# Why a mind that once could see

> What stays when sight goes and what fades — the science and the lives behind the mind ROUTIX Blind is modelled on, and why we chose it.

ROUTIX Blind is modelled on one particular mind: a person who could see, and then lost their sight. Not a camera. Not an eye. Not a mind that never saw. This page is the long answer to why — told through the science of what happens to a mind when the pictures stop arriving, and through the people who lived it.

One line to keep in view as you read. Nothing below measures ROUTIX. The studies describe people and brains, not software. They explain a choice; the only measure of the choice is the film you get, and that is yours to judge.

## Two ways of never seeing

There are two kinds of blindness that matter to this story, and they are not the same.

A person blind from birth has never had a picture. Everything they know of the visible world reached them through words, touch and sound. A person who lost their sight later — as a child, as a student, in the middle of a working life — had the pictures first. The question the science asks is simple: when the sight goes, do the pictures go with it?

The clearest single answer comes from a task where people are asked to hold a pattern of filled and empty cells in mind and work with it. People who can see, and people who lost their sight later in childhood or after, scored alike with people blind from birth or from their first years — but they got there by different roads. The sighted and the late blind said they built a picture of the pattern and simplified it. The people blind from birth or their first years said they stored each filled cell as a position, a pair of coordinates, with no picture at all. Same score. Different mind.

*Source: Vanlierde and Wanet-Defalque, Acta Psychologica, as summarised in [Frontiers in Human Neuroscience](https://pmc.ncbi.nlm.nih.gov/articles/PMC3958697).*

That difference — a picture against a list of positions — is the difference at the heart of ROUTIX's choice. A scene is not a list. A list cannot tell you what the lamp does to the wall.

## What stays

**Colour, lightness, shape and size.** Adults who went blind after having seen were asked, years and in some cases decades later, about the visible world: the shape of things, their size, their colour, how light or dark they are, the shape of letters and words. They answered as well as people who could see. Decades without a single picture arriving, and the knowledge of what red is, how big a horse is, what shape a letter has, stayed whole. That knowledge lasted. One kind of knowledge did not keep as well: faces. We come back to that below, because it matters.

*Source: Dietz, Malaspina, Albonico and Barton, Neuropsychologia — [the abstract](https://www.sciencedirect.com/science/article/abs/pii/S0028393221003638).*

**Depth and perspective.** Ask a person who lost their sight long ago to imagine an object and walk towards it in their mind until it fills their view. They do what the sighted do: a large object is imagined farther away than a small one; it grows as they approach it; there comes a point where it overflows the frame of the mind's eye. Most of those blind from their first years could not do it at all: an imagined thing, they said, can never overflow. Perspective — the thing that makes a scene a scene rather than a diagram — survives the loss of sight when sight was there first.

*Source: Vanlierde and Wanet-Defalque, Journal of Visual Impairment and Blindness — [the paper](https://files.eric.ed.gov/fulltext/EJ683483.pdf).*

**Dreams.** People who lose their sight after about the age of seven usually keep pictures in their dreams; those who lose it earlier keep fewer, and people blind from birth dream mostly in sound, touch, taste and smell — though researchers still debate whether their dreams hold anything visual. People blinded as teenagers or adults go on dreaming in pictures for years; in dream diaries, the pictures thin slowly over a long time. The dreaming brain keeps the pictures it was given.

*Source: Meaidi, Jennum, Ptito and Kupers, Sleep Medicine — [the abstract](https://pubmed.ncbi.nlm.nih.gov/24709309/); an overview at [Medical News Today](https://www.medicalnewstoday.com/articles/do-blind-people-dream).*

**The seeing part of the brain keeps its old character.** In a brain blind from birth, spoken language moves strongly into the areas that would have handled sight; even the grammar of a sentence shows up there. In a brain that lost its sight as an adult, those areas answer language only faintly and never follow its grammar, and the language network stays where it is in everyone else.

*Source: Pant, Kanjlia and Bedny, Developmental Cognitive Neuroscience — [the paper](https://pmc.ncbi.nlm.nih.gov/articles/PMC6994632).*

## Why words alone do not carry colour

Here is the finding that explains the choice best.

Adults blind from birth know a great deal about how animals look. Shape, size, the texture of skin or fur — they agree with sighted people substantially, and on shape almost as closely as sighted people agree with each other, and they get there by inference: a bat is textured like a mammal but shaped like a bird, because of what kind of thing it is and where it lives. The one dimension where they diverge most from the sighted is colour. Colour is the easiest visual fact to put into words, and words carried it worst.

*Source: Kim, Elli and Bedny, Proceedings of the National Academy of Sciences — [the paper](https://pmc.ncbi.nlm.nih.gov/articles/PMC6561279); a plain summary from [Johns Hopkins](https://hub.jhu.edu/2019/05/22/blind-people-animal-appearance).*

Other researchers answered that language carries more than you would think — in their words, "a surprisingly rich repository of visual knowledge". Even so, colour is where words fall shortest, and the next finding shows why.

*Source: Lewis, Zettersten and Lupyan, in the same journal — [the reply](https://pmc.ncbi.nlm.nih.gov/articles/PMC6765286).*

Brain imaging says why. The brain keeps the colour of things in two forms. One is a sensory form, found in people who can see, held partly in the visual areas themselves. The other is a form built from language and thought, held in a region of the temporal lobe, and found in people blind from birth and in the sighted alike. A person blind from birth has the word for red. A person who sees has the word for red and the memory of red — and a person who once saw was given both.

*Source: Wang, Men, Gao, Caramazza and Bi, Neuron — [the abstract](https://pubmed.ncbi.nlm.nih.gov/32386524).*

Put those together and the choice writes itself. A description can tell you where the lamp is. Only a mind that has seen lamplight knows what it does to the wall, to the edge of the table, to the colour of a face turned towards it. A revision model that must get colour and light right from a few words had to be modelled on a mind that owns both kinds of knowing.

## A description becomes a room

Modelling a mind on description only works if a description can become a scene at all. It can, and the science says how.

In one study, people who had lost their sight late, people blind from birth and sighted people in blindfolds learned layouts they could not see — from words, from touch, or by walking. Asked afterwards to scan them in their mind, they moved through them as if they had real distances: farther things took longer to reach. What they had learned had become a place with real shape, not a list of facts.

*Source: Afonso and colleagues, Memory and Cognition — [the paper](https://link.springer.com/content/pdf/10.3758/MC.38.5.591.pdf).*

Even among people blind from birth, those who built a mental image as they listened — and who got about on their own — matched the sighted; those who rehearsed the words as words did worse. In the authors' phrase, the difficulty was "not due to the absence of vision per se" — it was a matter of what the listener did with the words.

*Source: Schmidt, Tinti, Fantino, Mammarella and Cornoldi, Acta Psychologica — [the paper](https://doi.org/10.1016/j.actpsy.2012.11.010).*

That is the premise of ROUTIX Blind in one sentence: a scene told in words to a mind that builds pictures comes out as a room you could walk through — and a mind that once saw also knows what the light in that room does.

## The people who lived it

Studies count. People convince. Each of these is one person, told as one person, not as a rule.

**Zoltan Torey** lost his sight as a young man, in a factory accident in Australia. He was told to live by ear and by hand. He chose the opposite: to keep an inner picture of the world, continuously rebuilt from every sound, touch and smell, and to act on it. He visualised machines from the inside out. He described his inner life as "a kind of continuously-produced film strip". To his neighbours' alarm, he once replaced the guttering on his own roof, alone, at night — that part is his own account, and we leave it as his. He wrote about it all in a memoir.

*Sources: the Radiolab episode [Out of Sight](https://radiolab.org/podcast/308564-out-sight/transcript); Oliver Sacks, "The Mind's Eye" — [the essay](https://www.braillewithoutborders.org/ENGLISH/themindseye.html).*

**Sabriye Tenberken** saw faces and landscapes until she was twelve, then not at all. She went on to found a school for blind children in Tibet. Her imagination stayed not only pictured but coloured: numbers and words carry colours for her, and the colours seem to have grown stronger after her sight went. She builds whole landscapes in her mind — and they can be beautifully wrong. Standing at a famous lake, she pictured it vividly while facing the wrong way.

*Source: Oliver Sacks, "The Mind's Eye" — [the essay](https://www.braillewithoutborders.org/ENGLISH/themindseye.html).*

**Jacques Lusseyran** was blinded at eight in a school accident, and later led a Resistance network in occupied France. Within days of the accident he found what he called an inner light: the light, and all the colours, were still there, inside. The light dimmed, he wrote, whenever he was afraid, angry or impatient, and came back when he trusted again.

*Source: his memoir, excerpted in [The Sun](https://www.thesunmagazine.org/articles/23738-and-there-was-light).*

**Jorge Luis Borges** lost his sight slowly. In a late lecture he described living not in darkness but in a world of colours that had lost their edge: blue and green still there, white gone or blurred into grey, and yellow, above all, staying loyal to him. The world of the blind, he insisted, is not the night. (Accounts differ on which colours went first; his blindness was gradual, and the sequence is his.)

*Sources: [Hektoen International](https://hekint.org/2022/08/08/jorge-luis-borges-brilliant-blindness/); [Antiserious](https://www.antiserious.com/p/shakespeare-is-wrong-the-blind-do-not-see-darkness-borges-on-blindness-fe6e57e94f25).*

**Sargy Mann**, a British painter, went fully blind in his late sixties and kept painting until his death, from memory and touch: a palette laid out the same way each day, and his wife helping him mix the colours. As his sight went, he said, he had to invent new ways of making expressive patterns.

*Sources: [Apollo](https://apollo-magazine.com/sargy-mann-late-paintings/); [Isis Magazine](https://isismagazine.org.uk/2015/07/invisible-to-mortal-sight-sargy-mann-and-the-art-of-darkness/).*

Five people, five roads: by will, by colour, by light, by slow loss, by paint. In each of them the pictures stayed — rebuilt, coloured, lit, in perspective — and in all but Borges, who kept a little sight to the end, they were made, not received.

## The honest limits

If this page said only what stays, you should not believe it. Here is what goes.

**Faces fade.** In the study that found colour, shape and size intact after decades, faces were the exception. People who had not seen for many years still knew what things looked like, yet fell behind the sighted on famous faces — above all on their features. Of everything tested, the face kept least well.

*Source: Dietz and colleagues, as above.*

**The picture can soften.** In one small older study, the larger the share of a life spent without sight, the less a person's imagery leaned on pictures. And in a careful test of spatial imagery, adults who lost their sight late were above chance but below sighted people, however long they had been blind and whichever age they had lost it at; those who leaned on words did worse, and those with a good sense of direction did better. A mind that once saw is not a camera. It is a mind.

*Sources: Hollins, Neuropsychologia — [the abstract](https://pubmed.ncbi.nlm.nih.gov/4033908/); Occelli, Lin, Lacey and Sathian, Frontiers in Human Neuroscience — [the paper](https://pmc.ncbi.nlm.nih.gov/articles/PMC3958697).*

**John Hull.** A theologian in Birmingham, fully blind in middle age after years of failing sight, he kept an audio diary of what it was like. Within a few years he reached what he called deep blindness: he could no longer picture even a numeral without tracing it. He chose to stop picturing his wife, because any picture he made of her would be false; of his children, he said he did not know what they looked like. He came to know the world by sound instead; rain, he said, brought out its contours. He is one man, and we tell him as one man.

*Sources: Oliver Sacks's review in [The New York Review of Books](https://www.nybooks.com/articles/1991/04/11/the-dark-paradoxical-gift/); the Radiolab episode [Out of Sight](https://radiolab.org/podcast/308564-out-sight/transcript).*

What the limits teach is as useful as what the gifts teach. The pictures stay when they are used; they soften when they are not. Faces keep least well; the surest face is the one you give. A mind that once saw rebuilds a room from a description, and it rebuilds it wrong if the description is wrong, or missing.

## What ROUTIX took from it

These pages gave ROUTIX no method. They describe the way of thinking [ROUTIX Blind](https://routix.studio/help/why-routix/routix-blind) holds to, and the rules it keeps:

- **Rebuild; never patch.** A mind that once saw does not lay a new picture over an old one. It builds the whole place again, light and all, with the new thing standing inside it. ROUTIX Blind thinks about your scene that way: the whole of it first, then your change inside it.
- **Both kinds of knowing.** The word for red, and the memory of red. Colour and light are not read off your words; they are known, and your words say where they fall.
- **A scene is a room, not a list.** What stands where, how far, in what light, at what scale — held together, so one thing can change and the rest stays whole.
- **Faces are best given.** Because faces are what even a seeing memory keeps least well, the surest way to the face you want is to give it: your own picture, and [Focus](https://routix.studio/help/focus/what-is-focus) when words could mean two subjects. ROUTIX Blind does not pick one behind your back.
- **Use keeps the picture alive.** Every revision begins with the whole stretch, understood anew — never from a memory of a memory.

And ROUTIX Blind does not rest on the literature alone. ROUTIX Blind was developed over two years of research and engineering, with more than 500 people who lost their sight after having seen, and 222 experiments and studies of our own — how they imagine, how they remember, how they rebuild a scene from nothing but words. Three of those experiments are told on the [ROUTIX Blind](https://routix.studio/help/why-routix/routix-blind#the-research) page. The rest stays with us for now.

How ROUTIX Blind does what it does is ours, and this page does not describe it. What this page describes is why it thinks the way it thinks.

## Frequently asked questions

### Does this science measure ROUTIX's films?

No. It explains why we modelled ROUTIX Blind on a mind that once could see; it does not measure what that mind does in your video. Only your film can do that. Bring a clip, change one thing, and judge.

### Why not a mind that was blind from birth?

Because the pictures were never there. People blind from birth know the world richly — but as positions, kinds and words, not as pictures, and least of all as colour. A revision has to get colour and light right. That needed a mind that had seen them.

### Why not simply a sighted mind?

Because a sighted mind argues with what it sees. Point at a tree and say it is a car, and the eyes win. A mind that once saw builds what it is told, whole and lit, without that argument. The story is on the [ROUTIX Blind](https://routix.studio/help/why-routix/routix-blind#the-insight-a-tree-and-a-car) page.

### Are the people on this page typical?

No, and we do not say they are. They are individuals who left accounts. The studies are the measure; the people are the faces of it. Where a finding is debated, this page says so.

### Where does this leave me?

With a clip, we hope. Open **Studio**, press **Revise a clip**, choose a few seconds, change one thing, and watch everything that did not change. That is the mind on this page at work — see [What Optic does](https://routix.studio/help/optic/what-optic-does).

Source: https://routix.studio/help/why-routix/why-a-mind-that-once-could-see · Updated 2026-11-09