Cognitive Daily

All this talk about stereotypes can get you thinking. Perhaps some stereotypes reflect actual differences. Take color vision, for example: men often refer to themselves as “color-impaired,” letting the women in their lives make home design decisions and even asking them to match clothing for them. Maybe they’re just behaving in accordance with traditional stereotypes … but maybe there’s something more to it.

In the 1980s, vision researchers began to find some real physical differences between the eyes of many women and those of most men. “Normal” color vision is possible because we have three different types of cone cells in our eyes, each of which responds to a different wavelength of light. The process is basically the reverse of how a TV set or computer monitor works: on a TV, there are three different colored dots—red, green, and blue—and the millions of “colors” we see are based on mixtures of different proportions of those colors. In the eye, cone cells can have three different photopigments. These are usually generalized as red, green, and blue, but their actual values are closer to yellowish green, green, and bluish violet. To avoid confusion, psychologists typically refer to them to long-, medium, and short-wavelength sensitive cones. Supposing we’re looking at a yellowish-green thing, the long-wavelength cones are stimulated the most, the medium-wavelength cones are stimulated a bit, and the short-wavelength cones are not stimulated at all, and the appropriate signal is sent along the optic nerve to the brain, which then recognizes the color as “yellowish-green.”

What the researchers were finding when they actually looked at the structure of the eye is that many women—perhaps over fifty percent—possessed a fourth photopigment. Was this merely a genetic anomaly? Would the brain even be able to process this fourth input? The early research suggested that it would not. Women were no better at determining whether two very similar color patches were actually the same. They were only slightly better than men at detecting subtle spots of red light, a fact researchers attributed to individual difference.

However, Kimberly Jameson, Susan Highnote, and Linda Wasserman were not convinced by this evidence. Five- and six-year-old girls are better at naming colors than boys, and grown men are not as good at color-naming compared to women. They felt the existing measures of color sensitivity and color-matching did not capture all the differences between men and women, and devised a new experiment that they felt was more representative of real-world vision.

It’s quite difficult to examine an eye to determine if it has the fourth photopigment—the process generally involves removing the eye itself. Jameson and her colleagues might have had just a bit of difficulty recruiting volunteers to participate in an experiment requiring such extreme measures, so instead they used a genetic test to determine how many different photopigments participants were likely to possess (they estimate this process to be about 90 percent accurate—biologists will recognize this as the genotype versus phenotype problem). Of 64 participants in the study, 23 were women with 4 photopigments, 15 were women with 3 photopigments, 22 were men with 3 photopigments, and 4 were men with 2 photopigments (this is commonly called “color-blindness,” but most people with 2 photopigments can still distinguish between many colors).

Next, participants viewed a spectrum projected on a lucite window covered with tracing paper. Over the next hour and a half, they performed an array of tasks, including marking the edges of the visible rainbow, marking the locations of the “best example” of each of the major colors, and marking the edges of each “band” of color in the rainbow. Between each task, a camera flash was set off to mask the previous spectrum example, and the experimenter mounted a new sheet of tracing paper on the panel.

The most compelling results came from the number of spectral bands task:

Type of participant
Average number of spectral bands
Number of participants
Four-pigment females
Three-pigment females
Three-pigment males and females
Two-pigment males

Four-pigment females perceived significantly more bands of color than both three-pigment males and females. Further, three-pigment males and females are statistically indistinguishable, suggesting that the result is not due to some cultural difference between men and women.

So why were others unable to find significant results in a color-matching task when we see such dramatic results here? Jameson et al. suggest that there may be two (or more) different modes of seeing color, each processed differently in the brain. The brain may use the data from all four photopigments for some processes, but not for others. But this is still supposition. What’s clear from this study is that the stereotype of women being better with color may reflect real differences between men and women.

Jameson, K. A., Highnote, S. M., & Wasserman, L. M. (2001). Richer color experience in observers with multiple opsin genes. Psychonomic bulletin and review, 8, 244-261.


  1. #1 Cancer Defeater MT
    September 27, 2005

    Cognitive Daily covers gender differences in color vision

    Cognitive daily has this interesting article: men often refer to themselves as ¡°color-impaired,¡± letting the women in their lives make home design decisions and even asking them to match clothing for them. Maybe they’re just behaving in accorda…

  2. #2 Carl Manaster
    September 28, 2005

    So the number of 3-pigment males in the study is 22, right, and their average number of spectral bands is 7.1? That strikes me as possibly still significantly different from 7.6 – though of course not as significantly different as the four-pigment females. I think it would have been better to present the table this way, rather than double-counting the 3-pigment females.

  3. #3 Tommy
    September 28, 2005

    Im wondering why women have this extra feature. I mean a lot of things can be related back to stoneage while men are hunting etc, women stayed at home. And you can see that some skills are better developed of that, manipulating objects in your head versus communicating. What would the reason be here? Is this handy for cooking/eating to distinguish better between colors?

    Any ideas?

  4. #4 Dave Munger
    September 28, 2005


    The .5 difference between men and women with 3 photopigments did not approach significance—there was a trend, but there were also significant individual differences. The standard deviation for these scores was in the range of 2—meaning most women’s scores were spread about in the range from 9.6 spectral bands to 5.6 spectral bands, and most men were spread from 5.1 to 9.1 spectral bands. There’s too much overlap there to call this a significant difference.


    Good question. On the other end of the scale, there’s a much higher incidence of color-blindness among males. Because this is an x-linked trait, it’s partially due to the genetics of the situation (not going to go into too much detail here), but if it was really adaptive for men to have the extra photopigments, one would think that it would gradually enter the overall gene pool.

    I suspect this very subtle difference doesn’t help men or women too much, and doesn’t offer a very large biological cost, so it just becomes a part of natural variation, like blonde hair or brown hair.

  5. #5 Martin
    September 28, 2005

    The way I understand television and computer screens is that they have three different colours of pixels, each of which corresponds to a wavelength that the pigments of “normal” person can detect. They can thus mimic any colour by replicating the amount of stimulation that the colour would produce in each of the three pigments. When a person can detect four wavelengths however, there would not be a pixel that directly stimulates the fourth colour (although I guess it could be stimulated a little by the other colours if the wavelengths were close).

    Televisions and computer screens might therefore not perfectly recreate real colours for viewers with four pigments. Do you know of any studies that have tried to look into this?

  6. #6 Dave Munger
    September 28, 2005

    Martin, I don’t think that’s quite right. There’s not a one-to-one correspondence between the pixels on a computer screen and the colors the photopigments detect. As I said in the article, the photopigments correspond to yellow-green, green, and blue-violet, while pixels are generally red, green, and blue. A standard computer monitor, to my knowledge, cannot produce the full gamut of colors detectable by even a trichromatic human eye. This is because while the yellow-green photopigment is activated the most by the wavelength corresponding to yellow-green, it will also be activated by red and green—just to a lesser extent. A red pixel can only be varying brightnesses of red.

    The problem for the brain is to determine if a lower than peak firing rate for the yellow-green photopigment corresponds to a less-bright patch of perfectly yellow-green color, or to a more-bright red color. This is where the other photopigments come into play. If the green photopigment is firing as well, then the brain understands that the color must be in the range of the green photopigment, and is therefore greener than the true bright red. Make sense?

    Wikipedia has a pretty good article on how this all works.

  7. [...] FUNDAMENTAL DIFFERENCE between the physiology of men and women. [...]

  8. #8 Lea
    September 30, 2005

    Tommy & Dave,
    Actually I can think of many cases in which greater color differentiation would be useful. For example, in the garden or in the woods, plants display nearly infinite color variation, and one way to identify plants and to distinguish healthy plants, fruits, vegetables, etc.from diseased ones is through color. Then there are the nutritious vs poisonous plants, a life and death issue. If we accept the hypothesis that women were the gatherers, and furthermore probably invented agriculture, increased color perception would be very useful.

    In addition, people’s health problems often show up in subtle changes in skin coloration (see ancient Chinese medicine)–again, a well-developed ability to see color would be useful to the female person in charge of managing her family’s health in paleolithic times.

    As for hunters: well aren’t hunting animals color-blind? Perhaps having keen senses, recognizing shapes and noticing types of movement might be more valuable in that case–color detail wouldn’t be as important.

  9. #9 MeMo
    September 30, 2005

    In living color

    Sometimes science just explains everything. Women may actually see more colors than men. Really, you take it from here. This is my gift to you. It’s too easy, kind of like the Bill Bennett thing….

  10. #10 afigbee
    September 30, 2005


    That says the adaptation would be fundamental and really ancient, probably found in all primate females.

    Were there no males at all found with four-pigment eyesite?

  11. #11 Dave Munger
    September 30, 2005

    Actually, yes, about 10 percent of males probably have four pigments.

  12. #12 Susan Marshall
    October 1, 2005

    I am curious to know whether gay men were more prone to quadrochromatism. I am working on a genetic hypothesis re: XY vs XX and homosexuality.

    Thank you for clarifying my hypothesis re: color perception…something I have considered since childhood, but could not prove.

  13. #13 dr gordon c cole
    October 4, 2005

    female superiority in color naming is also related to alexithymia. Alexithymic women named colors slower than non-alexithymic women.

  14. #14 Eric
    October 16, 2005

    Did the quadrochromatic women detect a wider spectrum of visible light? The idea of women needing to detect healthy plants sounds plausible – or to select berries among greenery. Men may be better keyed to motion.

    But the spectrum issue: plant health is easily determined when near-infrared reflectivity is examined. A plant exhibiting healthy photosynthesis will reflect near-infrared (just beyond visible red) brightly.

  15. #15 julian F.
    October 17, 2005

    A full exploitation of the fourth pigment would allow women not just to discriminate colors better (such as between different hues of orange), but also to actually see more colors.While the subjective classification of colors, i.e., the number of different colors that we name or categorize, is probably cultural, or experience based (orange was not different from red a few centuries ago), there would be an objective measure to detect this on women.Using colors generated from four filters instead of three (i.e., a four-chrome tv), would generate many colors that would be indistinguishable by men, but not for women. this would be not just a better discrimination, as you could generate a new color that differs a lot in four-color space, so that women could perceive them as quite different, not just as sligthly different, but men could not, is they are similar in three-color space. Let us say, an analogous difference between three-chromats and di-chromats should appear between men and four chromat women. Am I rigth? is it so difficult to set up an experiment like this?

  16. #16 Robert Bate
    October 19, 2005

    Is it possible to isolate a population of 4 receptored individuals, or are you only isolating a population of higher probability?

    In a population of 4 receptored individuals, do they make extra distinctions equally around the spectrum or is their advantage limited to one particular area, the greens or the blues, or maybe even seeing a spectral color beyond violet or red?

    This goes to a question of whether it is possible to see other colors unavailable to the rest of us.

  17. #17 Dave Munger
    October 19, 2005

    It’s just a population of higher probability—but the probability is over 90 percent, so results are still quite robust. I believe the differences are most prevalent in the red area of the spectrum, which is near the wavelength where the extra photopigment is most active.

    What this study addresses is that 4 photopigment individuals seem to make more distinctions between colors. That, to me, indicates that they are seeing more colors. But clearly, more work needs to be done in order to address that question more precisely.

  18. #18 Greta Munger
    October 19, 2005

    I don’t have the original in front of me, but I think the 4 photopigment individuals had a broader range of visible colors, with more reds. The apparatus involved using a light bulb that produced light wider the visible spectrum, and ran that light through a prism (which would spread all wavelengths out). The task was both to mark where colors started, and where boundaries were visible.

  19. #19 Robert Bate
    October 20, 2005

    If short wavelength cones peak at about 420nm. middle ones at 530nm and long ones at 560nm, do the researchers know where the fourth cone peaks?

    Margaret Livingstone refers to opponent cell processing that calculates the reading of, in one instance, the long cone in opposition to readings for the short and middle cones. Do the researchers indicate how the brain could deal with this extra information?

    Four coners may, in addition to making extra hue distinctions, pervieve the standard distinctions differently. Yellow and greenish yellow appear lighter at peak intensity than does violet-red not because more photons are reaching the eye but because the majority of our cones peak in the yellowish green section of the spectrum. Four coners may percieve oranges or reds even lighter than do we civilians.

  20. #20 Claudia Vess
    October 22, 2005

    As a visual artist, I can attest to the fact that men, especially as they age, perceive the relations of blue and green differently than women. There are some articles floating around about this.

  21. #21 Addie Sandbeck
    October 31, 2005

    I know that both my brother and father see colors differently than I do. Neither of them are color blind, but my mother and I often get into arguments with them about the colors of certain things. For instance: we used to have a black Subaru, but my dad thought it was green and my brother thought it was grey, while both my mother and I knew that it was black.

  22. #22 Ray Oickenson
    December 19, 2005

    Think I remember research from long ago suggesting that in some other primates the females also had “wider” color perception.

    Can quite see in early mammals the survival value in better discrimination of colors – both for foraging and for recognition of camouflaged predators.

    So, once “invented” in mammalian females (main foragers _and_ protectors of the young) it would just stick – no evolutionary reason to drop it.