science:

Meet HM. You may have heard of him: his name was Henry Gustav Molaison, and he is one of the most famous patients in the history of medical science. It would probably be fair to say that he is the single individual who has contributed most to our understanding of how memory works. When he died in 2008, the New York Times went with the title “an unforgettable amnesiac” for his obituary.

The young HM was plagued by incapitating epileptic seizures. After all conventional treatments had been exhausted, he submitted to an experimental procedure performed by the neurosurgeon William Beecher Scoville, wherein large parts of his medial temporal lobes, including the majority of his hippocampus and amygdala, were removed. After the surgery, his epilepsy was reduced, but he had also, in addition to forgetting some years before the operation, lost the ability to form new memories. The first paper on HM was published in 1957, a few years after his operation. HM spent the last fifty years of his life willingly letting himself be studied by scientists.

One of the most profound lessons scientists have learned from the study of HM’s case is that there are different types of memory, centralized in different parts of the brain, and it’s possible to lose one form while preserving the others. While HM couldn’t recall things that had happened to him since the surgery, or learn new facts or definitions, he did show improvement on learning skills. After repeatedly practicing a task where he would draw a figure while looking at his hand and the paper in a mirror, HM could not recall having ever practiced, but he did show marked improvement. His short-term memory was also intact; he performed no worse on tests of short-term memory than control subjects, and his scores on intelligence tests actually increased after his operation. Scientists now divide long-term memory into two kinds: declarative memory (episodic and semantic memory), which is apparently highly dependent on the hippocampus, and procedural memory, which is located elsewhere.

HM practiced a task where he navigated a maze with a stylus. His number of error didn’t decrease with practice, indicating that he was unable to recall the correct route; he was, however, able to reduce his times, indicating that he could learn the motor skills necessary to perform the task. And although he couldn’t learn a spatial layout in the lab, he was able to draw an accurate floor plan of the house he had moved into after his operation—apparently, his spatial memory wasn’t completely gone, and moving around his house every day allowed him to learn the layout.

In another experiment, HM studied magazine pictures, and showed normal recall relative to healthy controls (who had studied the pictures for a shorter time period). The researchers argued that HM couldn’t consciously recall the pictures, but could make limited judgments based on familiarity. Their hypothesis “is that conscious recollection of the learning episode depends on the hippocampus, whereas familiarity judgements without episodic content rely on perirhinal cortex.”

Word stem completion is a task where subjects first read a list of words, and then later complete a series of word stems with the first word that comes to mind. Studies show that people are more likely to choose words that have been “primed” beforehand: if you read “thing”, you’d be more likely to later complete the stem “TH” with “thing” than with “thumb” or “thong”. The task tests unconscious memory. HM showed normal responses when primed with words he had learned before his amnesia. However, when primed with newer words that entered use after his surgery, HM didn’t respond to the priming. The hypothesis goes that priming activates existing memories, thus making us more likely to recall them later, but HM had no representation of the newer words, and thus couldn’t recall them.

Interestingly, although HM had learned next to nothing about popular culture, politics or public figures after his surgery, he was able to name John F. Kennedy and Ronald Reagen when looking at their photographs. He could also, after receiving phonemic cues (e.g., M.T.) name other people, like Mao Tse-Tung, although he couldn’t identify famous faces based on semantic cues (e.g., he was a leader in China). Perhaps what little was left of HM’s medial temporal lobe could still function, but what he could learn was clearly very fragmented and incomplete.

The picture that emerges from the study of HM and other amnesic patients is that memory is a lot more complex than previously thought. It consists of a series of related processes that depend on different areas of the brain.

Henry Molaison died in December, 2008. His brain, which he had agreed to donate to science, is now stored in a thousand slices at the University of California, San Diego. Although he still couldn’t recognize researchers who’d studied him for decades, Henry remained dedicated and motivated until the end, optimistic that what science had to learn from him could help others.