelengths rather than to green. Or their blue sensitive (S) protein are a bit misshapen and like more green light rather than blue. This is similar to a small child trying to swim out and play in 4 foot high waves. Or the red sensitive (L) is shifted to respond more to green. But how do Enchroma glasses fix this?
How do Enchroma glasses work?
Below, you can take another look at the normal (left) three color sensitive cones, Blue (S), Green (M), and Red (L) and how much respond (Y-Axis) to different wavelengths (X-Axis). The graph on the right is an example of a mutated green (M-opsin) cone which is mutated such that it’s response is shifted and it now responds to redder, longer wavelengths; the most common type of color blindness and the main target for Enchroma.
Because the green cones are now less responsive to medium length electromagnetic radiation, people with this form of color blindness have a lack of green discrimination. But! in a majority of color blind people, their defective cone is rarely shifted so much that is overlaps with the red cone’s color response curve 100%. Rather, they only overlap some, like seen in the graph above. Enchroma is able to exploit this non-complete overlap by filtering out the wavelengths of light which activate the green and red cones indiscriminately.
As an important aside, the brain mixes color by additive color mixing, so similar amounts of red and green make yellow. Red and blue make purple/magenta. This is how pixels on your phone and TV work.
Lets think through how this would work. When light around 570nm enters the eye, the red cones would normally be activated much more than the green cones. That is there would be a big difference in how much red cones were activated versus how much green cones were activated. Because of this, we would see red (probably more orange at 570nm but lets keep it simple). But, when the green cone opsin is mutated and thus shifted to be responsive to longer wavelengths, it is activated more by 570nm light than it should be. Second to this, rather than the red cone responding much more than the green cone, the two cones respond to a similar degree. Because the green and red cones are responding by a similar amount to 570nm light, the brain thinks it is seeing both green and red at equal levels and interprets that as a brownish-yellow color instead of just red.
The problem: Color sensitive cones are not differentially responding to some wavelengths of light. Rather, distinct cone types are erroneously being activated together.
Enchroma’s solution: Filter out the wavelengths that the cones do not differentially respond to. Thus, distinct cone types will not have so much erroneous activation overlap.
The above graph shows the cone response curves of normal color vision (not sure why Enchroma used normal vision in this graph). Overlaid in white and grey is the Enchroma filter where wavelengths that fall in the white area are allowed through the lens whereas wavelengths in the grey area are filtered out and thus do not reach the eye. The wavelengths of light that the lens filters are those in which the greatest overlap between any two cones would be. For example, light from 550nm to 570nm is filtered out by Enchroma lenses. This range corresponds to the most problematic wavelength range for red-green color blind folks because either their green (M) cones are shifted up from a 530nm peak to the longer, more red end or, red (L) cones are shifted down from 560nm to a shorter, more green wave length. Rather, the only light that gets through the lenses are of wavelengths where the least overlap in cone response happens. You can see this by the big gap between the three curves in the white area of the graph, ie for green and red cones from 570nm-650nm the distance from the green to red curve is large.
To summarize, Enchroma lenses filter out wavelengths of light that commonly confuse the mutated cone cells in color blind individuals and cause the retina to send indiscriminate color signals to the brain. By filtering out these troublesome wavelengths of light, color blind people’s brains can better discriminate colors from the visible spectrum because the particular noisy, confusing wavelength-color signals are no longer coming in. Almost like tuning the radio to pick up the music better and cut out the static (indiscriminate signals).
“There’s some differentiation in color that pop out.”
But, what is color actually?
Before I described colors by different wavelengths of electromagnetic radiation. But this is only sort of true. In physics and EM radiation there is no such thing as color. There is nothing special about 450nm that makes it blue. The wavelength distance of 600nm is not inherently red. Rather, our brains create all the color we see in the world. Lets start with the eye and work our way to the brain to better understand how color is processed.
Around 542 million years ago the Cambrian explosion occurred. It is refereed to as an “explosion” because of the rapid evolution observed during this time that gave rise to a massive and diverse range of living organisms. Slightly before this time we see the first fossil evidence of the eye-like structures. Eyes started out as just a bundle of cells, called an Eyespot, responsive only to brightness (like rod cells). These eye spots likely played an important role in an organism’s day and night cycle of activity, or circadian rhythm. Eyes may even be a cause of the Cambrian explosion due to the advantages organisms with eyes enjoyed. Over time, eyespots curved, that allowed organisms to figure out which direction light was coming from. Likely, a pinhole formed next allowing higher direction discrimination, followed by a lens for better acuity, and finally an iris to control brightness and a cornea for even better focus.
So, eyes evolve which are sensitive to light (electromagnetic radiation). However as we learned about above, we know there is a wide range of electromagnetic radiation. Why are most organism’s eyes sensitive to wavelengths around 400-700nm? Well, this is due to the sun and earth’s atmosphere! If we measure the wavelengths of the electromagnetic radiation coming from the sun, we can observe that most of the radiation comes in the human visible spectrum, from 400-700nm. There is still quite a bit of energy in the infrared range and also in the ultraviolet range.
Not only is the human visible spectrum correlated with the strongest wavelengths the sun generates, Earth’s atmosphere also filters out electromagnetic radiation from the sun and space outside of the visible spectrum. The atmosphere absorbs high energy gamma and X-rays protecting life form the harm it causes to DNA. Most wavelengths of high energy ultraviolet light are also absorbed but some longer UV wavelengths do pass through. Visible light, however, is in no way filtered and can pass through the atmosphere unimpeded. This is called the optical window.
Thus, the earth, and the life on it, receive a massive amount light in the 400-700nm wavelength range. So, if eyes were to evolve, the most readily available electromagnetic radiation to respond and adapt to would be from 400-700nm. What we call the visible spectrum is really nothing special, it is simply the most common spectrum of higher energy EM radiation available to earth. For those wondering about that Radio Window, radio waves are so massive (long wavelength) and contain such little energy, developing biological methods of detecting them would be very difficult.
So we have eyes. We know why we have eyes that respond to the wavelengths that they do, but how do we get eyes that respond to different wavelengths of light?
As we saw from color blindness, we know how just a simple tweak to DNA can cause a light sensitive protein to respond to a different wavelength of light. As the broad EM spectrum responsive opsin protein responsible for eyespots was fine tuned from natural selection pressures, it is likely a few simple mutations resulted in multiple opsin proteins. A likely mutation type to cause this is a simple duplication error.
The gene that encoded the brightness sensitive opsin may have been duplicated. Then, the duplicated gene could be fine tuned to be responsive to say 450nm light, which very common in the ocean which is where complex life has thought to evolved (water absorbs longer wavelengths of light). The resulting organism would have the original brightness detection opsin as well as be able to see longer wavelengths of light. You can easily see how this type of mutation could result in multiple genes for multiple opsin proteins that respond differentially to various wavelengths of light, such as long 560nm, medium 530nm, and short 430nm. But where does the color come from?
The Neuroscience of the Visual System
I personally find anatomy extremely boring and dry so here is my attempt to make it bearable (and immensely simplified). Oh, and enjoy a beautiful picture of the retina by Santiago Ramón y Cajal, the father of modern day neuroscience.
Lets start with our three cone cells, each cone likes one of the three different wavelengths of light; short, medium, and long. The cones are on the prowl for any bit of information pertaining to their wavelength. When their favorite wavelength of light enters the club (eye) and touches them, they get excited and tell all their neighbors what just happened. This excitation is carried via nerve pulses, or action potentials, down the optic nerve to the relay station of the brain, the thalamus. In the thalamus, the optic nerve starts chatting with the Lateral Geniculate Nucleus (LGN). The LGN is the vision gossip queen; she knows about all the activity going on both in the eye and what the brain is currently “seeing”. To sort all this information, the LGN has 6 distinct layers that only care about specific gossip. Magnocellular cells in layers 1 and 2 only listen to brightness, depth, and movement chatter. Parvocellular cells in layers 3, 4, 5, and 6 like the nitty gritty details of the visual world as well as the hot long (and medium) wavelengths. Between each of these 6 main layers are smaller layers of Koniocellular cells, they only care about B-list celebrities, also known as shorter wavelengths of light. After all the visual gossip is organized, the LGN sends the gossip to the primary visual cortex. The visual cortex is broken into subregions, where V1 is the start of vision information processing and processing complexity increases as you move away from V1 to the higher numbered visual areas.
This is where wavelength processing starts to get tricky and clean segregation of information from different wavelengths of light largely breaks down. For example, if it is bright, a cell may respond to long wavelengths. But if it is dark out it may then respond to medium wavelengths. However, a small subset of V1 neurons that receive information from the LGN are sensitive to specific cone wavelengths. There are 2 groups of wavelength sensitive cells and neuroscientists refer to them as Blobs. Blobs work by figuring out the amount of difference between two wavelengths. There are long-medium blobs that care about the relative amounts of long and medium wavelengths being seen in the world and there are short-long blobs that care about the different between short wavelengths and longer ones. After blobs figure out relative differences, this information, along with a massive amount of more complex visual information (brightness, lines, etc) is sent out to higher order processing parts of the visual cortex; V2 and beyond.
A particularly interesting set of cells, called Globs, are found in visual cortex area 4. These globs respond to a specific combination of wavelengths. One glob may respond most to the result of 402, 517, and 604nm wavelengths. It’s neighbor glob is similar but a little different, maybe most responsive to 402, 512, and 609nm. Across all of V4 there will be globs that respond to varied levels of the combination of all three perceivable short, medium, and long wavelengths of light. If the V4 is damaged, say in a car accident, the person will lose their ability to perceive all different wavelengths of light and their world becomes black and white.
But wait, you missed the color part?
Well, there still is no color. Color will not be found in the anatomy and physiology of the visual system. Color is not a tangible thing we can just point at and say “well there it is, that is where our brains make it up”. Rather, color is how our brain’s visual system interprets electromagnetic radiation for our higher order brain regions that are likely responsible for our conscious experience of the world. This is similar to sound and music being how our brains interpret pressure waves. And heat is our brain’s interpretation of quickly vibrating particles. And sweet is our brains interpretation of sugar molecules binding to taste receptors. And vanilla is the brains rendition of odor molecules.
When you shine a 700nm wavelength light into someone’s eye, short wavelength cones don’t respond much. A few medium wavelength cones may fire. A massive amount of long wavelength cones will discharge. From those resulting electrochemical signals, our brains create what we know in our conscious mind to be red. The brain has no use for red or color at all in signal processing. The visual cortex just knows a specific series of 1’s and 0’s (action potentials or no action potentials) originated from the retina.
So what makes color, color? What makes red, red? Why do some electrochemical signals make us see red and others green?
Philosophers and scientists have been trying to answer this question for a very long time. Erwin Schrödinger, the famous physicist and Schrodinger cat guy, wrote in his book titled What is Life? With Mind and Matter and Autobiographical Sketches:
The sensation of color cannot be accounted for by the physicist’s objective picture of light-waves. Could the physiologist account for it, if he had fuller knowledge than he has of the processes in the retina and the nervous processes set up by them in the optical nerve bundles and in the brain? I do not think so.
Feelings and experiences vary widely. For example, I run my fingers over sandpaper, smell a skunk, feel a sharp pain in my finger, seem to see bright purple, become extremely angry. In each of these cases, I am the subject of a mental state with a very distinctive subjective character. There is something it is like for me to undergo each state, some phenomenology that it has.1
Qualia is often applied to the problem of color. A famous example put forth in a paper by Frank Jackson is of Mary, the brilliant color scientist. Mary is born in a black and white room and cannot leave. Mary decided to learn everything there is to know about color. Her captors provide her with books and black and white videos. She learns about electromagnetic radiation, and wavelengths, and how cones work, and how the visual system works. Eventually she learns all there is to know about color. Then she is allowed out of the room and looks upon the green grass and exclaims, “this is what it is like to experience green”. Even if you knew everything there was to know about the color green, there is still more information that can only be obtained by the experience of green. Rather, color is not a thing out in the world and is rather a manifestation of conscious experience as our brain interprets electromagnetic radiation. Though qualia is useful for thinking about color, this does not get us much closer to an answer as to what causes the brain to create color.
Does language cause our brains to create color?
Many of the ideas in this section directly come from or are inspired by the amazing Radiolab episode Colors. This whole blog post originated from my idea to merge the concepts put forth by the Colors Radiolab episode with the effects of Enchroma glasses.
The Island of the Colorblind is Oliver Sacks’ account of his visit a small yet peculiar Pacific atoll, Pingelap, in which many residents are completely color blind due to a rare genetic mutation. Due to generations of inbreeding on the small island, the complete color blindness mutation became prevalent. Despite not being able to see color, the islander’s language has a wide variety of words to describe pattern, tone, luminance, and shadow bringing a particular richness to the pale world they live in. How much of an influence does language have on how our brain interprets sensory signals, say from our eyes?
Well linguist Dr. Guy Deutscher has a lot to say about language and how it shapes the colors people see. So much so he wrote a book on it, “Through the Language Glass“. To summarize the findings of Dr. Deutscher, William Gladstone (yes THE William Gladstone, four time prime minister of England) was a huge fanboy of Homer and his works. Gladstone produced a study on Homer’s work and while doing so noticed something odd about the way Homer described the color of different things. The face of a man pale with fear, but also honey, were referred to as green. The sea and oxen were “wine” colored. And both Iron and sheep were violet. All in all the Illiad and Odyessy together contain the word “black” 170 times and “white” 100 times. However, quite surprisingly, “red” only appears 13 times. “Yellow” and “green” are both used less than 10, and “blue” is no where to be found. Not a single mention.
Following Gladstone’s fascinating work, Lazarus Gieger, a German Philologist and Philosopher, analyzed many different ancient texts from all around the world including Icelandic Sagas, Chinese Handscrolls, the original Hebrew text of the Bible, and Vedic Hymns. And throughout all these texts, from all these cultures, scattered all across the globe, a pattern emerged. Words for black/dark and white/light appeared in the earliest of written records. As time progressed, our modern rainbow slowly emerged as these separated cultures began writing about reds. After red, yellows began to light up the ancient pages and parchment. Sprouts of green followed yellow suns yet oceans of blue never flowed across ancient scrolls. It wasn’t until relatively recently blue things came into being. There was one exception to the ancient lack of blue, the Egyptians, who famously produced a blue dye thought to be the first synthetic dye ever synthesized. Even though our skies are an endless void of blue and blue oceans extend to the infinity of the horizon, the majority of ancient civilizations failed to ever comment on them as being blue. This is not due to a lack of content either. All these texts refer to the sky and ocean, but they are never called blue (often just light). We also know ancient people were not biologically color blind. The three color cones found in humans and discussed above evolved well before homo sapiens came to roam the earth; Catarrhini primates and many New World female monkeys have the same three color cones. So why is it the color that surrounds us the most is lost to these ancient civilizations? Well Dr. Deutscher formulated a little experiment to test just this.
Dr. Deutscher, along with his wife, taught their newborn baby girl all the colors of the rainbow as good parents do. Bananas were yellow. Grass was green. Firetrucks were red. Blue blocks were blue. And as their little girl grew a bit older, she could start to distinguish all the different colors. They could point to a blueberry and ask “what color is that” and their daughter would repeat back “boo” (close enough to blue when you’re just learning to use language). However during all of this, Dr. Deutscher and his wife never told their daughter what the color of the sky was. Not once. They never even acknowledged the sky was a thing. So, once their daughter could distinguish colors consistently, Dr. Deutscher took his young daughter outside for a walk on a cloudless day and, as he had done hundreds of times before, pointed and asked “what color is that?” However this time, he pointed up into the sky. She looked up quizzically. And after an unusually long while, said nothing. She just couldn’t answer him. This continued on for months as Dr. Deutscher would take his daughter out on cloudless days and ask the color of sky and his daughter would not answer. However, one day, she did answer. But with “White”. For another two months or so, every time the question was asked, she’d say “white”, until one day when she hesitantly said “blue”. Blue didn’t last though, and for a few more months she would switch back and forth between answering with “blue” or “white”. Eventually she did settle on the sky being blue. To Dr. Deutscher this was fairly clear evidence that the blueness of the sky may not be as obvious as we had thought (even with the language to describe its color) and that ancient civilizations likely could not perceive the blue of the skies and oceans.
Now, ancient people could see blue. That is, they had blue cones in their retinas, these blue cones were functioning, they were stimulated by short wavelengths and sent signals to the brain, but these ancient people did not perceive blue in their conscious mind. This may seem sort of odd, but think of the last time you lost something and looked all over only to find it in plain sight. Light was reflecting off that “missing” object, landing on your eye, and ending up in your visual cortex, likely multiple times. But, despite “seeing” the lost thing, it did not come into your consciousness. Also think of the clothes you are wearing now. Prior to reading that last sentence the feel of your clothes was not in your conscious attendance of the world, but that information was still making it into your brain. Another example may be when you learn a new idea or concept and it shapes the way in which you see the world. For example, once you finally find Waldo on a page, its hard to not see him over and over again. (Hover over the image to see Waldo)
This may be like what language does with color. Without language it is very difficult to perceive differences in color and categorize these differences. Without being told what color things of the world are and building up a palette of colors to perceive with, the world is pallid. A theory Radiolab proposes is having words for color in our language gives a means for different colors to be distinguished. As we distinguish between colors it feeds back onto our perception, and we begin to perceive differences in color. This perceptual change feeds forward to the language categories and separates colors further. This loop continues as our language develops and we gain experience with colors. Eventually the world we perceive is filled with a wide variety of vivid color.
You may still think this is ridiculous. How could we ever know what ancient people saw or didn’t see? Until a time machine is invented we can’t ask them. What if they could perceive blue and just never wrote about it? But what if there were primitive people around today still developing in language and technology? Could we just ask them if they can perceive blue or not?
The Himba People of northern Namibia are the perfect candidates for such a question. Professor Jules Davidoff and Debi Roberson published a variety of studies around 2003 which were the result of their field work with the Himba and the influence the Himba language had on their color perception. Unlike English, which has roughly 11 basic color terms (black, white, grey, purple, blue, green, yellow, orange, red, pink, brown), the Himba only have 5 color terms (Roberson et al., 2004). For the experiment, a variety of colored tiles (below) were placed in front of Himba participants and they had to separate the ones that were different and group the ones that were similar and name them. This was repeated in English speakers from the UK. Below you can see the differences in how the Himba and English speakers differentiated the same colors. The Himba refer to red/purple/pink as “Serandu”, yellow/beige (and the color of white people skin) as “Dumbu”, black (but also blue and green, and dark color shades) as “Zoozu”, white (as well as a wide range of light colors) as being “Vapa”, and blue/green/purple all as “Burou”. However, the Himba have a lot of modifiers for shades of green, just like we have forest-green or neon-green. The color terms and groups are overlaid with the tiles.
The results of this study provide us with evidence that color may not be universal but rather color is relative and likely influenced by language. In a study they published a year earlier, Davidoff, Roberson, and Shaprio (2002) demonstrated that the ability to differentiate colors correlated with the development of language as both Himba and English children learned more and more words to color their world with. However, there was large individual variation in color order acquisition (for example not all children, Himba or English, learned red first) and this again may be due to what words children learn rather than the frequency of colors they are seeing, ie blue sky and ocean. One last experiment (not published, not peer reviewed, so take with a grain of salt) made famous by a BBC show “Do you see what I see“, Davidoff had Himba people look at colored squares. One of the squares was a different color from the others and the participants had to choose which one was different. Davidoff measured how long it took for the Himba to make a choice and their accuracy. For the Himba, picking out different shades of green was very easy and they could respond quickly (left image, below). This may be due to all the modifying words they have for different greens. The average English speaker may have more difficulty finding the green square as it does not immediately pop out. This possibly due to our lack of words and needs for different greens.
However, when the different color was a blue (right image, above), the Himba had a hard time responding and it took quite a while for them to make a choice. This difficulty may be due to the Himba having no word for blue at all and thus finding it difficult to pick out the color specified by that word. Since we have a word for blue it was likely easier for you see to distinguish the blue square from the other green ones. For a review of these studies and related work, see Roberson et al., 2006. On the other side of the world, a similar experiment was carried out in Russia by Winawer et al., 2007. Russians have two separate words for dark blue (siniy) and light blue (goluboy), similar to English red and pink. Russians were quicker to discriminate across shades of dark and light blues compared to English speakers.
From these studies I think it would be fair to say language, at the least, plays an important role in painting the wide variety of color that fills our perceived world. Language provides our brain with a framework, or schema, to organize information. By organizing different sensory stimuli, such as various wavelengths of electromagnetic radiation, language allows us to better distinguish across stimuli that are different, like red and green, or similar, say a leaf-green to a green apple. Without the scaffolding language provides, it may be very difficult to naturally discover and most importantly retain perceived differences in stimuli. The word “green” gives our brains a way to bind together the electrochemical signal 550nm light generates with a concept of green. Without the concept that green exists, the signal may be unattended to by our conscious mind. The Egyptians had a word for blue because they were able to create blue. However, no other ancient civilization could make blue, so language did not develop for blue, and blue wavelengths were lost in wine colored seas.
Update: Dr. Carrier’s evidence for ancient blue
Dr. Richard Carrier, a world renown philosopher and historian with particular expertise in ancient Greek and Roman philosophy, science and technology, has written a rebuttal to the Buisness Insider piece based heavily on Radiolab’s Color episode. While I encourage you to read Dr. Carrier’s article, he essentially points out, and with a lot of convincing evidence, that ancient cultures, and not just Egyptians, had many words for the many different shades of blue and had access to many things which were blue, such the blue cornflower, blue paints, and sapphires. Dr. Carrier simply is not buying that ancient people could not perceive blue, or any color for that matter. Rather, they could perceive all the colors modern humans do today and the historical records prove it; Homer like was indeed color blind and other sources should have been sought out to confirm or deny his color descriptions.
Enchroma glasses and their implications for what we know about color
To conclude I’d like to bring together all we’ve learned about color: Electromagnetic radiation, the neuroscience of color and vision, the mechanisms of colorblindness, and how it may be language influences our perception of color, with what Enchroma glasses and, most importantly, the experience of color the colorblind people have when trying them on.
If anyone has actually made it all the way down here take a break and watch another heart warming Enchroma video (skip to 1:00). Language warning.
“That’s a lot different than what I thought [blue] was. I guess I gotta re-learn my colors or somethin’ cause that’s a weird fuckin’ color”
Enchroma and Youtube are giving us an interesting experimental paradigm. Colorblind people using Enchroma glasses have all the language for color available to them as normal sighted people do. However, their experience of these colors had always been constrained. With our new knowledge lets dissect what’s happening when color blind people try Enchroma glasses.
All the men in these videos have acquired language to a high degree. Their brains know every word for every color. However, their brains never received the full range of sensory signals caused by the visible electromagnetic spectrum. Rather, due to their cone cells, they’ve only ever experienced a constrained palette of colors and thus a constrained diversity of color signaling to their brain. Enchroma glasses are now expanding the range of possible signals sent from the retina to the brain for interpretation by blocking out the signals that underlie problematic, erroneously overlapping color information. Enchromatic’s brains are now tasked with interpreting and sorting all these new signals into their different linguistic categories. In the video above, the signal his brain had always associated with “blue” is now different. He will need to re-learn what it is like to perceive blue in order to associate into his brain’s framework of what the experience of blue is. Enchroma wearers will need to relearn what it is colors are like to experience and modify their framework of what colors fit with what words.
Second, using the report of the man in the third video, rather than seeing the leaves of trees all as simply “green”, a diversity of green signals are now being sent to his brain. Before his brain had one large category of green because so much of his visual perception of medium wavelengths was a glob of the same green. His retinas could only send a limited range of green signals. Now, a wide array of distinct green, and likely some yellow, signals are available to his brain. These new colors, hues, and shades will need to be integrated into his green schema concurrently with expanding the possibilities of what it is to perceive green.
Despite needing to re-learn what colors are and expand the range of what colors can be, many people using the Enchroma glass are quite good at guessing colors. If we think back to Ethan in the second video exclaiming colors as he picks up various things, he never gets them wrong. In a follow up video Ethan answers a question (and many others) that may help us understand this, “How did he know the names of colors if he was color blind?” to which Ethan responds,
“I’ve learned the colors when I was a little kid and I knew what they were… I just didn’t actually see them. I know what green is, I know what yellow is, sometimes the shades in-between I get really lost on, I don’t know which one is which. For instance, purple, which is now my favorite color, and before I understood what it was, and I’ve always seen blue and red individually, the best way I can describe with purple is I can see them both at the same time, I can see blue and red all at once.”
and to another question,
“purple always looked blue, navy blue to me”
Ethan and his new love of purple gives us an insight into what his brain was doing and how it is now processing the resultant purple into his color understanding of the world. He had known prior that purple existed and he knew mixing red and blue made purple, but despite this knowledge he had never before experienced what purple was like. Quick reminder, for our brains to create purple, red and blue cones need to be activated and signal together while green cones are quiet. However, as we learned about red/green color blind people, the green and red cones are activated together more so than they should be due to the green or red cone shifting in wavelength sensitivity. So, Ethan’s brain never before experienced only blue and red signals, there was always some green making the signal noisy. Now that Enchroma is blocking the pesky green-red overlap, Ethan’s brain finally receives only blue and red together and into his brain’s conscious recreation of the world pops purple.
And this leads us to our final questions. Was it only because Ethan knew about purple and had a schema of information in his brain devoted to purple, did his brain make purple? If Ethan never knew purple existed, would his brain still have created purple from the signals it received by blue and red cones? Would his brain create purple but never bring it to his conscious attention without the idea of purple? Or, if without a word for purple, could it be the red-blue signal would be lost just as blue was lost to ancient civilizations? And finally, what colors are there that only color blind people can see?
What do you think? Is the rainbow of color universal to the human brain or is it a construct of language? Let me know in a reply
Thinking you may be color blind after reading this? Take EnChroma’s color blind test.
If you’re still interested in color, check out Vsauce’s video “Is your red the same as my red?”
All figures and graphs by Enchroma are property and copyright of Enchroma Inc.
- Michael Tye, “Qualia”, The Stanford Encyclopedia of Philosophy (Fall 2015 Edition), Edward N. Zalta (ed.), forthcoming URL = <http://plato.stanford.edu/archives/fall2015/entries/qualia/>
- Debi Roberson, Jules Davidoff, Ian R L Davies, and Laura R Shapiro. 2004. “Color categories: Evidence for cultural relativity hypothesis.” Cognitive Psychology, 50(4), 378-411. doi:10.1016/j.cogpsych.2004.10.001
- Jules Davidoff, Debi Roberson, and Laura R Shapiro. 2002 “The development of color categories in two languages: a longitudinal study.” Journal of Experimental Psychology General, 133(4), 554-571. doi:10.1037/0096-34188.8.131.524
- Debi Roberson, Jules Davidoff, Ian R L Davies, and Laura R Shapiro. 2006. “Colour categories and category acquisition in Himba and English” In: Nicola Pitchford and Carole P. Biggam, eds. Progress in Colour Studies. Amsterdam: John Benjamins Publishing Company, pp. 159-172. ISBN 9027232407 [Book Section].
- Jonathon Winawer, Nathan Witthoft, Michael C Frank, Lisa Wu, Alex R Wade, and Lera Boroditsky. 2007. Russian blues reveal effects of language and color discrimination. Proceedings of the National Academy of Sciences, 104 (19) 7780 – 7785. doi: 10.1073/pnas.0701644104
- Himba/English color language charts from Antonio Šiber at http://www.antoniosiber.org/cijanoza_en.html
- http://www.color-blindness.com/ was an immense help and wealth of information
- And of course Wikipedia
What is color? Enchroma glasses, neuroscience, and the mystery of color by Blake Porter is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
If you enjoyed this article and found it informative, consider donating! Funds first go to keeping the website running (domain name/hosting) and secondly go to me. Mainly for grad school things, so food and beer. Donations are secure and handled by PayPal.