What Colors Tell Us About Philosophy

Consensus Reality is Not True Reality

So one of my very favorite videos on YouTube talks about color. Echo, our lovely tour guide in the world of color mixing, explains that to mix colors with pigments, we need magenta, cyan, and yellow rather than red, yellow, and blue:

If you’ve never encountered this before, it’s definitely an interesting presentation. And poor Echo is so very annoyed that not everybody knows this: “you were lied to as a child.” But let’s not be sloppy here—what she really means is that she was lied to as a child. This comes through pretty clearly for anybody who watches the video; most people are taught, as she clearly was, that red, yellow, and blue pigments can be mixed to make any color of the rainbow, when simple testing reveals that it’s actually magenta, yellow, and cyan. Basic subtractive color mixing.

But for me, it’s not the lecture, it’s the frustration she feels that I notice first, because that’s the way that I felt after discovering these things on my own in college, excitedly showing my design and color teacher a vibrant wheel I had mixed from cyan, yellow, and magenta, and having him respond,

“No Apple Pie, that’s only with printing inks.”

I don’t have the original anymore. There were a few more intermediate colors there, but this gives the gist

I was holding these colors directly in front of my teacher’s face—not in the midst of lecture where the attention of the entire class would have been on him, not in a potentially embarrassing situation where his professional reputation was on the line and the need to save face could reasonably have overridden what his eyes were telling him. I was saying that I had “figured out why” he’d required us to purchase magenta paint, and that I had mixed red using magenta and yellow. I clarified that this wasn’t something I printed, but rather that I had painted it the previous evening using the standard pigments of the course. And well, he repeated that this was impossible, and that magenta and yellow only made red with printing inks.

My response was, basically, to drop out of school and spend the next few years obsessively picking through boxes of crayons for the purest magenta, yellow, and cyan in order to mix the richest and most vibrant red, green, and blue that I could.¹

Image courtesy of Anthony

It wasn’t until years later, when I went to university as a physicist, that I found everybody in the department took for granted that magenta, yellow, and cyan were it; subtractive color theory was posted on the walls, along with discussions of light wavelengths and diagrams of Snell’s Law. Suddenly consensus reality was admitting that the obvious way to mix color using magenta, cyan, and yellow was the right way, while shrugging about its triviality.

But once I’d completed my masters’ degree and returned to the regular world, this alignment of consensus reality with experimental reality vanished. Occasionally I’m able to just browbeat people into not arguing with me about stuff because I Am A Physicist and I Have A Degree. But when I try to actually convince laypersons by being gentle and patient—people like my son’s art teacher, or my distant relatives, or just people I talk to around—I find I’m not always able to make any headway at all.

Trying to make the issue as cut-and-dried as possible, I ordered a few plastic films with well documented transmittance spectra, and showed people how you could overlay magenta over yellow to get red, like this:

This worked for most people, but the films in real life are somewhat reflective, and show more yellow or more magenta depending on which way you hold them. I wasn’t able to be fully convincing to everybody I showed this to. People had little experience with plastic films, I realized; maybe the situation felt like a setup.

So thinking I was learning, I tried getting colored pencils or crayons and mixing yellow and magenta right in front of people, only to have some of them say the result is not red, but pink, purple, or orange. Thinking I was still learning, I mixed yellow and magenta in secret, showed them the result, asked them what color it was, waited for them to say “red,” and then had them closely examine the paper to see the bits of unblended magenta and yellow crayon wax at the fringes of the big red spot on the page, only to have some people say, “well maybe it’s orange; red is a shade of orange.”

Finally I realize what I really have here is a series of informal tests for dogmatism. Some people are so bad at understanding very simple sense impressions without a culturally agreed-upon framework that they trust the framework rather than their lying eyes.

The Perplexity of Consensus Realty

At the time I’m writing this, AI is all the rage; the latest search engine, Perplexity, is evidently poised to become the ultimate search engine: “Perplexity AI’s challenge to Google hinges on something simple but tough—being the best.” But it didn’t take me very long to determine Perplexity has a certain limitation, and it’s the same limitation Google has:

No, Perplexity, the three primary colors are not red, blue, and yellow; try reading your first two links

No, changing the order to red, yellow, and blue doesn’t solve the problem

For comparison, duckduckgo isn’t helpful enough to hand the user the answer, but it does get it right with both of the first two hits:

Not only is consensus reality confused about color mixing, it’s also confused about how good duckduckgo is compared to alternatives

The wikipedia article has “Cyan (C), magenta (M), and yellow (Y) are good chromatic subtractive primaries in that filters with those colors can be overlaid to yield a surprisingly large chromaticity gamut” while howstuffworks is more direct in saying “The subtractive primaries are really cyan, magenta and yellow… Other colors can be used as primaries, but they will not produce as wide a range of color mixtures."

For Philosophy to Work, it has to Beat Consensus Reality

I’ve been working lately on trying to figure out what to do to make philosophy work. Obviously plenty of people, like Jesse Tatum, don’t see the same problems with the discipline that I do, or Lorenzo Warby does, or David Stove does.²

But the disagreement within the field, on extremely basic questions, should give even the most optimistic proponent of conventional philosophical methods cause for disquiet:³

Some might argue that answering questions like this is philosophers’ job; unfortunately it’s been a few thousand years now, and they haven’t succeeded yet. In my own attempts to consider possible changes to the way philosophy as a discipline could be carried out, I’ve been appealing to consensus reality, either by suggesting that you’d have to be nuts to abandon the most basic aspects of consensus reality, or casually proposing reasoning like computer programmers and importing consensus reality as a way of initializing philosophical variables.

But, well, consensus reality is pretty clearly bad.

If you want an example of a successful discipline, look anywhere in STEM. The sciences have the advantage of being empirically testable, either directly or, as the models strengthen into respectable theories, indirectly. And mathematics has the advantage of an extremely narrow focus, on an area that begins in an extremely simple way.

STEM fields have a further advantage: a layman of average intelligence can immediately understand and agree with the lowest level findings. This builds trust, confidence, and consensus early on. But the key that makes math and the sciences so successful is that this consensus is built up productively, as science and math make convincing new discoveries.

Is doing it the other way around, and trying to replace a missing philosophical foundation with consensus, and then drawing inferences from there, going to work? So far as I can see, you’ll just end up sounding like my kids’ very, uh, well meaning art teacher, or like a perplexed artificial intelligence. Consensus is, at best, static and conservative; at worst, consensus is windblown, faddish, a castle suspended in the clouds.

Isn’t the entire point of a field like philosophy to generate new knowledge and understanding? In order for a discipline to progress, it has to be able to alter and in some cases completely overturn consensus. Five hundred years ago, we were living on an Earth that was orbited by its own sun and created 6000 years ago; science was able to completely overturn the consensus regarding the ground beneath our feet. When has philosophy ever done anything like this? What would philosophy need to be able to do to accomplish this?

Conclusion

Check for yourselves; if you want to make red paint, you can use yellow and magenta. And check for yourselves; perplexity and google don’t always get things right. This act of checking to falsify a claim demonstrates the power of empirical methods seen in science. So, maybe instead of being concerned with consensus reality, it makes more sense to trust scientific findings:

import Biology

a = evaluate(“Humans evolved from earlier primates”) # returns 1

If you’re going to trust consensus reality before math and science, you’re probably making a serious mistake.⁴

1. Yes, there’s some simplification here, but I’d already been lied to in math class enough times about how “you can’t divide by zero” that my trust in the educational system, and really in consensus reality as a whole, had crumbled away.

2. Stove, D. C. (1991). The Plato cult and other philosophical follies (chapter 7). Oxford: B. Blackwell. Available Online

3. Bourget, D. & Chalmers, D. J. (forthcoming) Philosophers on Philosophy: The PhilPapers 2020 Survey. Philosophers' Imprint.

4. Unless, of course, you are a politician