He did two crossword puzzles a day, sometimes more, working through the list of clues in strict order, as if to remember where he was.
And, perhaps, what he was doing.
Henry Gustav Molaison — known through most of his life only as H.M., to protect his privacy — became the most studied patient in the history of brain science after 1953, when an experimental brain operation left him, at age 27, unable to form new memories.
Up until his death in a nursing home in 2008, Mr. Molaison cooperated in hundreds of studies, helping scientists identify and describe the brain structures critical to acquiring new information. He performed memory tests; he filled out questionnaires; he sat for brain scans and performed countless research tasks, each time as if for the first time.
In between it all he did puzzles, books upon books of them, a habit he'd picked up as a teenager. Near the end of his life he kept a crossword book and pen with him always, in a basket attached to his walker. His solving opened a window on the brain and demonstrated puzzles' power, and their limitations, in stretching a damaged mind.
‘'For someone with this profound amnesia, the question was: Why, of all the pastimes out there, would he find crosswords so reassuring?" said Dr. Brian Skotko, a clinical fellow in genetics at Children's Hospital Boston. "Well, in a world that was buzzing by and not always so easy to understand, I think finding solutions gave him great satisfaction. He had those puzzle books nearby morning, afternoon, and night, and he turned to them if nothing else was going on. It was his go-to activity."
In a series of experiments, Dr. Skotko led a team from Duke University and MIT that used crosswords to test Mr. Molaison's language aptitude, as well as his capacity to learn new facts.
As expected, the puzzles stumped him badly if they drew on events that happened after 1953, the year that a surgeon removed two finger-shaped slivers of tissue from the brain to relieve severe, chronic seizures. Those slivers included a structure called the hippocampus, which scientists now know is critical for acquiring so-called declarative memories, those concerning people, places, and facts. The Nixon presidency, the gulf war, the rise of Starbucks, the Internet bubble and its bursting — they all came and went without leaving any accessible trace on his memory.
Yet the researchers found that Mr. Molaison was a competent puzzle solver, compared with healthy people his own age, when allowed to draw on what he'd learned in the years before the operation. He knew about the 1929 stock market crash, the Great Depression, and Pearl Harbor. His language skills, once acquired and uploaded from his then-intact hippocampito higher areas in the cortex, were entirely independent, protected from surgical damage.
‘'We found that learned language is a robust dynamic," Dr. Skotko said. "He often came up with answers to the clues that others did not."
Mr. Molaison stunned researchers over the years by learning some new facts, said Suzanne Corkin, a professor in the department of brain and cognitive sciences at MIT who worked with him in the last decades of his life. He knew that Archie Bunker called his son-in-law "Meathead," on the 1970s show All in the Family, one of his favorites. He knew that Raymond Burr played a detective on TV. He knew about the moon landing, she said, perhaps because of his lifelong interest in rockets.
In particular, Ms. Corkin said, he seemed to be able to update pre-1953 memories. To see whether his love for crosswords could help this process, Dr. Skotko, Ms. Corkin and others performed another study, this time testing Mr. Molaison's ability to solve post-1953 clues that had pre-1953 answers. For instance: "Childhood disease successfully treated by Salk vaccine" (announced in 1955); and "Pact, military alliance of 7 European communist nations enacted to counter NATO" (also 1955).
After repeated trials on the same puzzles, the man who lost his memory learned to fill in the right answers.
‘'We found that he could learn new semantic, factual information as long as he had something in his memory to anchor it to," Dr. Skotko said.
Mr. Molaison, who grew up in Hartford, certainly would have known about polio: the city suffered through an epidemic during his childhood, said Ms. Corkin, a fellow Hartford native who, though younger, remembers it herself.
"This would have been an emotional memory; he certainly would have know people who had polio, and maybe some who died of it," she said.
That emotion, she said, seemed to enhance his brain's ability to update previous memories.
At least two processes may have been at work when H.M. learned new facts, Ms. Corkin argues. One is a weak signal from tissue around his hippocampus that survived the surgery; recent research shows that these surrounding areas are active in acquiring new facts. But research also suggests that anytime the brain retrieves a stored memory, it updates and reshapes it. This process, called reconsolidation, occurs in higher cortical areas that were intact in Mr. Molaison's brain. It does, however, apparently depend on the hippocampus to retain the updates. H.M., after he stopped practicing the study puzzles, soon forgot the post-1953 answers he had learned.
Not that he minded. In a 1992 interview, he admitted to Ms. Corkin that he did have trouble remembering things.
‘'One thing I found out is that I fool around a lot with crossword puzzles," he said. "And, well, it helps me, in a way."
It helps you remember? Ms. Corkin asked.
It did, he said. And, he added, "You have fun while doing it, too."