In 1965, Lawrence Weiskrantz, a neuropsychologist at the University of Cambridge, surgically removed the primary visual cortex from the brain of a rhesus monkey named Helen. The primary visual cortex is the region most associated with visual experience. Without it, Helen should have been blind.
She was. Or she appeared to be. She didn't track objects. She didn't respond to visual cues. She behaved, for months, like a monkey who couldn't see.
Then a PhD student named Nicholas Humphrey began spending time with her. Not testing her — just being in the room. And he started noticing things. Helen avoided obstacles. She reached for objects that were offered to her. She moved through space with a confidence that didn't match her supposed blindness. Over months of patient, informal observation, Humphrey discovered that Helen could navigate her environment with surprising accuracy. She responded to movement, shape, and location. She behaved, in many ways, like a monkey who could see.
But she showed no sign of knowing she could see. No recognition. No engagement with the visual world as something she was experiencing. The information was getting in. The experience of seeing was not.
Humphrey had stumbled onto one of the strangest phenomena in neuroscience. A few years later, Weiskrantz would give it a name — an oxymoron he came up with under pressure when asked for a seminar title. He called it blindsight.
The human who saw without seeing
The phenomenon might have remained a curiosity of animal neurology. It didn't.
Following the findings with Helen, Weiskrantz turned to a human patient known in the literature by his initials: DB. In his early thirties, DB had undergone surgery to remove a growth affecting the visual cortex on the left side of his brain. After the operation, he was blind across the right half of his visual field. He was clear about this. He saw nothing there.
Weiskrantz asked him to do something unusual: guess. Point to where a light was flashed in the blind area. Say whether a line was horizontal or vertical. Tell us whether the shape was an X or an O. DB thought this was pointless — he couldn't see anything. But he agreed.
He was right 87% of the time.
Not vaguely right. Precisely right. In 150 presentations, his finger pointed to locations he insisted he couldn't see with accuracy that stunned both him and the researchers. When shown his own results, DB was baffled. As far as his experience was concerned, he was guessing. He had no visual sensation in the blind field. No flicker, no shadow, no hint of light. When pressed, he described a vague "feeling" that something might be there — not sight, not vision, just a faint sense that he couldn't explain and didn't trust.
Subsequent research with other patients confirmed and extended the finding. Blindsight patients can detect colour. They can discriminate the orientation of lines. They can track movement. They can respond to the emotional expressions on faces they insist they cannot see — distinguishing happy from angry at rates well above chance, without any experience of seeing a face at all.
The visual information is entering the brain. It's being processed accurately. It's producing correct behavioural responses. And the person has no visual experience of any of it.
What this actually means
Notice what you assume about seeing.
You open your eyes. The world appears. You see a room, a face, a colour. You assume the experience of seeing is what makes the seeing work. That awareness is the engine. That the felt sense of "I see this" is how sight happens.
Blindsight says otherwise.
DB's brain processes visual information in his blind field. It locates objects. It discriminates shapes. It tracks movement. It reads emotional expressions on faces. All of this happens accurately. And all of it happens without any visual experience. The processing works. The visual experience is absent.
DB himself is still here. Still conscious. Still a someone. Awareness, as a ground, remains. He thinks, he speaks, he wonders at his own results. What has disappeared is the visual experience in one region of his field. The processing in that region continues in silence — accurate, reliable, and completely dark from the inside.
Blindsight shows that processing and experience can come apart. The brain handles visual information without any of it surfacing as something you see. The work gets done in the dark. And yet awareness remains. DB is still present. The ground in which seeing normally appears is still there. It just receives nothing from that part of the field.
This raises a question. If processing works fine on its own, and awareness is still present regardless, then what is the relationship between them? Why does processing sometimes surface as lived experience — as the felt quality of seeing, hearing, touching — and sometimes stay silent?
What awareness adds
Nicholas Humphrey spent sixty years on this question. He first worked with Helen in 1966. In 2026, he received the Dennett Prize from the International Center for Consciousness Studies for his lifetime of work on the problem.
His answer is careful and specific. What awareness adds, he argues, is not better processing. Blindsight proves the processing works fine without it. What awareness adds is a felt quality — the difference between processing visual information and experiencing sight. Between modelling the world and being someone who sees the world.
Humphrey draws a distinction between two things most people collapse into one: perception and sensation. Perception is the brain modelling what's out there — extracting features, building a map of the environment, generating predictions about what will happen next. Blindsight has perception. It works.
Sensation is different. Sensation is what it's like to perceive. The redness of red. The warmth of sunlight on your face. The felt quality of a sound — not the information it carries, but the experience of hearing it. Sensation is subjective, personal, and qualitative. It's what DB is missing in his blind field. The information arrives. The quality is absent.
Humphrey's argument — developed across decades and laid out most fully in Sentience: The Invention of Consciousness— is that sensation is what creates the self. Not the cognitive self that plans and remembers. The felt self. The sense that there is a someone to whom things happen. A being with a stake in the world. A creature that doesn't just process its environment but cares about it — because the environment is felt, not just computed.
Without sensation, there is function but no subject. Information but no one who finds it interesting, beautiful, painful, or meaningful. The machinery runs. Nobody watches.
Why this matters now
The AI conversation has been asking the wrong question. "Can machines think?" "Can machines be conscious?" "Will AI become aware?"
Blindsight reframes all of this.
Thinking without awareness is not hypothetical. It's documented. DB's brain thinks about visual stimuli in his blind field — processes them, analyses them, produces correct responses to them — without any awareness of doing so. The question isn't whether thinking can happen without awareness. It already does, in a human brain, in a clinical setting, on record.
The question is what awareness adds to thinking.
AI processes language, images, data. It generates responses that are functionally accurate. By every external measure, it performs cognitive tasks. And it almost certainly has no experience of performing them. There is no felt quality of processing a query. No sensation of understanding. No one home.
This is exactly DB's blind field, scaled up. Accurate function. No experience. The processing works. Nobody watches.
If that's all intelligence is — processing without experience — then AI already has it. Blindsight showed that processing and experience can come apart. The question was never about intelligence.
The question was always about the experiencing.
What the experiencing is
This is where the neuroscience runs out of things to say and the contemplative traditions begin.
Humphrey can describe what experience adds — felt quality, subjectivity, the sense that there is a someone to whom things happen. He can show that it's absent in blindsight and probably absent in AI. What he can't explain, within his framework, is why it exists at all. He frames this as an evolutionary puzzle: if the processing works without experience, why did experience arise?
But the question assumes that experience is something the brain produced — that awareness is a biological invention that needs an evolutionary justification. The contemplative traditions start from a different place entirely.
The traditions take a different approach. They don't ask what awareness is for. They ask what awareness is.
And their answer, arrived at through sustained introspection rather than neurology, is that awareness is not a feature of the brain. It's not something the brain produces to improve its processing. It's the ground. The space in which all processing — conscious and unconscious — occurs. Not added on top. Not a bonus feature. The condition that makes experience possible.
In Advaita Vedanta, the word is sakshi — the witness. Not a witness who does anything. Not a witness who improves perception or adds value to cognition. A witness who simply is. Present before the seeing. Present during the seeing. Present when the seeing stops. The witness doesn't need the visual cortex to exist. The visual cortex needs the witness to be experienced.
This is not a claim neuroscience can verify. But it maps, with uncomfortable precision, onto what blindsight reveals. Perception without awareness is possible. The brain can run its programmes in the dark. But the darkness is not the default. The awareness was already here. The brain's programmes — all of them, conscious and unconscious — are occurring within it.
DB's blind field isn't missing something the brain needs to add. It's missing something that was always already there — and that, in the sighted field, makes the difference between computing the world and living in it.
The question you're left with
You're reading this sentence and there is something it is like to read it. Not just processing — experiencing. Not just decoding words — feeling them land. The sentence isn't only being parsed. It's arriving somewhere. In someone.
Blindsight shows that the parsing would work without the someone. The brain could decode these words, extract their meaning, produce a behavioural response — all without any felt experience of reading. The machinery doesn't need awareness.
And yet, here it is. Here you are. Reading, feeling, present. Not because the brain requires awareness to function. But because awareness is what makes functioning into living.
The neuroscience can measure where awareness is absent. It can photograph the brain processing in the dark. What it can't explain is why the light is on in the first place.
The traditions would say: that's because the light was never turned on. It was never off. It wasn't produced by the brain and it won't be extinguished by the brain. It's what was here before the brain, before the processing, before the question about why it exists.
Not a feature. Not an add-on. Not a product of evolution that needs justification.
The ground. Still here. Even now, as you finish this sentence and the words stop and the awareness that was reading them remains — quiet, present, and not in need of an explanation.
Sources and further reading:
- Weiskrantz et al., "Visual Capacity in the Hemianopic Field Following a Restricted Occipital Ablation," Brain (1974) — the original blindsight study with patient DB
- Nicholas Humphrey, "Sentience: The Invention of Consciousness" (2022) — on sensation versus perception and the evolutionary purpose of felt experience
- Nicholas Humphrey, "How Blindsight Answers the Hard Problem of Consciousness," Aeon (January 2026) — the Dennett Prize essay on blindsight and the birth of the self
- Humphrey, "A Blind Monkey Who Sees Everything," doctoral work at Cambridge (1966-1972) — on Helen the monkey and the discovery of residual visual capacity
- Weiskrantz, "Consciousness of the First Order in Blindsight," PNAS (2010) — on Type 1 and Type 2 blindsight in patient DB