Trepanation: The Oldest Surgery

The single most important thing to understand about trepanation is how we can possibly know a Stone Age patient survived it, and the answer is beautifully simple: healed bone. When a hole is cut in a living skull and the person lives, the raw edges of the opening slowly remodel, growing smooth and rounded as the body heals, a process that takes months to years and cannot occur after death. A completely sharp, unhealed edge tells the opposite story, a patient who died during or soon after the operation. This one distinction, healed versus unhealed margins, lets archaeologists read survival directly from a skull thousands of years old. And the practice was astonishingly widespread. Trepanned skulls have been found across Neolithic Europe, in ancient China, in the Caucasus, throughout the Americas, everywhere humans settled except, as far as we know, Australia. Because these cultures had no contact with one another, trepanation was not a single invention that spread; it was discovered independently, many separate times, by different peoples who each arrived at the same extraordinary idea. The oldest securely dated example comes from a Neolithic grave at Ensisheim, in the Alsace region of France, dated by its grave goods to roughly 5100 to 4900 BCE, more than seven thousand years ago. The skull belonged to a man of about fifty who had been trepanned not once but twice; the smaller opening had healed completely, and remarkably, the skull shows no sign of the injury or illness that might have prompted either surgery.

No place on Earth practiced trepanation like ancient Peru, which has yielded more trepanned skulls than the rest of the world combined. The leading modern authority is John Verano, a physical anthropologist who has personally examined more than eight hundred of them, spanning roughly two thousand years from about 400 BCE to the Inca period. His data tell a genuinely stirring story of a craft being learned. In the earliest samples, survival, inferred from healed bone, sits around 40 percent. But it climbs steadily over the centuries as technique improved: to as high as 91 percent in skulls from 1000 to 1400 CE, and averaging 75 to 83 percent during the Inca era. To feel the weight of those numbers, set them beside a far more recent benchmark. Cranial surgery during the American Civil War, in the 1860s, performed with steel instruments and some awareness of cleanliness, killed roughly half its patients, a survival rate of only 44 to 54 percent. Inca surgeons, working with no metal finer than bronze and no germ theory at all, were bringing patients through skull surgery at rates their better-equipped counterparts would not match for centuries. That comparison sounds like exaggeration, and it is not; it has been carefully worked out from the skulls themselves. How they did it is partly visible in their tools. The earliest Andean surgeons cut with blades knapped from obsidian, volcanic glass that fractures to an edge finer than modern surgical steel; later cultures used a curved bronze or copper knife called a tumi. They seem to have worked in the open air and, tellingly, not to have reused tools between patients.

The scientific study of trepanation began with a puzzle and a great anatomist. In 1867 the French scientist Paul Broca examined a trepanned skull from Peru and grasped what the healed edges meant: that prehistoric people had performed surgery, and that patients had lived. Broca and his colleagues went on to establish the whole field, though his own early theory of why, that Neolithic people trepanned to release the spirits believed to cause childhood convulsions, has long since been set aside. It is a useful reminder that our reading of these skulls has shifted enormously over a century and a half, and remains, in places, a work in progress. Two honest cautions matter here. The first is a subtle statistical trap called the osteological paradox: because only survivors leave healed bone, a high 'percentage of healed skulls' measures survival among the operations we can detect, not necessarily every operation ever attempted. The evidence is real, but it must be read with care. The second is a striking absence. Ancient Egypt, for all its documented surgical sophistication, seems barely to have practiced trepanation at all; the anatomist Grafton Elliot Smith reportedly examined some fifteen thousand Egyptian skulls and found no clear cases, and the Edwin Smith Papyrus, Egypt's great surgical text from around 1600 BCE, does not describe the procedure. Trepanation was not universal after all; it flourished in some cultures and was nearly unknown in others, for reasons we still cannot fully explain.

Here we reach the real limit of the evidence, and honesty requires admitting it plainly: for most trepanations, we simply do not know why they were done, and often we never will. A skull can show that an operation happened and that the patient survived, but it rarely reveals the reason. The leading candidates are all plausible and none is provable across the board. Some trepanations sit right beside a healed skull fracture, strongly suggesting they were a response to head injury, perhaps to relieve pressure or clean a wound, and this is the most medically intelligible motive. Others may have aimed at epilepsy, chronic headaches, or conditions we can only guess at. And some were likely ritual or spiritual: in certain cultures the disc of bone removed during the operation, the roundel, was kept and worn as an amulet, which points to meaning well beyond straightforward treatment. The question of pain is just as open. There is no evidence of true surgical anesthesia anywhere in the ancient record. Some researchers suspect Andean surgeons may have used chicha, a fermented maize beverage, to relax the patient; Verano himself doubts that chewing coca leaf, a mild stimulant, offered any real pain relief. With no written records from these cultures, such ideas remain informed guesses, not established facts. The astonishing courage this implies, on the part of patient and surgeon alike, is left almost entirely to the imagination. And the practice was not always safely locked in the distant past: trepanation survived as a living tradition among the Gusii people of Kenya into the twentieth century, documented by a physician-researcher in the late 1950s, performed without anesthesia and, by his record, survived by roughly nineteen of every twenty patients.

Because the facts here are so startling, they attract two opposite distortions, and both should be named and set aside. The first is to over-romanticize the medicine: to take a high survival rate as proof that trepanation 'worked,' that it was effective treatment in the modern clinical sense. It was not, and the distinction is crucial. Surviving the surgery is not the same as being cured by it. There was no anesthesia beyond, at most, a fermented drink, no understanding of germs, no antiseptic technique; and in many cases the operation likely did nothing at all for whatever condition prompted it, even when the patient lived through the procedure itself. A remarkably high proportion of people survived having their skulls opened, which is a genuine testament to skill; it is not evidence that doing so healed them. The second distortion runs the other way, into pseudo-archaeology: the claim that survival rates like Peru's must reflect some lost advanced technology, secret knowledge, or non-human assistance. This is simply false, and it insults the people who actually did the work. There is no mystery requiring aliens or Atlantis here. There is something far more impressive: ordinary human beings, with obsidian and bronze and steady hands, learning a difficult craft by observation and practice across many generations, and getting genuinely, measurably good at it. Modern neurosurgery does still open the skull, in the burr hole and the craniotomy, to relieve pressure or reach the brain. But it does so guided by imaging, under anesthesia, in sterile conditions, for a clear reason established in advance, a world away from the ancient operation. The thread of 'cutting a hole to help the head' runs unbroken across seven thousand years, and the two ends of it are not the same thing at all.