Mary Knew Everything About Red. Then She Saw It

Imagine a scientist who knows everything there is to know about the color red, and has never once seen it.

Her name, in the famous version, is Mary. She has spent her whole life in a black-and-white room, learning about the world through black-and-white screens. She is the world's greatest expert on color vision. She knows every wavelength, every cone cell, every neural pathway involved in seeing red: the light around 700 nanometres, the long-wavelength cones that respond most strongly to it, the signals that travel from the retina to the visual cortex. Every physical fact there is to know.

Then one day she is allowed out of the room, and she sees a red rose.

Does she learn something new?

Most people's gut says yes. Now she knows what red looks like. And if she learns something new, then knowing all the physical facts was not knowing everything, so something about experience is not captured by physics. That is the knowledge argument, put forward by the philosopher Frank Jackson in 1982. Philosophers have argued about it ever since, and Jackson himself later changed his mind and rejected his own conclusion.

I spent a while with this puzzle, and I came out with a different reading of it. I think the argument has been fought over the wrong question. My reading is that Mary's room is really about communication, and at the end there is a small test you can try the next time you look at any color.

A Problem in Disguise

The question that struck me was simple:

"Is Mary's room actually a communication problem in disguise?"

Consider what Mary learned in the room. Everything she knew came through human channels: textbooks, equations, lectures, diagrams. Language and mathematics. Those are ways of passing information from one mind to another.

What happens when she sees the rose? Her brain represents red to itself. That is a channel too, just a different one. It is internal, and it has no translation into words, because nobody has ever built a bridge between that channel and language.

So my suggestion is this. Mary does not discover a new kind of fact. She receives a message through a channel she had never been allowed to use before.

Every Channel Is Partial

It becomes more interesting when you push it.

If seeing red is one channel and reading about red is another, what is either of them carrying? What is the thing being communicated?

This is where I landed:

"Everything is epistemic communication because the thing in itself is inaccessible."

Take two cups on a table. You can describe them as an image in your visual field, a sentence, a number, a chemical formula, a quantum state. Each description is real and useful. None of them is the cups. What the cups are beneath all the descriptions, nobody can step outside the universe to check.

Philosophers have a word for that underlying reality: the noumenon, the thing-in-itself, in Kant's language. My suggestion is that everything we call knowledge, including the "hard" knowledge of physics, is a set of representations of it. Some are external and shared, like language and mathematics. Some are internal and private, like the redness of red. None of them hands over the thing itself.

On this reading, Mary was never going to learn "everything" in the room. All she could learn was everything that can be communicated through human channels. The rose opened a second channel.

Why the Brain Makes Qualia at All

This also suggests an answer to a question that has bothered me for years: why does the brain bother generating qualia, the felt qualities of experience?

The brain is a distributed system. Around 86 billion neurons have to coordinate. On this reading, qualia are how the system communicates with itself, packaging information into a format the whole can use. Pain says danger here in a way no list of numbers ever could. Red says this, right here, ripe or bloody or important.

It also explains why a dog and I can look at the same wavelength and have different experiences. Dogs have two kinds of cone where we have three, so we are running different communication systems over the same unknowable source. In the same way that humans invented many languages to talk about the world, nature may have invented many kinds of experience to "talk" about the same underlying reality.

What This Says About the Mind–Body Problem

The usual framing is a war between two kinds of stuff: physical matter and mental experience. How could one produce the other?

If I am right, that framing starts to dissolve. Physical descriptions and felt experience are not two different substances. They are two different channels, two representational formats, both trying to convey something that neither fully captures. The hard problem does not disappear, but it changes shape. Instead of "how does matter make mind?", the question becomes "what is it that every channel is trying to say?"

I will be honest about where this leaves me. I am an idealist; I think consciousness is fundamental. A stricter version of this argument leads somewhere slightly different, a view where the bottom layer is neither "mental" nor "physical" but simply unknowable, and both mind and matter are ways of representing it. Philosophers call something like this Russellian monism. I find myself pulled between the two, and I think that tension is honest.

Have People Thought of This Before?

Probably, in pieces. There are plenty of positions in the philosophy of mind that treat experience as a kind of representation. What struck me was how obvious the communication reading felt once I saw it. The thought experiment itself shows that "external" knowing and "internal" knowing are two different kinds of access, so the difference was sitting in plain sight.

It is an open debate, and I hold this lightly. But I have never looked at a red rose in quite the same way since.

The Message You Can't Forward

The next time you look at a color, any color, try this.

Notice that you will never be able to send that experience to anyone. You can name it, point at it, measure it. They can look at the same thing and have their own version. But the experience itself, exactly as it is for you, stays with you.

I don't think that is a failure of language. I think it is how knowing works: each of us, and each of our senses, carries a partial picture of reality, and none of them delivers reality directly.