A child is standing beside a water pump. One hand is held beneath the spout while, into the other, someone traces the same movements again and again. The water is cold and continuous. The movements against her palm are familiar by now, but their purpose is not. She has learned to reproduce them with her own fingers without understanding what they contain.
The person beside her persists. The shapes are traced again while the water runs over the child's other hand: W-A-T-E-R. This time something happens between the coldness rushing over one palm and the movements arriving in the other. They are no longer separate sensations. One belongs to the other. The thing pouring over her skin has a name.
Nothing has been added to the physical world. The water is the same water it was moments before; the child's hand has not changed. Yet something immense has happened to the world available to her.
The child was Helen Keller. She was not yet seven when, in 1887, Anne Sullivan began the extraordinary education that would become inseparable from both their lives. Keller had lost her sight and hearing after an illness in infancy. In the years before Sullivan arrived, she had developed signs for some of the people and things around her, but she did not yet possess language in the form that would later allow her to read, write, study and eventually graduate from Radcliffe. Sullivan, herself visually impaired, had no formal teacher training when she undertook the task many believed impossible.
The story of the water pump is so famous that it has hardened almost into parable, with Sullivan as miracle worker and Keller as the child rescued from darkness. Their real lives were far more complicated than that story allows. Sullivan's biographer Kim Nielsen cautions against reducing their relationship to the celebrated moment at the pump, noting that it developed into a complicated lifelong relationship of friendship, dependence and mutual influence. But stripped of the mythology, something extraordinary remains. A sensation already familiar to Keller became attached to a symbol, and the symbol could travel. Water no longer required water to be running over her hand.
It is an unusual place to begin thinking about perception because sight is absent from the scene altogether. Perhaps that absence also exposes something about the way sighted people habitually imagine knowledge itself. Even when we describe a child who could not see coming to understand the world, the metaphors most readily available to us tend to lead us back to vision.
We are so accustomed to describing knowledge visually that we barely notice ourselves doing it. We see someone's point. An explanation is clear. An idea is illuminating. Ignorance leaves us in the dark. Yet a human world does not have to arrive principally through pictures. Sound, touch, smell, taste, balance, temperature, proprioception and signals from inside the body all contribute to experience. Keller's life makes the sighted metaphor of perception as a window onto reality suddenly seem rather provincial.
There is another difficulty with the window. No human sense opens onto everything.
Our eyes respond to only a small region of the electromagnetic spectrum. Our ears admit only part of the range of vibrations physically present around us. Receptors in the skin, nose, tongue and body are selective too. Perception begins, in other words, with an astonishing amount of exclusion. Long before the brain has interpreted the information reaching it, much of physical reality has never entered the human sensory world at all.
Nor is what remains simply delivered intact to consciousness. Nervous systems integrate signals arriving through different sensory channels and interpret them in relation to context and prior experience. This usually happens so successfully that we do not encounter a collection of wavelengths, pressures, chemical compounds and electrical impulses. We encounter coffee, rain, another person's face, the distant smell of smoke, the warm give of fur beneath an open hand. And then, without warning, the certainty that something behind us has moved.
Colour offers an especially vivid glimpse of this process because the relationship between physical stimulus and experience is not as straightforward as it feels. Magenta, for example, has no single corresponding wavelength in the visible spectrum. It is nevertheless unmistakably real as a perceptual experience. Visual illusions expose similar acts of interpretation: identical shades can appear different in different surroundings, static patterns can seem to move, equal lines can look unequal. The point is not that the brain is hopelessly unreliable. Quite the opposite. An illusion often works because perceptual processes that are ordinarily useful have been presented with an unusual problem.
The world pushes back against us. A closed door does not become open because we perceive it badly, and a cliff does not become less fatal because somebody mistakes its edge. But the physical world and the world as experienced are not identical. The more interesting question is what happens when we stop treating one ordinary sighted human experience as the standard against which every other experience must be measured.
Daniel Kish walks through the world making small clicks with his tongue.
Blind since infancy, Kish uses the echoes returning from those clicks to gather information about his surroundings. To call this simply a replacement for vision is tempting because vision is the sensory language into which sighted people habitually translate space. But Kish himself has had to borrow that language when explaining an experience that does not map neatly onto ours. Talking to Ed Yong, he described density and texture as something like the “colour” of echolocation. A wooden fence and a wrought-iron fence return differently; bushes differ from a hardwood door; a tree can emerge through sound as a solid vertical form topped by something softer. On one walk described by Yong, an unusual combination of concrete, turf and a parked car caused Kish to stop and ask whether someone had parked on the lawn.
What is striking is not simply that a blind person can perform an impressive feat. Human echolocation has a history stretching back centuries, long before researchers understood what practitioners were doing. Some blind people who detected obstacles through reflected sound could not themselves explain the mechanism; the ability was once called “facial vision” or an “obstacle sense.” What interests me more is the problem Kish and Yong encounter when they try to describe it. One does not know what seeing is like. The other cannot know what sonar is like for Kish. They share a language, but the experience itself will not cross intact.
Perhaps this is one of language's strangest properties. Words allow experiences to travel astonishing distances while reminding us that something is always lost in transit.
Once Keller understood the word had meaning, the water no longer had to be running over her hand for it to be present in thought. Kish can tell a sighted writer that textures have something like colour. More than a century after Keller’s childhood, and far from the streets Kish navigates by sound, their experiences can still reach us through language. We can form some idea of what each describes, even though we cannot enter either experience directly.
Even more strangely, the language available to us may influence which distinctions become easy to make in the first place.
English speakers confronted with an unfamiliar smell often become oddly helpless. We say something smells like lemon, smoke, wet earth, petrol, old socks. In some languages, smell is not so linguistically awkward. Research by Asifa Majid, Niclas Burenhult and colleagues on Jahai, spoken by a hunter-gatherer community in the Malay Peninsula, has shown that speakers have abstract terms for odour qualities and can name smells with a facility that challenges the old assumption that olfaction is simply difficult for humans to put into words.
The human nose did not improve on the way to Malaysia. What changed was the linguistic and cultural world in which the nose was being asked to speak.
That does not mean language imprisons people in separate realities, or that learning a word magically creates a sensory capacity. The relationship between language and thought has survived far too much argument to permit anything so simple. But languages do not divide experience identically. They make some distinctions obligatory, some convenient and others cumbersome. A distinction made repeatedly can become something a speaker must remain ready to notice.
Space provides one of the strangest examples.
Stephen Levinson describes sitting at the mouth of a cave in northern Australia after a difficult day-long search through bush. He was disoriented and asked Dan, a speaker of Guugu Yimithirr, to point toward their base. Dan pointed immediately, straight through the hillside. Levinson checked him with a compass and later against maps. He was extraordinarily accurate.
This was not an isolated party trick. Guugu Yimithirr makes pervasive use of fixed directions where an English speaker would often use left, right, in front of or behind. A person might be north of a tree; an object might sit on the southern edge of a table; even the orientation of an image can inherit the directions of the physical world in which the book is being held. Speaking the language therefore requires continuous access to directional information that an English speaker can usually afford to ignore.
Levinson tells another story about a Guugu Yimithirr storyteller named Jack Bambi. Bambi described a shipwreck, including where people moved and where a shark appeared. Years earlier he had told the same story while facing another direction. When recordings of the two tellings were compared, the gestures had changed relative to Bambi's body but remained anchored to the directions of the original event. The remembered world had kept its bearings.
Other languages organise landscapes in ways that likewise refuse to line up perfectly with English categories. Cross-linguistic work on landscape terminology has found variation not merely in what geographical features are called but in the semantic distinctions that recur through place names, movement verbs and directional terms. Some languages map landscape and parts of the human body onto one another in systematic ways.
This is where claims about language shaping reality can become seductive, and where caution matters. Levinson himself considers several possibilities for how language and spatial cognition interact rather than treating speakers as prisoners of grammar. His evidence favours a more complicated position in which linguistic systems select, exercise and partly construct ways of organising spatial information from capacities available to human cognition.
Still, the implication is difficult to shake. Two people may cross the same ground while habitually attending to different relationships within it. One carries left and right with the body; another keeps the larger landscape continuously available.
And then there is the fact that words themselves are sensory objects.
You are looking at marks as you read this. Somehow they are not remaining marks. For a literate reader, black shapes become language almost involuntarily. David Abram, writing about the sensory character of literacy, describes reading as an encounter in which visible marks are transposed into something associated with sound; alphabetic reading recruits relationships between eye, ear, mouth and language that have become so familiar we cease to find them peculiar.
Keller's route was different again. Language travelled through her hand. The same human invention can enter one nervous system through eyes, another through ears, another through fingertips, and somehow arrive as meaning.
Perhaps this is why the temptation to ask which version is the real world begins to feel like the wrong question. None of these people has escaped physical reality. Neither has any of them received all of it. A nervous system works with what its body can detect, what experience has taught it to distinguish, what attention selects, and what culture and language have made available to thought and communication.
But we have still been talking about one species.
Consider a flower.
It may be yellow to us, or white, or blue. We can lean towards it, inspect the petals, smell it, touch it. We might reasonably feel that we have encountered the flower rather thoroughly.
Then a bee arrives.
Now the limits become harder to ignore. Many flowers display ultraviolet patterns invisible to human eyes but available to bees and other insects. The markings were not hidden until the bee arrived. They were present while we were staring directly at the flower.
And ultraviolet is only the beginning.
Flowers exist within Earth's atmospheric electric field. Because they are grounded and bees can acquire positive charge while flying, the meeting between bee and flower is electrical as well as visual and chemical. Experiments led by Daniel Robert found that bumblebees could learn to distinguish artificial flowers using electric cues alone, even discriminating differently shaped electric fields. The tiny hairs covering a bumblebee's body can be deflected by those fields and contribute to their detection. Real flowers, Yong writes, are surrounded by distinctive electric patterns that are as invisible to us as their ultraviolet markings.
The flower has been carrying all of this while we stood beside it.
What, then, is the flower? The object visible to us? The ultraviolet pattern visible to the bee? Its odour? Its texture? Its temperature? Its electric field? The chemical information available to another creature? All of these belong to the same physical thing, but no single observer necessarily receives them all.
The biologist Jakob von Uexküll gave a name to the perceptual world available to an organism: its Umwelt. Ed Yong uses the idea throughout An Immense World to lead readers through sensory worlds that overlap geographically while differing radically in what they contain. Some animals detect electrical fields; some use reflected sound; some are exquisitely sensitive to vibrations travelling through surfaces; some respond to ultraviolet or polarised light. Magnetoreception may provide information from Earth's magnetic field, although precisely how that information is experienced by different animals remains extraordinarily difficult for humans even to imagine.
That last difficulty is not incidental. Yong describes magnetoreception as especially counterintuitive because, unlike echolocation or electroreception, it resists easy comparison with familiar human senses. Researchers themselves struggle to imagine the perceptual experience they are trying to investigate. Yong makes a quietly unsettling observation: a scientist's questions are guided by imagination, and imagination itself is constrained by the senses the scientist possesses. Our own Umwelt can therefore limit not merely what we perceive, but what we think to ask about another creature's.
For most of human history, there must have been realities around us for which we possessed neither sensation nor explanation. Ultraviolet light did not begin when humans discovered it. Earth’s magnetic field did not wait for the compass. The electric fields around flowers did not appear in 2013 when scientists demonstrated that bumblebees could discriminate them. They were already here.
This could leave us imprisoned inside our biology, except that humans have developed a peculiar habit of refusing to remain entirely within the limits of our senses. We build instruments.
A detector converts radiation we cannot see into numbers, colours or sounds that we can. A microscope brings structures below the resolution of the naked eye into view. Telescopes collect signals from distances no human eye could bridge. Other instruments register frequencies, fields, particles and chemical traces for which evolution supplied us with no conscious sensation at all. We take something unavailable to the human body and translate it into something the human body can perceive.
In that respect, there is an unexpected kinship between the instrument and the word. Both allow something that cannot arrive directly in a particular form to arrive by another route. The comparison is imperfect: a scientific instrument detects and transforms physical information, while a word is a symbol whose meaning has to be learned. But both can extend what becomes available to a human mind.
A detector can turn invisible radiation into a visible trace. A word can make an absent thing available to thought. Neither contains the thing it represents. The movements against Keller's palm did not contain water, any more than the word ultraviolet contains ultraviolet light. Yet both can point beyond what the body is receiving at that moment.
This returns us to the child at the pump. Sullivan could not give Keller sight or hearing. What she found was another route by which something in one mind could reach another. Once the connection between sensation and symbol was understood, Keller could think about water when no water touched her skin.
Science does something related whenever it gives us ways to think about things we cannot experience directly. Ultraviolet. Ultrasound. Magnetic field. We can measure them, describe them, build mathematical models of them, translate them into images and sometimes manipulate them with extraordinary precision. Language and instruments together can carry us astonishingly far beyond the limits of unaided perception.
But knowing that something exists is not the same as living inside the experience of it.
I can know that a bee detects information at a flower that I cannot. I can look at an image in which ultraviolet information has been translated into colours my eyes can see. I can understand the wavelengths involved. None of this tells me what ultraviolet looks like to a bee, if “looks like” is even the right way to ask the question.
Yong encounters the same boundary with Kish. They can speak to one another. Kish can reach for visual analogies; Yong can reach for imagined acoustic ones. But neither can hand the other his experience. Their shared language takes them remarkably far and then stops.
There is something both humbling and beautiful in that limit. The world available to us is not false because it is partial. Keller's water was real. Kish's returning echoes are real. The directions held in a Guugu Yimithirr speaker's memory are real. The colour of the flower is real to us, and the ultraviolet and electrical information available to the bee belong to the same physical encounter. What changes is what can enter a life.
Perhaps that is the more interesting meaning of The World Your Brain Gives You. A brain does not sit alone inside a skull inventing a private universe. It belongs to a body with particular senses, a history of learning, habits of attention, a language, a culture and an environment. It receives some things readily and others not at all. It learns distinctions. It predicts, combines, remembers and sometimes misreads. Through language, technology and imagination, it can reach beyond some of the limits with which it began, but never all of them.
There are wavelengths passing through the world that no unaided human eye will see, frequencies we will never hear, fields we do not consciously feel, chemical messages we cannot read and sensory distinctions for which we may have no experience at all. Some are ordinary features of another creature's day. Others we have learned to detect only by building something that translates them for us. And there may be aspects of physical reality we have not yet learned to detect, distinctions we have not thought to make, questions our particular sensory world has not taught us to ask.
The child at the pump could already feel the water. What changed was that another way of knowing it became possible.
More than a century later, we have names for things Helen Keller never knew existed and instruments that reveal worlds no human body could discover unaided. Still, every new means of seeing seems to reveal another edge beyond which we cannot follow.
Perhaps the astonishing thing is not that the brain sometimes fails to show us reality exactly as it is. It is that from such a small opening onto the physical world, we have discovered that the world is larger than the one we can experience.
And it was larger all along.
Sources and further reading
Primary historical material on Anne Sullivan's teaching of Helen Keller, including the April 1887 water-pump breakthrough.
American Foundation for the Blind. Teaching Helen.Read the account →Biographical chronology of Keller's childhood, education and the arrival of Anne Sullivan in 1887.
American Foundation for the Blind. Helen Keller Biography and Chronology.Read the biography →Keller's own account of her childhood, education and acquisition of language.
Helen Keller. The Story of My Life. 1903.Read the book →A biography that complicates the familiar “miracle worker” story and examines the lifelong relationship between Anne Sullivan Macy and Helen Keller.
Kim E. Nielsen. Beyond the Miracle Worker: The Remarkable Life of Anne Sullivan Macy and Her Extraordinary Friendship with Helen Keller.Discussion of Daniel Kish and human echolocation, including the difficulty of translating one perceptual world into another person's sensory vocabulary.
Ed Yong. An Immense World: How Animal Senses Reveal the Hidden Realms Around Us.Earlier reporting on Daniel Kish and the use of reflected sound by blind human echolocators.
Ed Yong. The Brain on Sonar: How Blind People Find Their Way Around with Echoes.Read the article →Research showing that Jahai speakers use dedicated abstract terms for odours and can name smells with a consistency that challenges claims that olfaction is inherently difficult to express in language.
Asifa Majid and Niclas Burenhult. Odors Are Expressible in Language, as Long as You Speak the Right Language.Read the study →Cross-linguistic research on spatial frames of reference, including Guugu Yimithirr direction systems, wayfinding and the Jack Bambi shipwreck recordings.
Stephen C. Levinson. Space in Language and Cognition: Explorations in Cognitive Diversity.Research on the linguistic and cultural categorisation of landscapes across languages.
David M. Mark, Andrew G. Turk, Niclas Burenhult and David Stea, eds. Landscape in Language: Transdisciplinary Perspectives.A philosophical and phenomenological account of perception, literacy and the sensory character of language.
David Abram. The Spell of the Sensuous: Perception and Language in a More-Than-Human World.The experiment showing that bumblebees can detect and learn the electric fields surrounding flowers.
Clarke, Whitney, Sutton and Robert. Detection and Learning of Floral Electric Fields by Bumblebees.Read the study →Research identifying mechanosensory hairs as a way bumblebees can detect weak electric fields.
Sutton, Clarke, Morley and Robert. Mechanosensory Hairs in Bumblebees (Bombus terrestris) Detect Weak Electric Fields.Read the study →The broader sensory-ecology framework used in this essay, including Umwelt, ultraviolet vision, electroreception, echolocation and magnetoreception.
Ed Yong. An Immense World: How Animal Senses Reveal the Hidden Realms Around Us.