Henry Molaison and the Two Memory Systems

On September 1, 1953, neurosurgeon William Beecher Scoville removed large portions of Henry Molaison's medial temporal lobes at Hartford Hospital. The surgery cured Molaison's debilitating epileptic seizures. It also rendered him permanently unable to form new conscious memories. From that morning on, every conversation, every person he met, every new experience vanished within minutes.

What he could do was stranger. Over years of testing by psychologist Brenda Milner, Molaison learned to trace a star using a mirror — a task that requires repeatedly correcting for the reversal of movement. Each session, when given the mirror and the star, he would say he had never done this before. Yet his performance improved trial by trial, session by session, exactly as a person with normal memory would improve. The skill was accumulating in a system his conscious mind could not access.

This was the first clean proof that memory is not one thing. Declarative memory — explicit knowledge of facts and events — depends on the hippocampus and surrounding medial temporal structures. Procedural memory — skill learning, conditioning, habit — does not. Molaison had lost the former entirely while the latter remained intact.

Millner published these findings in 1957, and they restructured the study of memory. Molaison went on to become the most studied individual in the history of neuroscience. He was referred to in the scientific literature only as "H.M." for 55 years, until his death in 2008, when his full name was finally published. He had participated in experiments almost every time Milner asked, never knowing he had done any of them before.