How do we see color? | Hacker News Reader

How do we see color?

kids.frontiersin.org

13 pointslimbicsystem20 comments

the amazing mantis shrimp (see Figure 6) is thought to see the world in more colors than any other animal: some species of mantis shrimp have over 8 different photoreceptors (cones)

Mantis shrimp actually have quite poor color vision despite their exotic equipment.

https://www.science.org/doi/full/10.1126/science.1245824?sid...

Author here: Yup - that's a great paper (published just after the Frontiers piece). There >are< some other organisms that have more than three photoreceptors and seem to use them for 'proper' color vision though: many birds and fish (and perhaps some women).

While this probably means the mantis shrimp only really sees 12 colors instead of millions (one per photoreceptor type), the article notes that this may result in substantially faster color recognition, i.e. trading off fidelity for latency.

For software engineers wanting to know more about this subject I highly recommend this article:

https://en.wikipedia.org/wiki/CIE_1931

Human color vision is a lot more complicated than it seems at first. The sensitivity curves of the actual receptors in the eye do not directly match our physiological perception curves for red, green, and blue. The red perception curve even has some negative (!) values. There is a lot of neurological processing involved and that is why colorspace models are so complicated and imperfect. There is no simple absolute "RGB" mapping like I and many others have naively assumed before digging into it.

The question in my head has always been, is what I'm seeing as 'red' for example, the same that others see? And, in the end, does it matter? I see what I recognize as 'red' and I use the word 'red' to describe what I see. You see what you perceive as 'red' and use the same word for it. We would have a shared word for the color, but what we are actually seeing is not the same at all.

These are the things that make my brain hurt.

The question in my head has always been, is what I'm seeing as 'red' for example, the same that others see?

Yes, unless you have some kind of color vision deficiency. The human visual cortex has little differentiation among individuals.

https://journals.plos.org/plosbiology/article?id=10.1371/jou...

The question you're responding to is (although of little if any practical importance) a more philosophical one than can be answered by cortex differentiation, which cannot tell us much at all about the experience of qualia.

which cannot tell us much at all about the experience of qualia.

Same brain, same qualia. Your emotional response to the color may be different but the color is the same.

The quantifiable characteristics of the light spectrum are the same--the experience of the qualia are not necessarily the same. This is common with taste and smell; it could easily be the same story with the experience of color.

This feels like a "god of the gaps" religious argument. Every time we learn more the degree to which two people's perception of color can vary shrinks as we find them functionally identical at every level of our understanding. Will it take the invention of a machine that can perfectly read a human's mind to convince armchair philosophers there are no more shadows for it to hide in?

Your analogy doesn't work and I suspect it's because you may fundamentally misunderstand what qualia actually are--this isn't about gaps; this relates much more to the logical impossibility of reducing qualia to something that can be understood through quantitative methods. You dropped the key in the dark and are looking for it under a beam of light three feet away by trying to claim that any any amount of neuroscience could answer a question of this nature.

That said, much like the problem of induction, its insolubility is not necessarily of much practical importance day-to-day. We know that certain wavelengths will be recognized by people with normal visual faculties as "red"; whether the way they experience that qualia is different or not is not of much practical importance. It's more of a reminder of the limits of our empirical knowledge.

What is the difference between something that is undefinable and something that is fictitious?

The answer is very probably but there is no possibility to be certain

I only recently discovered the term for this is quaila, https://en.wikipedia.org/wiki/Qualia.

Qualia are one of the more interesting subjects in the study of mind because they're inextricable from our perceptions and our experience of everything, but they can't be reduced or quantified in any meaningful way.

(If anyone is tempted to respond to this with anything from Daniel Dennett, please consider citing a serious philosopher instead.)

What if we all have the same favourite colour, but we see them differently.

There might be an analogy to AI here: suppose you take an embedding network (e.g. word2vec) and instantiate two instances with random weights. You then train each network on the same training data up to the same level of accuracy.

Given any particular pair of words (e.g. "red", "book"), the models should produce pairs of vectors that have similar similarity scores, and thus one could argue that they have the same "understanding" of the word "red" and its relationships with other words. But the actual vectors that the two model instances output for the word "red" will be totally different, and entirely incomparable, due to different random initializations. (It's overwhelmingly likely that they will converge to different local minima).

Likewise, the actual internal representation of the color "red" inside your brain differs completely from the color "red" inside my brain, in terms of e.g. the set of neurons activated and the weights they use. That effect should run quite deep, and will be influenced by a myriad of life experiences - for example, maybe someone fears red because they saw too much blood as a kid, while someone else loves red because their culture views it as a symbol of luck & prosperity.

It's nice that we have language as a common protocol for these deeply convoluted internal representations, even if it fails to communicate the full nuance of someone's subjective experience.

In short, as long as our perception is based on biological and imperfect systems, these slight differences will exist.

It's only weird if you overthink it though.

If I say 'can everybody see in the dark equally well?', or 'can everyone read at a distance equally well', rather than 'does everyone see the same 'red' as me', I think you'd be much more comfortable with saying 'no'. And yet it's essentially the same question: when exposed to the same stimulus, are people's perceptions the same?

I am partially colourblind. I can tell which colour a red light is, my red t-shirt is. I'd never say that a strawberry looks green. And yet, darker shades I struggle to differentiate between red, green, and brown in a way others don't. So fundamentally, my perception of red must be different to that of a normally-sighted person.

Don't lose sleep over it. The answer is 'no, everyone sees their own, unique version of the world, but it doesn't really matter'.

Color, Color Science, color vision, reproduction, perception, etc. often surprises engineers when they realize how incredibly complex the subject can be. Many years ago I took a course on Color Science at the Rochester Institute of Technology. I was working in a domain where I realized I had to know more about this subject than I could learn on my own at the time. This has come up many time during my career (even today). I am so glad I took the time to learn more about this from the folks who, quite literally, wrote the books on the subject (Fairchild, etc.).

Here's a fun video to watch:

https://www.youtube.com/watch?v=mf5otGNbkuc

Another interesting topic to understand:

https://en.wikipedia.org/wiki/Metamerism_(color)

Interesting to also consider the science of Optical Illusion in response to how we see color.

See 151 Illusions & Visual Phenomena with explanations by Michael Bach

[1] https://news.ycombinator.com/item?id=25045392

Reply on news.ycombinator.com