# Rainbows: Why Not Just Three Colors?

**URL:** <https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243>\
**Category:** Factual Questions\
**Created:** [January 23, 2019, 2:36am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243 "2019-01-23T02:36:08Z")\
**Posts on this page:** 20\
**Page:** 3

<div class="post-metadata">

**Author:** ![DPRK](https://avatars.discourse-cdn.com/v4/letter/d/4491bb/32.png) [@DPRK](https://boards.straightdope.com/u/DPRK)\
**Post date:** [January 24, 2019, 12:54am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/41 "2019-01-24T00:54:43Z")

</div>

> [@CalMeacham](#):
>
> We can actually answer the question of what color Newton was naming “indigo” by using his own scheme of equating the musical notes with the characteristic lengths associated with the colors – which we now know to be wavelengths. I ran the calculations, and found that Blue came out to be about 465 nm, indigo about 417, and Violet about 390. By comparison, Wikipedia and other sources give about 470 nm for blue, 450 for Indigo, and 420 for Violet.
> 
> You get slightly different values if, instead of the “Classical” values you use even tempering of your scale, but not that different – 488 for Blue, 450 for Indigo, and 387 for Violet. You can look these up on a spectrum (although if you look at an on-line source, you won’t really get proper color reproduction).
> 
> This is from a piece I wrote titled “The Well-Tempered Spectrometer”, which hasn’t yet been published.
> 
> If, instead of a simple octave, you go with a Chromatic Scale (which includes the “black keys” on the piano), you’ll end up with a new color at 690 nm between Red and Orange, one aty 615 nm between Orange and Yellow, one at 516 between Green and Blue, one at 460 between Blue and Indigo, and one at 410 nm between Indigo and Violet.
> 
> Imagine that you’re a modern-day Newton trying to buttonhole your friends into seeing these new colors, and come up with names for them.

I get that Newton decided to mark seven (or 5+2?) colours, but what was his exact procedure? It seems a stretch to squeeze a full octave in there, for starters. And did he just name those seven colours and ask his friend to mark those off on a spectrum, or did he construct a tempered scale first, according to mathematical principles, and superimpose that onto the spectrum?

Looking at Newton’s [colour circle](https://upload.wikimedia.org/wikipedia/commons/0/0a/Newton%27s_colour_circle.png) pictured here, it appears that the interval between Blue and Indigo is (approximately) equal to that between Indigo and Violet, but your equally-tempered ratios 488/450 and 450/387 do not match? Also, if we go for your chromatic version, wouldn’t we need to split each of Red, Yellow, Green, Blue, and Violet into two, eliminating each of those colours in favor of two new ones?

---

<div class="post-metadata">

**Author:** ![TokyoBayer](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/tokyobayer/32/13989_2.png) [@TokyoBayer](https://boards.straightdope.com/u/TokyoBayer)\
**Post date:** [January 24, 2019, 3:56am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/42 "2019-01-24T03:56:09Z")

</div>

> [@septimus](#):
>
> Some linguists have observed the following:
> 
> All languages have at least 2 colors - white and black
> 
> Languages with 3+ colors have red.
> 
> Languages with 4+ have green or yellow
> 
> Languages with 5+ have green and yellow
> 
> Languages with 6+ have blue
> 
> Languages with 7+ have brown
> 
> Languages with 11+ have purple, pink, orange, gray.

Do you have a cite for this? I wonder how many languages there are around with only two words for colors, etc.

---

<div class="post-metadata">

**Author:** ![dtilque](https://avatars.discourse-cdn.com/v4/letter/d/d6d6ee/32.png) [@dtilque](https://boards.straightdope.com/u/dtilque)\
**Post date:** [January 24, 2019, 5:27am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/43 "2019-01-24T05:27:54Z")

</div>

> [@TokyoBayer](#):
>
> Do you have a cite for this? I wonder how many languages there are around with only two words for colors, etc.

[Berlin & Kay](https://en.wikipedia.org/wiki/Basic_Color_Terms:_Their_Universality_and_Evolution)’s book on it was published back in 1969. The former Perfect Master known as Cecil did a [column](https://www.straightdope.com/columns/read/449/could-early-man-only-see-three-colors/) summarizing B&K’s book. There’s been a lot of research on this topic since, but I’m not aware of any good summary of it.

---

<div class="post-metadata">

**Author:** ![septimus](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/septimus/32/410_2.png) [@septimus](https://boards.straightdope.com/u/septimus)\
**Post date:** [January 24, 2019, 8:32am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/44 "2019-01-24T08:32:08Z")

</div>

> [@TokyoBayer](#):
>
> Do you have a cite for this? I wonder how many languages there are around with only two words for colors, etc.

The source I found yesterday, which cites the B&K book \*\*dtilque \*\*mentions, was “The Linguistic Significance of the Meanings of Basic Color Terms” ([The Linguistic Significance of the Meanings of Basic Color Terms on JSTOR](https://www.jstor.org/stable/412789)) which  
(a) may need you to register (for free) to read the 37-page article;  
(b) needs $12 before presenting a version which allows copy-paste;  
(c) was intended as a partial refutation of B&K’s result;  
(d) is more about vision physiology than language variations.

It does mention the Dugum Dani language of which [Wikipedia writes](https://en.wikipedia.org/wiki/Dani_people):

> [@](#):
>
> The Dani languages differentiate only two basic colours, _mili_ for cool/dark shades such as blue, green, and black, and _mola_ for warm/light colours such as red, yellow, and white. This trait makes it an interesting field of research for language psychologists, e.g. Eleanor Rosch, eager to know whether there is a link between way of thought and language.

In refutation of B&K, the jstor article mentions that 69% of Dani informants focused white-warm _mola_ at Red.

---

<div class="post-metadata">

**Author:** ![BigT](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/bigt/32/12044_2.png) [@BigT](https://boards.straightdope.com/u/BigT)\
**Post date:** [January 24, 2019, 9:06am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/45 "2019-01-24T09:06:42Z")

</div>

> [@Francis\_Vaughan](#):
>
> Right to a very good first approximation. But there is a sting in the tale. We see violet, and that is past pure blue. We should not perceive a spectral colour past the peak of blue sensitivity if you use a simple RGB mixing model. The sting is that there is a tiny blip in the red sensitivity in the far blue, and our brain actually does get a small mix of red with blue when seeing spectrally pure violet.
> 
> Part of the reason is that our eyes don’t have red green and blue sensors anyway. They have three sensor types that have very broad sensitivity, but with different slopes and break points across the spectrum. Subtract the signal of one from the others in different combinations and you can get the differences as signals in the three colour bands. However even this isn’t done, they eye processes the three different sensors into a luminance and two difference signals, and they is what is pumped back into the brain. (I was really impressed with some work done a while ago where the neurons in the back of the retina had their structure analysed and the logic of the interconnections reverse engineered, and out popped the sum and difference signals. Previously the nature of the signalling had been painstakingly derived by working through the precise colour sensitivity the eye brain system exhibits, and fitting the model to the way we see both spectrally pure and mixed colours.)

And this is ultimately the answer to a question I asked in GQ long ago, of why we see red in violet. It’s not in the cones (sensors), as I had originally supposed, but in the process of sending the visual signal to the brain. It’s a quirk that allows us to identify colors beyond blue by mapping them partly into the red spectrum.

I presume the eye doesn’t just send the raw signals to conserve bandwidth, as it would have to send signals from all the cones and the rods. Instead, some minimal processing occurs.

---

<div class="post-metadata">

**Author:** ![BigT](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/bigt/32/12044_2.png) [@BigT](https://boards.straightdope.com/u/BigT)\
**Post date:** [January 24, 2019, 9:54am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/46 "2019-01-24T09:54:28Z")

</div>

> [@CalMeacham](#):
>
> We can actually answer the question of what color Newton was naming “indigo” by using his own scheme of equating the musical notes with the characteristic lengths associated with the colors – which we now know to be wavelengths. I ran the calculations, and found that Blue came out to be about 465 nm, indigo about 417, and Violet about 390. By comparison, Wikipedia and other sources give about 470 nm for blue, 450 for Indigo, and 420 for Violet.
> 
> You get slightly different values if, instead of the “Classical” values you use even tempering of your scale, but not that different – 488 for Blue, 450 for Indigo, and 387 for Violet. You can look these up on a spectrum (although if you look at an on-line source, you won’t really get proper color reproduction).

I did actually use an [online source](https://www.johndcook.com/wavelength_to_RGB.html), for convenience. I agree with you that it isn’t super accurate, but I think it can be useful, if what I say turns out to be true on most screens.

First set of numbers: **Blue Indigo Violet**  
Second set of nums: **Blue Indigo Violet**

I then used GIMP to try and equalize the brightness, for a clearer comparison:  
First set: **Blue Indigo Violet**  
Second set: **Blue Indigo Violet**

Based on these results, the classical gives me what I would call blue, blue-violet, and red-violet. But the tempered gives me the cyan, blue, purple (slightly redder than violet).

I would expect that, to make it harmonious and fit the pure math, he’d use classical. Still, there are those color wheels that look more like he used even temperament. I could see him doing so to try and shoehorn in his theory.

I will finally leave you with this, since you seem to have some expertise, and can critique it: [Indigo - Wikipedia](https://en.wikipedia.org/wiki/Indigo#Classification_as_a_spectral_color)

---

<div class="post-metadata">

**Author:** ![CalMeacham](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/calmeacham/32/35_2.png) [@CalMeacham](https://boards.straightdope.com/u/CalMeacham)\
**Post date:** [January 24, 2019, 12:26pm UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/47 "2019-01-24T12:26:09Z")

</div>

> [@DPRK](#):
>
> I get that Newton decided to mark seven (or 5+2?) colours, but what was his exact procedure? It seems a stretch to squeeze a full octave in there, for starters.

I told you the procedure in my earlier post. Why do you think it’s a “stretch” to “squeeze” an octave in there? You can divide any length by seven or eight.

---

<div class="post-metadata">

**Author:** ![CalMeacham](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/calmeacham/32/35_2.png) [@CalMeacham](https://boards.straightdope.com/u/CalMeacham)\
**Post date:** [January 24, 2019, 12:32pm UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/48 "2019-01-24T12:32:13Z")

</div>

Incidentally, I suspect that people continue to use “Indigo” and “Seven Colors” because of its long history (and social inertia), because there’s more of a mystic association to “seven” than to “six”, and so that Roy G. Biv will have a pronounceable last name. But they still use only six colors in a rainbow flag ( [Rainbow flag - Wikipedia](https://en.wikipedia.org/wiki/Rainbow_flag#/media/File:Gay_flag.svg) ) and in the Resistor Code ( [https://www.instructables.com/id/Resistor-Color-Code-Guide/](https://www.instructables.com/id/Resistor-Color-Code-Guide/) )

---

<div class="post-metadata">

**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\
**Post date:** [January 24, 2019, 12:43pm UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/49 "2019-01-24T12:43:55Z")

</div>

The original Rainbow flag actually had 8 colors (pink and cyan/turquoise) were the additional two colors.

---

<div class="post-metadata">

**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\
**Post date:** [January 24, 2019, 12:51pm UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/50 "2019-01-24T12:51:08Z")

</div>

> [@BigT](#):
>
> I would expect that, to make it harmonious and fit the pure math, he’d use classical. Still, there are those color wheels that look more like he used even temperament. I could see him doing so to try and shoehorn in his theory.

If you look at the wikipedia article posted on indigo above, there is [chart from Newton](https://en.wikipedia.org/wiki/Indigo#/media/File:Newton_prismatic_colours.JPG) that does show that each color band was not exactly the same size.

For me, when I look at a [spectrum](https://www.tissuegroup.chem.vt.edu/chem-ed/light/graphics/em-visib.jpg), there just seems to be a blindingly obvious stripe of light blue that seems like it should get a name.

---

<div class="post-metadata">

**Author:** ![dtilque](https://avatars.discourse-cdn.com/v4/letter/d/d6d6ee/32.png) [@dtilque](https://boards.straightdope.com/u/dtilque)\
**Post date:** [January 24, 2019, 12:57pm UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/51 "2019-01-24T12:57:01Z")

</div>

> [@septimus](#):
>
> In refutation of B&K, the jstor article mentions that 69% of Dani informants focused white-warm _mola_ at Red.

Not sure that’s a significant refutation of B&K, but that’s beyond the scope of this thread. B&K was unpopular among certain sectors of the linguistic world because it violated [Sapir-Whorf](https://en.wikipedia.org/wiki/Linguistic_relativity). Of course, not everyone liked S-W, so it was popular in other sectors.

---

<div class="post-metadata">

**Author:** ![CalMeacham](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/calmeacham/32/35_2.png) [@CalMeacham](https://boards.straightdope.com/u/CalMeacham)\
**Post date:** [January 24, 2019, 1:15pm UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/52 "2019-01-24T13:15:53Z")

</div>

> [@pulykamell](#):
>
> The original Rainbow flag actually had 8 colors (pink and cyan/turquoise) were the additional two colors.

I know – which is why I qualified my original citation of it. But the rainbow flag I link to is the current officially used one and the most common. The original Rainbow Flag uses two colors that break up its resemblance to a proper spectrum; the current one has six colors and looks more like what people expect to see in a rainbow.

> **[Rainbow flag (LGBT)](https://en.wikipedia.org/wiki/Rainbow_flag_(LGBT_movement))**
>
> The rainbow flag or pride flag is a symbol of LGBT pride and LGBT social movements. The colors reflect the diversity of the LGBT community and the spectrum of human sexuality and gender. Using a rainbow flag as a symbol of LGBT pride began in San Francisco, California, but eventually became common at LGBT rights events worldwide.
> Originally devised by the artists Gilbert Baker, Lynn Segerblom, James McNamara and other activists, the design underwent several revisions after its debut in 1978,...

---

<div class="post-metadata">

**Author:** ![CalMeacham](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/calmeacham/32/35_2.png) [@CalMeacham](https://boards.straightdope.com/u/CalMeacham)\
**Post date:** [January 24, 2019, 1:28pm UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/53 "2019-01-24T13:28:30Z")

</div>

> [@pulykamell](#):
>
> If you look at the wikipedia article posted on indigo above, there is [chart from Newton](https://en.wikipedia.org/wiki/Indigo#/media/File:Newton_prismatic_colours.JPG) that does show that each color band was not exactly the same size.
> 
> For me, when I look at a [spectrum](https://www.tissuegroup.chem.vt.edu/chem-ed/light/graphics/em-visib.jpg), there just seems to be a blindingly obvious stripe of light blue that seems like it should get a name.

My initial reaction is that the “blindingly obvious” light blue stripe is really an artifact of the way your monitor renders colors (and, of course, would render pictures of rainbows the same way). But, upon studying some actual spectra here at my desk (I’ve got a set of diffraction grating glasses that make me look like a dork when I put them on, but give me instant rainbows), I’m not so sure. That region of blue between green and violet can contain what looks like a lighter stripe of blue.

It was correctly stated above that the colors are not of equal width, whether this is the way they’re seen if plotted on a scale linear in wavelength, or the way the spectrum actually appears when generated by a prism or a raindrop. Yellow is definitely much narrower in extent than, say, red or blue, which take up huge areas. We clearly did not decide upon colors by trying to segregate the spectrum into regions of equal size. But Yellow is, despite its narrowness, so qualitatively different from what surrounds it, and so strikingly obvious that any scheme that rejected “Yellow” as a color would never be accepted and used.

By the way, as I observed above, not all rainbows look the same – the presence or absence of colors, and their relative widths, depends upon the sizes of the raindrops 9and, knowing the appearance, you can tell what size the drops are). And the appearances of the rainbow spectra are different from the appearance of a spectrum made by a prism, and what comes out of a glass prism is different from that produced by a salt prism, which is different from what is created by a diffraction grating,… etc and etc. and so on. The relative widths of different spectral regions really does critically depend upon the dispersion of the medium or the number of diffraction lines and the geometry of the situation.

---

<div class="post-metadata">

**Author:** ![DPRK](https://avatars.discourse-cdn.com/v4/letter/d/4491bb/32.png) [@DPRK](https://boards.straightdope.com/u/DPRK)\
**Post date:** [January 24, 2019, 9:29pm UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/54 "2019-01-24T21:29:11Z")

</div>

> [@CalMeacham](#):
>
> I told you the procedure in my earlier post. Why do you think it’s a “stretch” to “squeeze” an octave in there? You can divide any length by seven or eight.

You certainly can, but that does not mean you can play an entire octave between adjacent C and D on a guitar string. Taking [from Wikipedia] the extreme bounds 380-740 nm on the visible spectrum, they _nearly_ make an octave, just a quarter-tone or so short.

From Newton’s description of his procedure and your own description of it, I gather that he projected a spectrum onto the wall through a prism, then asked his friend to delineate both the seven bands and the center of each band, having named the seven colors beforehand. (Also, he only tested one friend?)

---

<div class="post-metadata">

**Author:** ![DPRK](https://avatars.discourse-cdn.com/v4/letter/d/4491bb/32.png) [@DPRK](https://boards.straightdope.com/u/DPRK)\
**Post date:** [January 24, 2019, 9:47pm UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/55 "2019-01-24T21:47:30Z")

</div>

> [@CalMeacham](#):
>
> I told you the procedure in my earlier post. Why do you think it’s a “stretch” to “squeeze” an octave in there? You can divide any length by seven or eight.

You certainly can, but that does not mean you can play an entire octave between adjacent C and D on a guitar string. Taking [from Wikipedia] the extreme bounds 380-740 nm on the visible spectrum, they _nearly_ make an octave, just a quarter-tone or so short.

From Newton’s description of his procedure and your own description of it, I gather that he projected a spectrum onto the wall through a prism, then asked his friend to delineate both the seven bands and the center of each band, having named the seven colors beforehand, including Orange and Indigo. (Also, he only tested one friend?) He concluded that the boundaries between bands “about” divided the interval into certain proportions: 2/9, 1/3, 1/2, 2/3, 4/5, and 7/8 of the length, starting from the violet end, confirming to himself the musical analogy. Do I have all that right?

---

<div class="post-metadata">

**Author:** ![Littleman](https://avatars.discourse-cdn.com/v4/letter/l/ccd318/32.png) [@Littleman](https://boards.straightdope.com/u/Littleman)\
**Post date:** [January 25, 2019, 12:54am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/56 "2019-01-25T00:54:18Z")

</div>

This is all wildly over complicating the question.

Why are rainbows seven colors instead if three primary colors?

Because the three blend , so we get to see secondary colors in rainbows. It’s that simple.

---

<div class="post-metadata">

**Author:** ![SigMan](https://avatars.discourse-cdn.com/v4/letter/s/13edae/32.png) [@SigMan](https://boards.straightdope.com/u/SigMan)\
**Post date:** [January 25, 2019, 1:25am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/57 "2019-01-25T01:25:11Z")

</div>

Hah! I was right. 😃

---

<div class="post-metadata">

**Author:** ![Chronos](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/chronos/32/134_2.png) [@Chronos](https://boards.straightdope.com/u/Chronos)\
**Post date:** [January 25, 2019, 1:25am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/58 "2019-01-25T01:25:25Z")

</div>

Rainbows aren’t seven colors _or_ three colors; they’re an infinite number of colors. And those infinite number of colors aren’t from blends of some finite number; the finite number is from the limitations of how human eyes work.

---

<div class="post-metadata">

**Author:** ![Jinx](https://avatars.discourse-cdn.com/v4/letter/j/c6cbf5/32.png) [@Jinx](https://boards.straightdope.com/u/Jinx)\
**Post date:** [January 25, 2019, 1:30am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/59 "2019-01-25T01:30:51Z")

</div>

Just wanted to check in. I don;t have time right now to read everything. I will report back with my thoughts ASAP.

---

<div class="post-metadata">

**Author:** ![Jinx](https://avatars.discourse-cdn.com/v4/letter/j/c6cbf5/32.png) [@Jinx](https://boards.straightdope.com/u/Jinx)\
**Post date:** [January 25, 2019, 1:33am UTC](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243/60 "2019-01-25T01:33:28Z")

</div>

> [@Littleman](#):
>
> This is all wildly over complicating the question.
> 
> Why are rainbows seven colors instead if three primary colors?
> 
> Because the three blend , so we get to see secondary colors in rainbows. It’s that simple.

Quick reply because your comment caught my eye: This cannot be correct. I have seen a simple demonstration at the Benjamin Franklin Institute in Philadelphia (once upon a time when this museum really cared about science). When one blends the three primary colors of light, one creates white light.

[Previous page](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243.md?page=2)

[Next page](https://boards.straightdope.com/t/rainbows-why-not-just-three-colors/828243.md?page=4)
