The patient was wide awake. A flap of her scalp had been folded back, a window of bone lifted out, and the soft, glistening surface of her brain lay open under the lights of an operating room in Montreal. She felt no pain; the brain itself has no pain receptors, and the scalp had been numbed with local anesthetic. A surgeon in a white gown touched a thin wire to one spot, then another, sending a few volts through the tissue for a second or two at a time. Each time, someone asked her what she noticed. Then the wire touched a spot deep in the side of her brain, and she began to hum a tune, keeping perfect time. She was not imagining the song, she insisted. She could hear it, played by an orchestra.
The surgeon was Wilder Penfield, and over three decades he and his colleagues did this more than a thousand times. Their purpose was medical: to find the scrap of damaged tissue that set off a patient’s epileptic seizures, cut it out, and leave everything else alone. But in the process Penfield drew the most famous map in neuroscience, a strip of brain that holds a lopsided little body with giant hands and lips. He found the brain’s speech areas by watching patients fall silent mid-sentence. And he reported something that still sounds like science fiction: that a touch of current in the right place could make a person hear a song, see a face, or relive a moment from years before.
If José Delgado’s radio-controlled bull showed in the 1960s that electricity could steer behavior, Penfield’s operating room was where scientists first learned, patient by patient, what electricity could do inside a conscious human mind. His map still guides surgeons, his memory claims turned into one of psychology’s most stubborn myths, and the strip of cortex he charted is exactly where today’s brain implants are placed.
A football coach goes to Oxford
Penfield did not set out to be a brain surgeon. He was born in Spokane, Washington, on January 26, 1891, and grew up mostly in Hudson, Wisconsin. At Princeton he played on the football team, and after he graduated in 1913 the university briefly hired him as a coach. A Rhodes Scholarship took him to Merton College, Oxford, where he came under the spell of the great physician William Osler, who became a lifelong hero, and studied under Charles Scott Sherrington, the physiologist who had spent years mapping the motor areas of monkey and ape brains with electrical stimulation, work that later helped earn him a share of the 1932 Nobel Prize.
The First World War intruded. Penfield volunteered at a military hospital near Paris, and in 1916 he was wounded when the cross-Channel ferry he was riding, the SS Sussex, was torpedoed. He went on to earn his medical degree at Johns Hopkins in 1918, spent a year as a surgical house officer under Harvey Cushing in Boston, and returned to Oxford to finish his studies. He studied brain cells in Madrid with Pío del Río Hortega and, crucially, spent six months in 1928 in Breslau, Germany, with the neurosurgeon Otfrid Foerster.
Foerster was cutting scars out of the brains of wounded soldiers to cure the seizures those scars caused, and he did it with patients awake under local anesthesia, testing the surface of the brain with an electrode as he went. Penfield brought that method home. Later that year he moved to Montreal, at the invitation of McGill University and the Royal Victoria Hospital, and became the city’s first neurosurgeon.
Sparks in a dog’s brain
The idea that the brain’s surface could be tested with electricity was about 60 years old by then. In 1870 two young Berlin doctors, Gustav Fritsch and Eduard Hitzig, showed that a weak current applied to particular spots on one side of a dog’s brain made specific muscles on the other side of its body twitch. It was a shock to a scientific establishment that had long believed the cortex worked as an undivided whole. Within a few years the Scottish physiologist David Ferrier had charted movement areas in monkeys, and Hitzig produced drawings of monkey brains dotted with the points he had stimulated.
The first person to try it on a human being did so in a way that still makes ethicists wince. In 1874 a Cincinnati physician named Roberts Bartholow treated Mary Rafferty, a 30-year-old woman with a cancerous ulcer that had eaten a hole about two inches across through her skull. Bartholow pushed needle electrodes into the exposed brain and found that current made her opposite arm and leg move. When he turned the current up, she convulsed and fell into a coma, and she died a few days later. Bartholow said she had consented, though he himself described her as “feeble-minded,” and the American Medical Association censured him.
Surgeons gradually found safer ways to use the technique. In 1909 Cushing, Penfield’s future teacher, stimulated the strip of cortex just behind the brain’s central groove in two conscious patients and found that it produced sensations rather than movements. Foerster tested hundreds of patients in Germany. What nobody had done was turn those scattered observations into a systematic map of the human brain. That became Penfield’s project, almost by accident, because his real goal was to stop seizures.
A tumor in the family
Before he could build an institute, Penfield faced the hardest case of his career. His only sister, Ruth Inglis, had suffered occasional seizures since childhood, and by her early forties she had swelling at the back of her eyes, a sign of pressure inside the skull, and weakness on her left side. In December 1928 Penfield opened her skull himself and found a large tumor in the right frontal lobe. Ruth was awake throughout most of the six-hour operation, apart from a brief, frightening hemorrhage, and at the end, he later wrote, she “apologized for having made so much trouble!”
To get as much of the tumor as he could, Penfield removed a substantial part of her right frontal lobe. Ruth went home and at first seemed well, though she later found it hard to plan and carry out tasks she had once done easily, such as getting a simple supper ready for a guest. The tumor came back. Cushing operated on the recurrence, without success, and she died in 1931. In 1935 Penfield and his colleague Joseph Evans published her case in the journal Brain as “Case 1” of a study of large frontal lobe removals, without saying that the patient was his sister. Like the iron bar that went through Phineas Gage’s head almost a century earlier, her case showed how much a person could lose from the front of the brain and still remain recognizably herself.
The Neuro
Penfield had dreamed of an institute where surgeons, neurologists and laboratory scientists worked side by side, and the Rockefeller Foundation agreed to fund it. On September 27, 1934, the Montreal Neurological Institute opened on University Street, on the slope of Mount Royal, with Penfield as its first director. The province of Quebec, the city of Montreal and private donors paid for the hospital side. Montrealers took to calling it simply “the Neuro.” In his opening address Penfield hoped it would mark “the birth of a Canadian School of Neurology,” and the same year he became a British subject, as Canadians then were.
The treatment that made the Neuro famous became known as the Montreal procedure. A patient whose seizures could not be controlled with drugs was taken to the operating room, given local anesthetic, and kept conscious while Penfield opened the skull. He and the physiologist Herbert Jasper, who joined the institute and recorded the brain’s electrical activity directly from its surface, looked for the patch of cortex where abnormal discharges began. Penfield then probed the area with a stimulating electrode. Often a patient’s seizures began with a warning sensation called an aura, such as a smell, a feeling in the stomach or a flash of light. If stimulating one spot brought back that familiar aura, he had probably found the source.
Just as important was finding what not to cut. Before removing anything, Penfield stimulated the surrounding brain while the patient moved, felt, talked or counted aloud. If a touch of current made a hand jerk, or stopped the patient from naming a picture, that tissue stayed. Each site was marked with a small numbered paper ticket laid on the brain’s surface and photographed, and a secretary in the gallery took down what the patient said. Those numbered tickets and verbatim notes, piled up over decades, became one of the richest records ever made of how the human brain works.
Generations of Canadians know a dramatized version of the procedure from the one-minute Heritage Minute film above, in which a patient cries out, “Dr. Penfield, I can smell burnt toast!” The film is a dramatization, but smells really were among the auras Penfield chased, and a patient’s report of one at the right moment told him where the trouble lay.
The little man in the cortex
By 1936 Penfield had stimulated the motor and sensory strips of 126 patients. With his resident Edwin Boldrey, he went through the operating notes, photographs and sketches, plotting every point that had produced a movement or a sensation in some part of the body. Boldrey drew about 170 summary maps by hand. On December 1, 1937, the two men published the results in Brain in a 55-page paper, “Somatic Motor and Sensory Representation in the Cerebral Cortex of Man as Studied by Electrical Stimulation.”
The pattern was orderly but strange. Along the strip in front of the brain’s central groove, the motor cortex, the body was laid out upside down: toes and legs near the top of the head, then trunk, arm and hand, then face, lips, tongue and throat lower down the side. The sensory strip just behind it followed much the same order. And the amount of brain devoted to each part had nothing to do with its size. The hand, and especially the thumb, took up a vast stretch of cortex. So did the lips and tongue. The back and the thighs got barely a sliver.
To make the idea stick, Penfield commissioned a drawing from Hortense Cantlie, a medical illustrator trained at Johns Hopkins under Max Brödel. She drew a little man sprawled along the edge of the brain with hands and lips swollen out of all proportion. Penfield called it a “grotesque creature.” It became the homunculus, Latin for “little man,” and in its 1950 form, redrawn by another illustrator for the book The Cerebral Cortex of Man by Penfield and Theodore Rasmussen, it is probably the most reproduced picture in neuroscience. The sculptures at the top of this article, on display at the Natural History Museum in London, show what a man would look like if each part of his body grew in proportion to the cortex that serves it.
The map was never as neat as the cartoon. Penfield and Boldrey themselves noted that the zones overlapped, that many motor responses came from the sensory side of the groove and many sensations from the motor side. A 2017 reanalysis of their data by the neuroanatomist Marco Catani called the homunculus a brilliant teaching aid that hides a messier truth. In 2023 a team led by Evan Gordon and Nico Dosenbach at Washington University in St. Louis reported in Nature, using fMRI, that the strip is interrupted by patches tied not to any single body part but to planning whole-body actions, which they called a “somato-cognitive action network.” The little man, in other words, has gaps in him. But for the purpose a surgeon needs, knowing where the hand and the mouth are before cutting, Penfield’s map still holds.
Penfield and the neurosurgeon Lamar Roberts went on to chart the language areas the same way, summarized in their 1959 book Speech and Brain Mechanisms. A patient would be shown a series of pictures and asked to name each one while the electrode touched different spots. At certain points the patient could not get the word out, or said the wrong word, and recovered the moment the current stopped. Those stimulation-induced “arrests” showed where speech lived in each individual brain, which varied more than the textbooks suggested. That idea, test each brain rather than trust a diagram, is the basis of the awake brain surgery performed in hospitals around the world today.
An orchestra on demand
Then there were the responses nobody expected. Most of the cortex, when stimulated, produced crude sensations: tingling, a flash of light, a buzzing sound, an urge to move. But in a small number of patients, all of them with seizures arising in the temporal lobe on the side of the head, the electrode seemed to call up whole experiences. Penfield called these “experiential responses,” and he spent the rest of his career trying to understand them.
In 1963 he and the neurosurgeon Phanor Perot published what they called a “final summary” in Brain, running to more than 100 pages. They had reviewed the records of 1,288 operations on 1,132 patients at the Neuro since 1934. The temporal lobe had been explored with the electrode in 520 of them, and 40 of those patients, about 7.7 percent, had reported an experiential response. In the other 612 patients, whose operations exposed other parts of the brain, there were none.
The transcripts read like eavesdropping. One young patient, stimulated at one point, said, “I heard someone speaking, my mother telling one of my aunts to come up tonight.” At another point: “My mother is telling my brother he has got his coat on backwards. I can just hear them.” Asked whether it was like a dream, he said no: “It is just like I go into a daze.” One woman, when the electrode returned to a spot that had already produced a tune, said “White Christmas” and that it was the orchestra playing. Another patient heard a song again and again and hummed it so clearly that the operating room nurse recognized it as “Rolling Along Together.” The patient insisted that she was not just thinking of the song but actually hearing it, and when she hummed along, the tempo was exactly right.
One man heard a piano and someone singing. When the same point was stimulated again without warning, he said, “Oh Marie, Oh Marie, someone is singing it,” and explained that it was the theme song of a radio program: “The program is called the ‘Life of Luigi.'” He ended by singing the refrain, and, the paper notes, “All in the operating room recognized the song.” Penfield and Perot also cited a 1958 report from the Mayo Clinic, where electrodes implanted in an 11-year-old boy’s temporal lobe repeatedly brought on the sound of friends “hollering for me across the street. They want me to come and play.” That one experience was reproduced by stimulation 62 times.
Penfield was careful to check that patients weren’t simply responding to suggestion. Sometimes the team announced a stimulation and then didn’t deliver one; the patients usually reported nothing. He also noticed that the responses felt vivid yet strangely partial, as if a single strand had been pulled from the past. They lasted only while the current flowed, and stimulating the same spot a few minutes later could bring back the same scene, or a different one. He described the effect in a public lecture in New York in November 1957, published in the Proceedings of the National Academy of Sciences, and in his Lister Oration in London in 1961, titled “Activation of the Record of Human Experience.”
The tape recorder that wasn’t
That title captures the conclusion that made Penfield famous outside medicine. He came to believe that the brain kept a continuous record of conscious experience, like a strip of film, and that his electrode was simply rolling a stretch of it forward. Only things that had once come “within the spot-light of attention” were stored, he said, but those were stored in detail. Popular books ran with it. The 1960s bestseller I’m OK, You’re OK leaned on his findings to argue that every childhood feeling lay preserved in the brain, waiting to be replayed.
The idea spread through psychology too. In 1980 the psychologists Elizabeth Loftus and Geoffrey Loftus asked 169 people how memory works. Of the 75 with graduate training in psychology, 84 percent chose the statement that everything we learn is permanently stored and that lost details could in principle be recovered with hypnosis or other special techniques. Many cited Penfield as their evidence.
Then Loftus and Loftus went back to his data. Experiential responses had occurred in 40 of 1,132 patients, about 3.5 percent of the whole series. Strip out the patients who heard only music or voices and those whose reports were too vague to classify, they wrote, and “less than 5%” of patients produced the lifelike recollections the work was famous for. Even some of those looked less like playback than like dreams. One woman heard a mother calling her little boy and placed the scene in her own neighborhood; when the same spot was stimulated 18 minutes later, she heard the same voice but said it was at a lumberyard, a place she said she had never been in her life. The psychologist Ulric Neisser had made the same point in 1967: the content of these experiences, he wrote, “seems entirely comparable to the content of dreams.”
Today most memory researchers agree. Remembering is reconstruction, not replay, which is why confident memories can be wrong and why techniques meant to “recover” buried memories can plant false ones. Loftus went on to build much of her career on exactly that problem. Penfield’s patients really did hear songs and voices; modern surgeons have reproduced such responses. What the electrode probably triggered was a mixture of memory, expectation and the brain’s habit of making sense of whatever it is fed.
The patient who couldn’t remember
The most important thing Penfield’s institute discovered about memory came not from what patients experienced on the table but from what some of them lost afterward. In 1950 a young British psychologist named Brenda Milner, studying for a doctorate under Donald Hebb at McGill, began testing Penfield’s patients before and after their operations. Two of them, known in the literature as P.B. and F.C., were left with a severe and lasting inability to form new memories after Penfield removed part of one temporal lobe, including the hippocampus. Milner and Penfield concluded that the hippocampus on the other side must already have been damaged, so the surgery had in effect removed both.
When Penfield and Milner reported those cases at a meeting in the mid-1950s, a surgeon in Hartford, Connecticut, named William Beecher Scoville got in touch. On September 1, 1953, Scoville had removed the inner parts of both temporal lobes from a 27-year-old man with crippling epilepsy, Henry Molaison. The seizures eased, but Molaison could no longer remember anything new for more than a short while. Milner traveled to Hartford to test him, and in 1957 she and Scoville published “Loss of Recent Memory After Bilateral Hippocampal Lesions,” the paper that made the patient known for half a century only as H.M.
Milner showed that H.M.’s intelligence, perception and older memories were largely intact; what was gone was the ability to turn the present into lasting memory. In one famous experiment he learned to trace a star while watching his hand only in a mirror, getting better day after day while insisting each time that he had never done it before. That split, between memories we can consciously recall and skills the body learns without our knowing, became one of the founding ideas of memory science. Milner went on to win the Kavli Prize in 2014 and kept working at the Neuro past her 100th birthday. The lesson for surgeons was immediate: never remove both hippocampi, and test memory carefully before taking out either one.
The map in today’s operating rooms
Penfield retired as director of the Neuro in 1960. He spent his last years writing novels, a biography and an autobiography, promoting second-language learning in early childhood and helping found the Vanier Institute of the Family. In 1975 he published The Mystery of the Mind, in which he concluded, after a lifetime of electrodes, that the mind could not be fully explained by the brain’s machinery alone. He died in Montreal on April 5, 1976.
His methods outlived his theories. Awake craniotomy, with stimulation mapping of movement and language, is now standard for removing tumors and seizure foci that sit close to critical areas. Epilepsy centers routinely lay grids or strips of electrodes on the brain, or thread them into it, and leave them in place for days to record seizures and map function, the modern version of the numbered tickets Penfield laid on the cortex. Researchers still learn from patients in this position. In 2009 a French team reported in Science that stimulating part of the parietal lobe gave patients a conscious urge to move, and at higher currents a false belief that they had moved, when they had not.
The homunculus also turned out to be a useful address for engineers. When the first BrainGate volunteer, Matthew Nagle, moved a computer cursor by thought in 2004 and 2005, the tiny Utah electrode array that read his intentions sat in the hand region of his motor cortex, the region Penfield and Boldrey had drawn with such enormous fingers. The wireless implants now being tested by companies like Neuralink target the same territory, as our Neuralink timeline describes, and implants that help paralyzed patients speak again read the face and mouth areas just below it.
Stimulation went the other way, too. Deep brain stimulation, which quiets tremors by delivering steady current through electrodes placed deep in the brain, grew out of the same mid-century willingness to test the living brain with electricity that drove Penfield and Delgado. Science fiction noticed early: in Philip K. Dick’s 1968 novel Do Androids Dream of Electric Sheep?, the hero and his wife dial up their emotions each morning on a home device called the Penfield mood organ.
Why Penfield still matters
There is a reason the people in Penfield’s notes are remembered mainly by initials and case numbers. They were not volunteers for an experiment. They were sick, often desperate, and they let a surgeon open their skulls in the hope of being free of seizures. Many were helped; the 1963 paper notes, for example, that the woman who heard “Rolling Along Together” had been free of attacks for 12 years and was living “a happy, normal life.” But everything Penfield learned about the waking brain came from them, from the patience of people lying conscious on an operating table for hours, answering the same question again and again: what do you notice now?
Penfield understood the debt. He credited his patients as, in effect, co-authors, and he never pretended the electrode was more than a crude tool for a delicate instrument. What he gave in return was a map. It was imperfect, and his boldest reading of it was wrong. But it turned the brain’s surface from a uniform gray mystery into territory with addresses: here the thumb, here the tongue, here the place where a song might suddenly start to play. More than 90 years after the Neuro opened its doors, surgeons still consult that map before they cut, and engineers consult it before they implant. Most people have never heard of Wilder Penfield, and yet the little man he drew lives in all of us.
Further reading
The Man Who Mistook His Wife for a Hat: And Other Clinical Tales
Oliver Sacks’s classic case stories, including "Reminiscence," about two women who suddenly heard old songs playing in their heads, a chapter that leans on Penfield’s findings.
Patient H.M.: A Story of Memory, Madness, and Family Secrets
Luke Dittrich, grandson of the surgeon who operated on Henry Molaison, tells the story of the most studied patient in memory science and the psychosurgery era around him.
Axis Scientific Deluxe 8-Part Human Brain Model with Arteries
A life-size brain that comes apart into eight pieces, with the frontal, parietal and temporal lobes and the hippocampus labeled, so you can see where Penfield’s map runs.
MUSE 2: The Brain Sensing Headband
A consumer EEG headband that reads your brain’s electrical rhythms from the forehead, the same kind of signal Herbert Jasper recorded straight from the cortex at the Neuro.