This is to complement an older post of mine, The Blue That Is Green (And More). Here’s what I just read in the Guardian: Do you see blue or green? This viral test plays with color perception – A visual neuroscientist realized he saw green and blue differently to his wife. He designed an interactive site that has received over 1.5m visits.

It started with an argument over a blanket.

“I’m a visual neuroscientist, and my wife, Dr Marissé Masis-Solano, is an ophthalmologist,” says Dr Patrick Mineault, designer of the viral web app ismy.blue. “We have this argument about a blanket in our house. I think it’s unambiguously green and she thinks it’s unambiguously blue.”

Mineault, also a programmer, was fiddling with new AI-assisted coding tools, so he designed a simple color discrimination test.

If you navigate to ismy.blue, you’ll see the screen populated with a color and will be prompted to select whether you think it’s green or blue. The shades get more similar until the site tells you where on the spectrum you perceive green and blue in comparison with others who have taken the test.

“I added this feature, which shows you the distribution, and that really clicked with people,” says Mineault. “‘Do we see the same colors?’ is a question philosophers and scientists – everyone really – have asked themselves for thousands of years. People’s perceptions are ineffable, and it’s interesting to think that we have different views.”

Apparently, my blue-green boundary is “bluer” than 78% of others, meaning my green is blue to most people. How can that be true?

This is so wrong on so many levels!

First, this is not just a matter of sensorial perception. It’s a matter of education. As I mentioned in my previous post, the “calibration” at the other end of the optical nerve, the brain, depends on two factors: how many objects of each color (and in which shades) was one exposed to as a kid; how was each object’s color defined to the kid by whoever defined the color for them: a parent, a grandparent, etc. The fact that a parent told you as a kid, “this object is green, not blue,” has “calibrated” your mental sense of colors!

Secondly, it’s also a matter of national or local culture. In some countries, languages, or regions, “turquoise” is referred to as “turquoise green,” in others as “turquoise blue,” whereas the correct definition would be “a color in-between green and blue.” It’s greenish-blue, or blueish-green.

In computers and printing, cyan is what neutral turquoise used to represent in some cultures in the past. That’s because its hex code is #00FFFF, halfway between blue and green in the RGB color model: RGB(0, 255, 255), meaning green = blue = 255.

But when you’re mentally conditioned by the “turquoise green” naming, to you turquoise is green! Surprisingly enough, for many people, cyan is nonetheless “blue”!

Unfortunately, people are not educated by using pure colors of a unique wavelength, so they can’t be told, “this is the threshold between green and blue, at so many nanometers.” Even worse, turquoise is even more ambiguously defined and perceived: where does it end, and where do green and blue begin?

The Guardian articles adds:

There’s no word for “blue” in ancient Greek, for example, which is why Homer described stormy seas as “wine-dark” in The Odyssey. By contrast, Russian has distinct words for light blue and dark blue. However, recent research suggests a greater vocabulary may only be beneficial for remembering colors and not for perceiving them.

I’m not buying the results of that study. The ancient lack of the concept of “blue” is mesmerizing (I even linked to a video). But language is more than important. To quote myself,

I would add to what the video says that blue and green still have special statuses in some cultures. In Romanian, “bleu” means “light blue” (unlike in French, where “bleu” means just “blue”) and “vernil” means “light green”; in Russian, “dark blue” (синий) and “light blue” (голубой) have distinct terms; in Italian, “azzurro” is “blue” (with a tendency to be used for lighter shades, as it literally means “sky blue”; to definitely mean “sky blue” one has to use “azzurro chiaro”), whereas the generic blue is “blu” (duh).

When people just use “blue” and “green” for no matter what shades, they tend to become bat stupid, just like our neuroscientist-cum-programmer, who designed a completely stupid test at Is My Blue Your Blue?

The second screen I was shown was this one:

I added a text to show you that the green and blue components are too close in value for a normal eye (and the brain backend!) to decide. So it’s neither green nor white, but turquoise! Also, almost cyan, but tending to green.

The screen I’ve used is an IPS with a decent gamma, and yet the above “turquoise that is sort of cyan but greener” somehow looked a turquoise tending to blue, rather. So I answered honestly to each question without checking the RGB values, and this is what I got:

First of all, it’s all a lie! Turquoise is neither green nor blue, and it’s not a unique RGB triplet, but a range of shades! What’s important to me is not where people are considering its CENTER to be, but where it ENDS! In other words, where some people still consider a shade to be turquoise, whereas most people would say it’s “definitely blue” or “definitely green”! But the mediocre mindset only cares about the CENTER, seen as a BOUNDARY: to the left, it’s green; to the right, it’s blue. That’s because the question is wrong! The proper question would have been: is this ❶ green, ❷ turquoise, or ❸ blue?

Garbage in, garbage out.

There seems to be such a fine line between your being considered a “greener” or a “bluer”:

  • My result: Your boundary is at hue 175, bluer than 59% of the population. For you, turquoise is green.
  • The screenshot from the article: Your boundary is at hue 171, greener than 72% of the population. For you, turquoise is blue.
  • For the author of the article: my blue-green boundary is “bluer” than 78% of others, meaning my green is blue to most people.

So, from 171 to 175, the decision changed from “turquoise is blue” to “turquoise is green”! ONCE AGAIN, TURQUOISE IS NEITHER BLUE NOR GREEN, YOU FUCKUPS!

Also, “hue 171” or “hue 175” are not telling much to most people. Let me clarify:

  • The perfectly neutral shade of turquoise, which is cyan aka RGB(0, 255, 255), has a hue of 180°.
  • Hue 175° means RGB(0, 255, 234) if S=100% and V=100%, a turquoise tending toward green.
  • Hue 171° means RGB(0, 255, 217) if S=100% and V=100%, a turquoise tending even more toward green.
  • Perfect, pure green is 120°. Perfect, pure blue is 240°.
  • To have shades more toward blue than toward green, they need to be above 180°.

Therefore:

  • The OBJECTIVE green-blue boundary is at a hue of 180°.
  • The SUBJECTIVE green-blue boundary is not specified by that stupid web app, but the fact that between the hues of 171 and 175 the verdict changes from “your turquoise is blue” to “your turquoise is green” means that the median boundary for people is somewhere in-between, say around 173°, which is in the green territory! Most people lean to the left, green side.

What I already knew, as per the aforementioned post, is that some shades labeled as blue by some people are rather greenish to me, so yeah, let’s say that “my turquoise is green”!

However, according to this web app, not only many people consider shades of blue to be green (in contradiction with the case of Tiffany Blue®), but I and the author of the article (not the neuroscientist) are “bluer” than most people, meaning that we’re closer to the objective truth! With my boundary at 175°, closer to the OBJECTIVE one at 180° than the median that lies around 173°, my interpretation of colors is more accurate than most people’s!

Quick question: is the color aquamarine more blue or more green? According to Wikipedia, there is an aquamarine called “brilliant green” (hue 160°, definitely green, with saturation at 50%), another one called “vivid green” (hue at 165°, still distinctly green, but with 100% saturation, hence “vivid”), and a “medium aquamarine” (still 160°, still “brilliant green” and still 50% saturation, but only 80% brightness, hence darker). From the article: “Aquamarine is a color that is a light tint of teal, in between cyan and green on the color wheel.” And yet, aquamarine is typically perceived as more blue than green by most people!

A final clarification: teal is, just like cyan, on the boundary between green and blue:

  • cyan: #00FFFF, RGB(0, 255, 255), HSV (180°, 100%, 100%) — green and blue at the maximum
  • teal: #008080, RGB(0, 128, 128), HSV (180°, 100%, 50%) — green and blue at half-value

And let’s not dive into shades such as “teal blue” (around 194°), “teal green” (same hue of 194°, so it’s blue and abusively called green just because it’s darker) and “deep sea green” (hue 182°, so ever so slightly bluish, but very dark), or “process cyan” aka subtractive cyan (with a hue of 193°, it’s definitely blue, but there is a reason for CMYK to be the way it is).

All in all, allow me to say that people are incredibly stupid, and they stick the labels “blue” and “green” at random, often based on the brightness of the color!

Fucking retards. Misanthropy is the right attitude.

BONUS: The language connection

I mentioned that what each individual believes to correspond to the words “green,” “blue,” “turquoise,” “cyan,” “teal,” and so on, is more a matter of what they were told about colors in their early childhoods. I also believe that the way some cultures use different words to designate different shades of certain colors also plays a role, but right now I want to focus on the previous aspect.

Family or society, here’s a comparison between colors and sounds.

A baby hears different people uttering words that sound all different to her or him. When mom says “baby,” this doesn’t sound quite the same as when dad says “baby.” And the sound that we call “b” doesn’t sound always the same, does it? It will have taken an infant thousands and thousands of different hearings of the same words to reach the conclusion that different sounds are supposed to be considered identical! It’s a necessary classification, similar to what Machine Learning does. And it’s unassisted, more like Deep Learning, because in the absence of a known language, the meaning of words has to be inferred from context. It’s hard being a baby.

But what happens when you’re not exposed to a certain sound as a kid? Consider the following examples.

In Spanish, the letters “b” and “v” represent practically the same sound, which is typically a soft “b.” This can cause native Spanish speakers to mix up “b” and “v” in English. Unless they have an important exposure to English (which might be easier to get in northern Mexico than in Andalucía), they’ll struggle to utter a proper “v.”

Then, in Japanese, the language lacks distinct “b” and “v” or “f” and “p” contrasts as those found in English. The “p” sound (when aspirated) is less common in Japanese, so it’s difficult to grasp for native speakers of Japanese, and certain sounds are adapted in other ways: “v” often becomes “b” and “f” becomes “h.”

Korean has three types of stop consonants for sounds like “p,” “t,” and “k,” which are distinguished by tenseness, aspiration, and voicing. The “default” pronunciation is “lenis” (lax, plain); no strongly aspirated, and not tense. Then there’s “fortis,” a glottalized pronunciation: tensed, but no aspiration. Finally, there are also aspirated utterances, with a strong burst of air.

NOTE: Linguistics is a dumb animal quantified by (I have no word for it) retarded people. In linguistics, “aspirated” actually means “with a release of a strong burst of air,” hence literally “expirated”! (In English, the “p” in “pit” is aspirated, meaning there’s a puff of air following the release of the “p” sound. In contrast, the “p” in “spit” is unaspirated, with much less air released after the “p.” There are better examples for other sounds, more contrasting, but let’s leave it this way.) I’m happy I’m not a linguist, as I cannot use the term “white” for “black” and vice versa. Let me state it again: The term aspiration refers to the expulsion of air (not inhalation) in linguistic contexts.

Either way, Korean is not an easy language, except maybe in writing.

But as, in most cases, “b” is voiced and “p” is aspirated in English, this would confuse the Koreans, as they have a “lenis” sound that’s more like something between our “b” and “p”! As an approximation, the Romanization of the “lenis p” is “b,” and the Romanization of the tense “fortis p” and of the aspirated p is “p.”

To oversimplify, it’s like you were taught as a kid to classify a category of sounds into a “lenis p” which is more like “b.” Let’s call it “turquoise.” Then you must learn and reproduce some foreign sounds they claim to be either “b” or “p” but which sound to you like variations of this “lenis” which is sort of “turquoise,” but you have to obtain the proper, unmistakable “b” and “p” sounds, just as if you were asked to paint proper “green” and “blue” shades that people wouldn’t mistake one for another! But when you try to say “p,” what you can utter is something close to one of your “lenis” sounds, and some people will hear “b”! You’re persuaded you painted a wall in green, but it’s a shade of turquoise that to some people looks blue!

I’m not sure that I explained the parallel convincingly enough.

Let’s not leave aside the Mandarin Chinese, where the “b” and “p” distinction exists, but they are separated by aspiration rather than voice. Again, not only is this difficult for English speakers, but also when a Chinese tries to utter proper English “b” and “p” sounds, they might be perceived as sounding strange.

In true honesty, the real difficulty in Chinese stays in their 4+1 tones: four, and a neutral one. The classic example: “mā” = mother, “má” = hemp, “mǎ” = horse, “mà” = scold, and the neutral “ma” is a question particle placed at the end of a statement instead of a question mark.

As for the classifications made by linguists, I’ll probably die before being able to understand their definitions of open and closed, front and back. Yes, I’ve seen pictures and animations, but there is no brain connection between those pictures and me while uttering a sound. Examples:

  • You have two different front open vowels in “cat” [æ] and “trap” [a].
  • You have two different back open vowels in “father” [ɑː] and “lot” [ɒ].
  • You have two different front closed vowels in “see” [iː] and “sit” [ɪ].
  • You have two different back closed open vowels in “goose” [uː] or “foot” [ʊ].

Already those IPA signs… when Americans struggle even with French accents, so that Jay Leno writes “coupe” instead of “coupé”…

And there is more. The sounds for “u” in “uniform” (which is [j]) and “w” in “whiskey” (which is [w]) are considered consonants in English, despite them being classified as semi-vowels in other languages. If something sounds like a vowel but has the functions of a consonant, how do you call it? Well, if your mind is 1-bit, then consonant. If you can think on 2 bits, then it’s a semi-vowel.

What’s easier, colorimetry, or linguistics?

Colorimetry, but only if you are using a colorimeter, calibrated displays, and rigorous definitions of color names.

In real life, there are no proper definitions of colors in dictionaries and color charts, people aren’t taught anything properly, and they use computer and smartphone screens with atrocious gamuts, poor color calibration, and random gamma.

So let’s learn Mandarin. It only takes about 3–4 years to reach B1, and 7+ years to reach C2.