Science, Journalism, and the Diverging Narratives of Patient H.M.

When Henry Molaison died at a Connecticut nursing home in 2008, at the age of 82, a front-page obituary in The New York Times called him “the most important patient in the history of brain science.” It was no exaggeration: Much of what we know about how memory works is derived from experiments on Molaison, a patient with severe epilepsy who in 1953 had undergone an operation that left him without medial temporal lobes and the ability to form new memories.

The operation didn’t completely stop Molaison’s seizures — the surgeon, William Beecher Scoville, had done little more than guess at the locus of his affliction — but by chance, it rendered him a near-perfect research subject. Not only could postoperative changes in his behavior be attributed to the precise area of his brain that had been removed, but the fact that he couldn’t remember what had happened 30 seconds earlier made him endlessly patient and eternally willing to endure all manner of experiments.

It didn’t take long for those experiments to upend our understanding of the human brain. By the mid-1950s, studies on Molaison (known until his death only as Patient H.M.) had shown that, contrary to popular belief, memories were created not in the brain as a whole, but in specific regions — and that different types of memories were formed in different ways. Molaison remained a research subject until his death, and for the last 41 years of his life, the person who controlled access to him, and was involved in virtually all the research on him, was an MIT neuroscientist named Suzanne Corkin.

Molaison’s story has always been tinged with tragedy, dependent as it was on his unnecessary disfigurement. But neither Molaison’s history nor Corkin’s role as his scrupulous and caring gatekeeper has ever been controversial.

Or at least they hadn’t been until August 3, when The New York Times Magazine posted a 7,700-word excerpt from “Patient H.M.,” journalist Luke Dittrich’s new book about Molaison and his legacy. (I am a professor at MIT, although I do not know any of the principals involved in this story. I also reviewed “Patient H.M.” for The New York Times Book Review.) Dittrich did not come to this story as an impartial observer: He is Scoville’s grandson, and he spends most of his often gripping book wrestling with his family’s legacy.

The Times excerpt skipped over that compelling intergenerational reckoning and drew instead on the much smaller portion of the book that focuses on Corkin, who, as it turns out, was the best friend of Dittrich’s mother while growing up. As was the case with his grandfather, who Dittrich implies violated the Nuremberg Code governing treatment of human test subjects, this relationship did not result in gentle handling: Dittrich accuses Corkin, who died of liver cancer in May, of a host of scientific and ethical missteps.

The fact that Corkin was unable to defend herself did not mean The Times excerpt went unchallenged. On August 5, 229 scientists signed a letter praising Corkin and calling Dittrich’s account “biased and misleading.” That was soon followed by a response from Dittrich as well as more detailed rebuttals from Corkin’s colleagues at MIT and elsewhere.

Journalists and scientists are both in the business of telling stories, though the tools of the trade are different: Journalists are more likely to rely on apt anecdotes, while scientists prefer to use an accumulation of data. But on a fundamental level, both fields interpret evidence in an effort to make sense of the world.

Ideally, that means journalists and scientists are both striving toward the same goal: the truth. But as the controversy over Henry Molaison’s legacy illustrates, weighing competing narratives is not always straightforward. The facts Dittrich uses to make his case are, for the most part, inarguably true, but is his interpretation of them fair or accurate? And if some of his analysis is off target, does that mean the rest of it should be ignored as well?

Dittrich’s most explosive accusation against Corkin is that she was more concerned with her own reputation than with the advancement of knowledge. His main piece of evidence concerns Jacopo Annese, a neuroanatomist at the University of California, San Diego, who was selected by Corkin to dissect Molaison’s brain — and, in the process, discovered an unsettling complication: a lesion outside the medial temporal region. At some point during their collaboration, Annese and Corkin had a falling-out, and when he submitted a paper about his findings he failed to alert her. This was not a smart move: When Corkin happened to be chosen as one of the paper’s anonymous reviewers, she argued that it did not deserve to be published at all.

In writing his book, Dittrich had extensive access to some materials, but only limited, contentious interviews with Corkin.

Visual: PeskyLIbrarians/Flickr

Dittrich, who had extensive access to Annese and his lab — and only limited, contentious interviews with Corkin — acknowledges that Annese was guilty of “a violation of academic etiquette.” But Corkin’s violation was much more serious, Dittrich writes: She was opposed to publication because she feared that the “discovery of this new lesion” might “require a re-examination of all that old data.”

When I pointed out to Dittrich that by the time Molaison died, the primary importance of the medial temporal lobes in memory formation had been reconfirmed in countless other experiments, he responded by highlighting the comments of another of the paper’s anonymous peer reviewers, who wrote that damage to previously unidentified areas of the brain “may have implications for a retrospective reinterpretation of H.M.’s neuropsychological findings.” But that hypothetical scenario has proved false: In 2014, information about the new lesion was published in a paper on which both Annese and Corkin were listed as authors — and there has been no wholesale reappraisal of the half-century’s worth of studies on Molaison since then.

Dittrich’s other main accusation against Corkin is that she systematically shredded her files on Molaison. At first blush, he would seem to be on solid ground here: She told him so in a tape-recorded interview that he later posted online. Unfortunately, Corkin is also Dittrich’s only source here, and even in the truncated portion of the interview Dittrich posted, it is clear that Corkin is frustrated and annoyed. (The interview came after years of Corkin’s denying Dittrich’s requests for access to Molaison and his records. It also occurred less than a year before Corkin’s death, and in his book, Dittrich set the scene by describing her as “battling a serious illness” and “unusually thin.”) When I interviewed several of Corkin’s colleagues, as well as one of her old assistants, they all said she had been working to catalog her voluminous material on Molaison in the months before she died. In the weeks after The Times excerpt was published, Corkin’s former department says that it conducted its own investigation and that all of the Molaison files appear intact.

Of course, in the absence of witnesses, it’s impossible to know what Corkin did or did not do — so let’s say, for the sake of argument, that she did destroy some of her raw data. Dittrich is on tenuous ground when he claims that that would have violated both “basic research ethics and scientific accountability.” Human test subject data needs only to be kept for between three and six years, according to government guidelines, and Dittrich’s interview with Corkin took place more than six years after Molaison’s death. When I asked him about this, Dittrich emphasized “an important distinction between ethics and the law.” This is a fair point in the abstract, but in interviews with more than half a dozen neuroscientists and several medical ethicists, I couldn’t find anyone who felt Corkin had an obligation to keep all of her data. Arthur Caplan, the director of New York University’s Division of Medical Ethics, said it would be unfair to accuse Corkin of acting unethically if she had failed to do so. “It would be nice if she kept it,” he said, “but really unusual if she did.”

But what of Dittrich’s argument that Corkin was committing a transgression that was “harder to articulate,” namely, destroying important historical documents in an effort to get rid of evidence that might cause future scholars to re-examine her conclusions? The problem with Dittrich’s interpretation is that it depends in large part on his framing of Corkin’s feud with Annese. Absent that, her quote would not have been evidence of a pattern of behavior — and therefore would not have been nearly as noteworthy.

In addition to playing a significant role in shaping Molaison’s scientific legacy, Corkin was also largely responsible for our understanding of him as a human being. She was the person who revealed his identity to the public after his death, and in the coming years, she told the story of his life in numerous interviews and in her own biography of Molaison, “Permanent Present Tense.” The picture she put forth was of a basically happy, mostly gentle man with a sense of humor and an altruistic streak.

Dittrich vehemently argues that this portrait of Molaison is incomplete — and it does seem that, at the very least, Corkin sanded off some of Molaison’s rougher edges when she described him to the world. Unfortunately, in attempting to make a case for a more nuanced understanding of Molaison, the evidence Dittrich provides is often presented out of context. In one glaring example, he lists a fraction of Molaison’s answers to a standard depression questionnaire but fails to provide the actual result of the test. When I asked Dittrich about this, he told me he quoted from the test because Molaison’s description of his state of mind differed so starkly from the portrait Corkin and others had painted of him. “The words were what interested me,” Dittrich wrote to me, “not whatever score researchers derived from them.”

In this same section, Dittrich also selectively — and misleadingly — quotes from Corkin’s work in an effort to show that Corkin had claimed that Molaison showed no signs of any depression whatsoever. What Corkin actually wrote was that Molaison’s test results showed no indication of major depression, a clinical term with a specific meaning. (Dittrich told me that he paraphrased instead of quoted Corkin to tighten the sentence and did not intend to distort Corkin’s statement. He also said he would include Corkin’s full quote in the paperback edition of his book.)

It is, of course, possible that the questionnaire Dittrich quotes from did indicate that Molaison suffered from major depression. Unfortunately, the lack of any source notes, bibliography, or supplementary material in Dittrich’s 411-page book means that readers have no way of knowing why he was confident reaching the conclusions he did. Given that one of Dittrich’s main lines of attack is that Corkin tried to cover her tracks so that her work could not be re-examined, this was a stunning omission — and one that Dittrich, to his credit, acknowledges and intends to rectify. “I think the book’s lack of notes warrants criticism,” Dittrich told me in an email message, “and I regret not having included them in the hardback edition.”

He said he is in the process of assembling his source material, which will be included in the paperback edition and posted online.

A recurrent thread in Dittrich’s attacks on Corkin’s judgment and behavior as a scientist is his use of “common sense” arguments to make specific points about scientific conclusions. This is always a tricky road for science writers to walk: While our primary obligation is to the public and not the specialists whose work we’re translating, that doesn’t get rid of our obligation to put our criticisms and conclusions in context. Too often, Dittrich fails to do this — and that failure makes it easier for Corkin’s defenders specifically, and the neuroscience community more generally, to shrug off the legitimate and important issues that he does raise, most notably about informed consent.

Take, for instance, the revelation in “Patient H.M.” that I found most shocking: that Corkin and MIT had permitted Molaison to sign his own consent forms beginning in 1980. Federal guidelines involving human subjects require that they fully understand the tests they are undergoing and that they have the right to withdraw from experiments at any time. Given that Molaison could not remember what had happened to him 30 seconds before, it is hard to imagine how he could have met these requirements. More than two months after this information was first brought to light, this has still not been addressed publicly by MIT. When I gave Corkin’s supporters the chance to defend this move, they all demurred.

This seems to be a topic the neuroscience community is hesitant to discuss even outside of Molaison’s case: I made multiple efforts to learn the standard protocol used for amnesiac patients, to no avail. In one instance, an institutional review board specialist at Columbia University’s human research protection office told me several times that a member of the “management team” would respond to me; nobody did. In another, I was told by one of Corkin’s defenders that a specific researcher would be willing to speak with me; he was not.

By the early 1990s, either Corkin or someone else at MIT had misgivings of their own about the arrangement, and by 1992 Molaison had a court-appointed conservator, Thomas Mooney, the son of his former landlady, who had served as Molaison’s de facto caretaker after his parents were incapacitated. Dittrich uncovered several pieces of troubling information regarding Mooney’s appointment, including his almost complete lack of involvement with Molaison outside of rubber-stamping his consent forms and agreeing to donate his brain to MIT and Massachusetts General Hospital — and the fact that, contrary to Corkin’s contemporaneous claims, Molaison had living relatives at the time of Mooney’s appointment.

Corkin’s allies have attempted to address this issue. A statement released by Corkin’s former department head at MIT stressed that he was unaware of “anything untoward” that had occurred regarding Mooney’s appointment, which had followed a “valid legal process.” When I interviewed several people involved in MIT’s investigation, they all said Corkin had had nothing to do with Mooney’s appointment. This would certainly have been the proper call: It would be unseemly for a researcher whose career was so entwined with a single subject to also have a role in deciding who would look after that subject’s interests. But despite her colleagues’ claims, Corkin clearly did have a role in Mooney’s selection. “I looked after his medical care,” she wrote in her own book, “and found a dependable and caring conservator for him.”

That Corkin was allowed to exert so much control over a human being on whom she depended to establish her career and cement her legacy seems indefensible.

The vociferousness of the response to “Patient H.M.” has made one thing clear: Suzanne Corkin was a popular and well-respected scientist. But more than once in conversations over the past six weeks, I’ve found the vitriol directed at Dittrich by the neuroscience community to be curious. While I agree with much of the criticism of his analysis of Corkin’s behavior, his reporting on the treatment of the most famous human test subject of all time raises crucial questions about informed consent that remain relevant today. The scientific community would be wrong to use his book’s shortcomings as an excuse to ignore them.

This article has been updated to clarify a passage examining rules for preservation of research data, and the legal and ethical implications of destroying that data.

Seth Mnookin is the director of MIT’s Graduate Program in Science Writing. His journalism has appeared in numerous publications, including The New Yorker, New York Magazine, Wired, The New York Times, The Washington Post, The Boston Globe, Spin, Slate, and Salon.com.