Memory and the hippocampus

On September 1st 1953, William Scoville used a hand crank and a cheap drill saw to bore into a young man's skull, cutting away vital pieces of his brain and sucking them out through a metal tube. But this wasn't a scene from a horror film or a gruesome police report. Dr Scoville was one of the most renowned neurosurgeons of his time, and the young man was Henry Molaison, the famous patient known as "HM", whose case provided amazing insights into how our brains work.

As a boy, Henry had cracked his skull in an accident and soon began having seizures, blacking out and losing control of bodily functions. After enduring years of frequent episodes, and even dropping out of high school, the desperate young man had turned to Dr Scoville, a daredevil known for risky surgeries. Partial lobotomies had been used for decades to treat mental patients based on the notion that mental functions were strictly localised to corresponding brain areas. Having successfully used them to reduce seizures in psychotics, Scoville decided to remove HM's hippocampus, a part of the limbic system that was associated with emotion but whose function was unknown.

At first glance, the operation had succeeded. HM's seizures virtually disappeared, with no change in personality, and his IQ even improved. But there was one problem: His memory was shot. Besides losing most of his memories from the previous decade, HM was unable to form new ones, forgetting what day it was, repeating comments, and even eating multiple meals in a row.

When Scoville informed another expert, Wilder Penfield, of the results, he sent a PhD student named Brenda Milner to study HM at his parents' home, where he now spent his days doing odd chores, and watching classic movies for the first time, over and over. What she discovered through a series of tests and interviews didn't just contribute greatly to the study of memory. It redefined what memory even meant.

One of Milner's findings shed light on the obvious fact that although HM couldn't form new memories, he still retained information long enough from moment to moment to finish a sentence or find the bathroom. When Milner gave him a random number, he managed to remember it for fifteen minutes by repeating it to himself constantly. But only five minutes later, he forgot the test had even taken place. Neuroscientists had thought of memory as monolithic, all of it essentially the same and stored throughout the brain. Milner's results were not only the first clue for the now familiar distinction between short-term and long-term memory, but show that each uses different brain regions.

We now know that memory formation involves several steps. After immediate sensory data is temporarily transcribed by neurons in the cortex, it travels to the hippocampus, where special proteins work to strengthen the cortical synaptic connections. If the experience was strong enough, or we recall it periodically in the first few days, the hippocampus then transfers the memory back to the cortex for permanent storage. HM's mind could form the initial impressions, but without a hippocampus to perform this memory consolidation, they eroded, like messages scrawled in sand.

But this was not the only memory distinction Milner found. In a now famous experiment, she asked HM to trace a third star in the narrow space between the outlines of two concentric ones while he could only see his paper and pencil through a mirror. Like anyone else performing such an awkward task for the first time, he did horribly. But surprisingly, he improved over repeated trials, even though he had no memory of previous attempts. His unconscious motor centres remembered what the conscious mind had forgotten.

What Milner had discovered was that the declarative memory of names, dates and facts is different from the procedural memory of riding a bicycle or signing your name. And we now know that procedural memory relies more on the basal ganglia and cerebellum, structures that were intact in HM's brain. This distinction between "knowing that" and "knowing how" has underpinned all memory research since.

HM died at the age of 82 after a mostly peaceful life in a nursing home. Over the years, he had been examined by more than 100 neuroscientists, making his the most studied mind in history. Upon his death, his brain was preserved and scanned before being cut into over 2000 individual slices and photographed to form a digital map down to the level of individual neurons, all in a live broadcast watched by 400,000 people.

Though HM spent most of his life forgetting things, he and his contributions to our understanding of memory will be remembered for generations to come.