Wednesday, May 15, 2013

Is your green the same as my green?

Here is a common question when I give a class on color theory: "Do we all see color the same way?"

Often the question starts with the words "My wife...", or "My husband..."  There is, at the very least, a perception that men and women see color differently. But what about me and some other guy? Do we see color the same?

There are lots of answers to this question: How much variability is there is color perception? That makes it fun to blog about!

High level evidence of gender differences

There is a well known anecdote that if you ask a guy about his car, he will give a lengthy discourse about the engine. A woman will tell you the color. If you can get a guy to mention the color of his car at all, he will use basic color names like "green" or "blue". A woman will use a nuanced color name like cerulean or mauve to describe her car.

My wife and I, discussing automotive colors

I am not aware of any experimental evidence of gender bias in car automotive descriptive preference, but I know of a brilliant study from a brilliant color researcher [1]. He provided some evidence that somewhere between a photon entering the eye and a neuron recalling a word, there is some difference in the way men and women perceive color. This incredibly smart and good-looking color scientist asked 50 people to write down all the single word color names that they could think of in two minutes. Here is his conclusion:

"There was a statistically significant difference between the number of colors that the men could recall versus the number that the women could recall. Men averaged 15, and women averaged 18."

(A bit more on this groundbreaking experiment can be found in the section "Eleven" in the blog post "How many colors are in your rainbow?")

Where does this happen? Is this in the physiology of the eyes? Or maybe it's a brain thing -- woman are, on average, better at verbal skills than men? Or is it a cultural bias because women in general are encouraged to be interested in fashion?

Let's consider the differences in physiology...

Color blindness

Color blindness is one explanation I have heard for why there is a gender difference in color perception. Men are more likely to be colorblind than women. Dogs are colorblind. Therefore men are more likely to be dogs than women.
Comparison of how I see apples, and how my colorblind dog sees apples

It's true that women are less likely to be colorblind. Color blindness is a sex-linked trait [2]. The weakness is something expressed on the X chromosome. Basically, you need to have one copy of the color vision genes on either chromosome to have full color vision. Since women have two X chromosomes, they have two chances to get all the pieces for color vision. Men, who only have one X chromosome, only have one shot. As a result, color blindness is more prevalent in men than in women. About one man in a dozen is colorblind. For women, it's about one in 200. [3]

As I said, this is true, but it's not an explanation for why there is a gender difference in the color naming experiment. Even with the elimination of any possible colorblind men from the experiment, the results were still statistically significant. And what about the 11 out of 12 men with normal color vision who still prefer to talk horsepower and miles per gallon?


As I described in Organizing your crayons, color is three-dimensional. The reason for that is simple. We have three types of cones (light receptors) in our eye. Each cone responds to a different range of wavelengths. These three cones are called "L" (which stands for red), "M" (which stands for green), and "S" (which stands for blue). [4]

There is some controversy on this, but there are reportedly a number of people who have four different types of cones. To them, color is four-dimensional. 

Just to give a feel for what that means, it is a rough approximation that each cone allows us to distinguish about a hundred levels. For a combination of two cones, it is possible to distinguish 100 X 100 different colors. (For those who don't have a calculator handy, that number is 10,000, or ten thousand.) This is a rough idea of the number of colors that a dichromat (someone who is color blind with two types of cones) can see. 

For most of the population, there are three types of cones, so there are 100 X 100 X 100 different combinations that could be distinguished; one million perceptible shades of color.

A tetrachromat has four different types of cones. As a result, there are one hundred times as many different shades of color that this person could distinguish. One hundred million colors. Such a person could sense subtle differences in hue that I could only see with the help of LSD. Imagine how long it would take such a person to match their socks in the morning! [5]

Some folk's eyes are just more colorful than others

This is all pretty new research, but by one estimate, only a few percent of the female population (and none of the male population) has this super-human ability. This estimate came from a researcher who used to be right here in Milwaukee by the name of Jay Neitz.

Always wanting to make significant contributions to science, I did a little research on my own on the prevalence of this anomaly. I stopped a hundred woman in the mall and asked them if they were tetrachromats. Unfortunately, I had to curtail my data gathering when the police showed up.

As I have said, the existence of people like this is still under debate. As near as I can tell, the argument against goes something like this: "People who claim to be tetrachromats are just people looking for attention. If the whole tetrachromat thing didn't come up, they would claim to have been abducted by aliens, or would be singing Justin Beiber songs at a karaoke bar." Scary stuff indeed.

But... this still doesn't answer the question of why there is an apparent difference in the way men and two women see color. A few per cent of the female population is not enough to allow all women to be able to point to a color and tell whether it is ecru or oyster.

Physical variation in the eyeball

I just received a copy of the PhD thesis from a fellow by the name of Abhijit Sarkar [6]. The first comment I want to make--and this applies to everyone who is now working on a PhD thesis--is this: Why do you gotta make those things so darn long? And why do you gotta use such big words? I'm just saying... if you want your thesis to be tweeted about, you need to shrink it down to like two or three pages tops. And cut out all those darn equations, ok?

Dr. Sarkar describes several mechanisms within the eye that could cause two people to see color differently. The simplest one to describe is the yellowing of the lens in our eye. This happens gradually as we age. This yellowing has the effect of subduing the effect of light at the blue end.

I see the equivalent of this effect every day. My wife has our bedroom painted a pale lavender color. On a sunny day when the shades are open, the ceiling is a delicate pastel lavender. At nighttime, though, when we view the ceiling under incandescent light, the color is more of a gray [7].
The effect of a yellow filter on purple flowers

The yellowing of the lens in our eye has a similar effect. This effect is most striking for colors in the purple family. Sadly, as we grow older, we start to lose some of the subtle shades of purple. When I am old, I will wear purple, but I will need to have a younger woman around to appreciate it. [8].

Dr. Sarkar talks about a number of other mechanisms in the eye that act like filters to change the tint of what we see. His thesis is littered with phrases like "macular pigment optical density function". This of course, limits my ability to understand what he is talking about. But the big idea is that these tinted sunglasses that are built right into our eyes vary from one person to the next.

Luckily, this gentleman's thesis had a very clear and succinct abstract that even I could understand. There are differences in the make up of people's eyes [9] that mean we all see color just a bit differently. This effect, called observer metamerism, is a bummer. In industry, it makes it really darn hard to match colors. One person with "normal color vision" might say two colors match, and another may say they don't.

The whole crux of his thesis is that the problem is potentially solvable. His research says that people can be grouped into eight categories for their color vision, and he has developed a relatively simple test to categorize them. [10]

The practical application of Dr. Sarkar's work is when you are trying to match a color on a computer monitor with the color of a physical object held next to that monitor. Doing this is common in the printing industry, where people display what is called a proof on a computer screen. Customers sign off on a color based on this "soft proof", and press operators use that soft proof as a reference for when they are printing.

If you knew which category an observer fell into, it would be possible to adjust the computer monitor accordingly. There are some holes in implementation of this. In particular, I am not sure what happens when you have two people from different categories looking at the same screen.

But none of this sheds any light on gender differences. There is nothing in the thesis that is gender specific.


I don't have a final pronouncement on this, but I have yet to find a physiological explanation for why men and women seem to think about color differently. Until I hear differently, I am going to assume that if such a difference really exists, then periwinkle and wisteria and orchid and lavender originate somewhere in the murky canescent matter of my wife's brain, and not in the gray matter of my brain.

As for the simpler question, is there variability in the color response to different eyes, the answer is most certainly "yes".

[1] When I say something like "brilliant study from a brilliant color researcher", it is generally a euphemism for "I did this work". I prefer the euphemism because I am far too modest to take credit for it.

[2] I am not aware of any research that goes into how often color blind people get lucky, so I would prefer to call color blindness a "gender-linked trait" rather than "sex-linked".  On the other hand, there is a certain blue pill that is taken by some (often older) men that may cause a temporary color deficiency, and use of this blue pill is strongly correlated with getting lucky. I have heard of this pill, but have no personal experience to verify the effect. 

[3] That whole X/Y chromosome thing is a rip off, if you ask me. Men are also more likely to have hemophilia and male pattern baldness. I would give my manhood for another X chromosome!

[4] Ok... LMS stands for long, medium, and short, referring to the range of wavelengths that each accepts. It  is a simplification to call them red, green, and blue, since they are not exactly in the red, green, and blue region of the spectrum. The S cone is close to what we would call blue, but the M cone is broader than just the green part of the spectrum. And that darn L cone overlaps just a whole lot with the M cone. It would not be such an awful faux pas to call this the yellow-green cone. 

[5] If sock matching is an onerous burden for your lifestyle, I recommend checking out this iPhone app for sock matching.

[6] "Identification and Assignment of Colorimetric Observer Categories and Their Applications in Color and Vision Sciences", 2011 from Ecole polytechnique de l’Universit√© de Nantes.

[7] Of course my wife keeps insisting that the color is neither lavender nor gray, but rather a grayish lavender.

[8] My wife objected to this statement. But I won't tell you which one of us is older.

[9] Just to be clear, he is not talking about makeup, as in eye shadow and mascara. He is talking about how the eye is built.

[10] In my own experience, there are two categories of people: those who like to be put in categories, and those who don't.


  1. No footnote for the whole "men are dogs" implication? LOL

  2. Color vision changes even for the same individual during the day. We are, generally, better able to discern color differences during the morning than at night. Fatigue is the influence here, I've read. And, as we get older we lose color perception and discernment.

    When I consult with executives in the printing industry, I explain to them that, statistically, those with the best color vision are young women, around the age of 21 to 25. This news always finds favor with them.

    Studies have been done with color vision in cats, too. Using colored lights and electrodes in the cat brains, it has been determined that they see things all green, as if they have night-vision goggles. Their luminescence channel is cranked way up. This ought to be no surprise; indeed, this finding supports a bunch of color theory, but don't tell that to your cat.

  3. Here is some real science that explains the gender differences: