Huh. Neuromarketing is good for something.

Not really — it’s a buzzword and a scam. But at least it provoked this very good summary of color vision and color blindness in response to an absurd claim by a neuromarketer.

It’s a solid response to the claim that 25% of people are tetrachromats…although I’d never heard such an extravagant number before. Tetrachromats — people who effectively have four different functional classes of cones in their retina — are incredibly rare, exist largely in the domain of a theoretical possibility, and have been demonstrated only a couple of times. It takes some chutzpah to argue that they’re ubiquitous!


  1. blf says

    It takes some chutzpah to argue that they’re ubiquitous!

    Nah. In marketingland, 1 + 2 = 4, except when it is more profitable to be π, or whatever other value the client wants, such as “cute hedgehog”.

    Very similar to lawyers (except no ambulance chasing is involved), and politicians (except no rubber chicken kissing or law degree is involved).

  2. zenlike says

    Also ‘Professor’ Diana Derval who wrote that drivel is apparently totally unaware how fucking badly consumer monitors are in terms of colour representation and most people leave their monitor in a terribly calibrated state. I myself saw a lot of overlap in her ‘test’, but my monitor is also calibrated in a bad way. 2 minutes fiddling with the settings probably fixes that.

    Does that mean that fiddling with monitor calibration settings restores colour-blindness? Does that mean I can get my findings published in the same hack journals she publishes in? Does that mean I can also get a professorship at the scam institutions she frequents?

    Sadly, snake-oil saleswoman and general scam-artist Diana Derval seems to be doing pretty well for herself. How much that tells you about marketing people in general I will leave as an exercise to the reader.

  3. Marshall says

    This popped up in my facebook newsfeed and it made me laugh out loud. Human trichromats can distinguish about 10 colors, and they’re trying to claim that seeing 30+ means you’re a tetrochromat?

    Also, the image in the original article is compressed to hell; I checked in Matlab and there are actually 312 different hues along the horizontal axis of the image at the midline.

    A much better test is to use the Farnsworth-Munsell 100 Hue Test, like the version found here:

  4. Marshall says

    You can be a trichromat and still get a perfect score on the Farnsworth test, which is much harder than the bunk “tetrachromacy” test in the original article.

  5. llewelly says

    The marketing industry generates an endless stream of crap ideas. Most of those people are afraid they wouldn’t make a dime if there was a single fact in their messaging.

  6. llewelly says


    … most people leave their monitor in a terribly calibrated state …

    First fact is, most digital cameras have crappy and variable calibration too, so getting your own monitor right is not actually very helpful.

    All it will do is get you involved in appallingly ridiculous and irrelevant arguments, like whether or not John Boehner has orange skin.

    There’s no reason to do it unless you’re a professional or a dedicated hobbyist. I haven’t calibrated a monitor since I left the video game industry.

  7. zenlike says


    True, in fact as I said in my comment, my own monitor is badly calibrated and I don’t bother fixing it because it doesn’t really matter. It just means that any ‘test’ for viewing colours is meaningless when they are done on a consumer monitor. Diana Derval’s ‘test’ fails on a very basic level already. But she doesn’t obviously doesn’t care or doesn’t know any better (probably both).

  8. cvoinescu says


    … most people leave their monitor in a terribly calibrated state.

    That used to be very true with CRT monitors, and still true to some extent with LCD displays with VGA (analog) connectors, but, with the advent of DVI and HDMI, the average monitor is now decently calibrated out of the box (at least until you mess with the settings). Not good enough for professional use, of course, but color reproduction on monitors is no longer the complete nightmare it used to be. (The average digital camera is still pretty bad, and most webcams and cell phone cameras are still abysmal.)

  9. woozy says


    Even more basic (as well and frequently stated in the article): For an image produced with trichromate techniques, as *all* internet imaging are, viewed on a trichromate medium (as all monitors and tv screens are) it is simply logically impossible to reproduce and to see any tetrachromate data. It’s simply absurd. You might as well zoom in on a Panorama of San Francisco from the Marine headlands and expect to make out the license plate of the car on Lombard street, or to sniff your monitor and expect to smell the odor of the flowers photographed. The information just isn’t there and isn’t recorded or produced.

  10. saganite says

    Dumbfuck M cells. =)
    If I ever write an even vaguely ophthalmological paper, I have to get that in there somehow.

  11. blgmnts says

    Serious question: What is the cause of tetrachromacy in women?

    Since reading an article about the evolution of trichromacy in primates (which said that it was probably caused my an imperfect exchange of regions between the 2 X chromosomes thus leaving one with two copies of the gene for the L receptor) I thought it is connected to X inactivation. I guessed that part of the L or M cones get “their” genes from different X chromosomes and thereby are slightly different in their spectral sensitivity, but I would really like to know how it is.

  12. vewqan says

    woozy: I’ve actually seen RGBY televisions for sale, and on close inspection they do indeed have four different pel colours. I’ve tried buttonholing salespeople about why manufacturers would want to cater to the <0.1% of the population with true tetrachromacy but no one seemed to have any idea. (The sales material was just rubbish about four colour channels being obviously better than three.)

    Of course every extant video signaling standard presumes three colour channels anyway, so even if you had one you still wouldn't be able to view a tetrachromacy test on it.

  13. dorght says

    I would be interested to know if the verified tetrachromates actually had a multiple yet slightly different copies on one chromosome versus one x chromosome still expressing the gene despite x-inactivation. The third possibility being that the women actually have a patch work of areas in their retina that have one of the different cones being expressed but not the other, Depending on when inactivation occurs.