Consensus: Could Two Hundred Scientists Be Wrong?
September 2, 2016
The recent publication of a book about neuroscience’s most famous amnesia patient—known for decades only as H.M.—has stirred up a controversy in the world of science. On August 3, 2016, the New York Times Magazine released an article adapted from Luke Dittrich’s book, Patient H.M.: A Story of Memory, Madness, and Family Secrets. Two days later, on August 5, over two hundred neuroscientists from around the world had signed a letter to the Times in support of Professor Suzanne Corkin, the Massachusetts Institute of Technology scientist who did most of the research with H.M.
Henry Molaison (H.M.) suffered profound memory loss as a result of an experimental brain operation conducted in 1953 in an effort to control his epilepsy. The surgery removed most of Molaison’s hippocampus and some nearby structures on both sides of his brain, leaving him incapable of creating new episodic memories. Henry, who was the inspiration for the popular film Memento, could recall many things that happened to him prior to 1953, but after the surgery he couldn’t tell you what he had done five minutes before the present moment. As Dittrich put it in the New York Times Magazine article, “Each of the hundreds of times [he and Professor Corkin] met, it was, for Henry, a first meeting….”
The Questions Raised
Sadly, Suzanne Corkin died in May of 2016, months before Dittrich’s book and the Times article appeared, and so it was left to her colleagues to defend her against what they believed was Dittrich’s “biased and misleading” description of Corkin’s work with H.M. On August 9, six days after the Times article was released and four days after the letter signed by two hundred scientists appeared, Professor Corkin’s colleagues in the Department of Brain and Cognitive Sciences got specific about their concerns, posting a letter on the MIT website outlining what Department Head James J. DiCarlo, characterized as the “allegations” in Dittrich’s article:
- Allegation that (H.M.’s) research records were or would be destroyed or shredded.
Given the importance of H.M. to the history of neuroscience, any destruction of primary data would be considered an enormous loss and possibly a violation of research ethics.
- Allegation that the finding of an additional lesion in (H.M.’s) left orbitofrontal cortex was suppressed.
When Molaison died, his brain was donated to science, and postmortem analysis revealed a previously unknown lesion in his left frontal lobe. The discovery of this injury could affect the interpretation of earlier studies of H.M.’s memory loss. DiCarlo objected to the suggestion that Corkin had ever tried to block the publication of an article revealing this brain injury. The article did eventually appear with Corkin as a coauthor.
- Allegation that there was something inappropriate in the selection of Tom Mooney as Henry’s guardian.
Perhaps the most troubling questions raised by Dittrich were about the way Corkin obtained Henry’s consent to be studied. Given Molaison’s disability, it was doubtful that he was capable of understanding and evaluating Corkin’s requests to participate in research, and yet informed consent is a basic hallmark of research practice. According to Dittrich, from 1981 to 1992, Henry was the only person signing the informed consent forms for Corkin’s research studies. Eventually, although three of Henry’s first cousins were alive at the time, Corkin arranged for Thomas F. Mooney, a second cousin, to become his conservator. Mooney went on to sign research consent forms and—importantly—an agreement to donate Henry’s brain to MIT upon his death.
Video clip taken from the internet broadcast of Henry Mollison’s brain being sliced in a laboratory at the University of California at San Diego on December 4, 2009. (No sound.) As an indication of the historic nature of this event, the entire fifty hours of this process was broadcast live.
The August 5 letter to the New York Times signed by two hundred scientists in defense of Professor Corkin appears on the MIT website with the heading “by International Community of Scientists”—a clear attempt to suggest a consensus opinion in support of Professor Corkin. Without question, Suzanne Corkin was much admired, and her death must have been a great loss to her colleagues. On a human level, the urge to defend her honor is completely understandable, but was it a good idea? Should these scientists have put their names and institutional affiliations on this letter?
When Scientists Appeal to Consensus
Two other recent cases of scientists making appeals to consensus come to mind. In my July 2015 “Behavior & Belief” column, I argued that Lumosity and similar brain training programs were scams. I based my assessment, in part, on a 2014 consensus statement signed by seventy neuroscientists and memory researchers that found the claims of the brain training industry to be unjustified.
Similarly, in arguments about climate change, scientists frequently cite the overwhelming consensus among climate change scientists that global warming is real and caused by human activity. For example, in a recent climate change debate (see video below) physicist Brian Cox made an appeal to scientific consensus, and the Australian climate change denier Malcolm Roberts, a member of the far-right One Nation Party, attacked Cox for citing consensus rather than data. Fortunately, Cox brought a temperature trend graph and other evidence in further support of his viewpoint, whereas Roberts could only defend his position with vague conspiracy theories about NASA and other scientific groups “manipulating” the data.
Physicist Brian Cox debates climate change denier Malcolm Roberts.
As Malcolm Roberts was trying to suggest, consensus alone is not evidence. There was a time when, according to the prevailing scientific consensus, the Earth—not the sun—was at the center of our planetary system, and it took a long time for geocentricism to be tossed aside. Appeals to consensus and authority should always be suspect. At the same time, we cannot all be experts. In the contemporary world, citizens must confront many important issues without having the necessary skills to judge the evidence. As a result, we are often forced to rely on authorities. Furthermore, if the experts have reached a level of agreement, knowing that they agree can sometimes be a useful bit of information.
Consensus is not a given. Often well-intentioned investigators working with shared sets of data fail to agree. Problems remain unsolved, and scientists retreat to their corners in support of conflicting pet theories. This is common and to be expected. Eventually, future advances in technology or theory may make additional progress possible, but until then disagreements and incomplete answers are the normal state of affairs. So, when scientists are able to converge on a shared understanding, consensus can add a little weight to that view.
But how should regular folks know when—and when not—to accept a consensus viewpoint? To help decide, it is useful to ask, “What is the basis of the claim?” In the strongest cases, we have the word of scientists who have worked directly with the relevant evidence. For example, the consensus in climate change comes from researchers who work directly with climate data. Similarly, the brain training consensus statement was written by seventy scientists who routinely conduct and evaluate research on memory and learning. So, in these cases, the presence of a consensus seems noteworthy. It is not a guarantee that the dominant paradigm is correct, but it adds weight to the claim.
Unfortunately, the apparent consensus produced by the two hundred signatures on the MIT website in support of Professor Corkin is something else entirely. It is unlikely the signers had direct experience with the relevant evidence.
The Questions Raised
In the August 9 letter, Professor DiCarlo highlighted and responded to the three “allegations” above. According to DiCarlo, two of Professor Corkin’s department colleagues at MIT had investigated the matter and their report “rebutted each claim” made by Dittrich in the Times Magazine article. Unfortunately, that appears to be far from clear.
In response to the August 9 letter, Dittrich published a piece on Medium addressing each of the objections raised by the MIT evaluation. In defense of his claim that Corkin destroyed some of H.M.’s data, Dittrich posted an audio file of his interview with Corkin in which she can be heard saying that she had already shredded data and planned to shred more. This means that either Corkin did what she said and destroyed data, or she lied (or perhaps misspoke) to Dittrich.
In a subsequent piece posted on the MIT website (that’s number three, if you are counting) on August 20, Professor DiCarlo further defended Corkin, arguing: (1) that an internal investigation had discovered that no files were destroyed, (2) that there was nothing wrong with the assignment of Thomas Mooney as Henry’s conservator, and (3) that any dispute over publication and the newly discovered lesion was overblown.
Weighing the Evidence
This is a spat that will not be resolved. Professor Corkin is no longer with us, and she alone could answer some of these questions. In her defense, it should be acknowledged that the ethical standards for research with human participants has been evolving somewhat gradually since World War II, and Corkin began working with H.M. over four decades ago. Today’s standards for obtaining informed consent and for the preservation of research data are much more rigorous than they were when she first met Henry. In addition, squabbles among authors of scientific publications happen from time to time. As a result, it is unclear—to me, at least—whether she violated professional research standards.
Having said that, the questions raised by Dittrich are ones that, as a journalist, he has a right—perhaps even a duty—to raise. Our ethical standards will not continue to evolve unless we soberly consider the various research dilemmas of the past, present, and future. Perhaps there were good reasons for assigning Thomas Mooney conservator for H.M.—despite his not being Molaison’s closest relative—but what principles should guide cases like this in the future? Like Dittrich, I would prefer that every bit of Henry’s data be preserved, and yet in Dittrich’s audio recording, Corkin can be heard defending the shredding of data. It is clear to me that Dittrich considers Henry Molaison’s legacy to be a public good, something that should be preserved for future generations, and I tend to agree.
But now back to the question of scientific consensus.
When Should a Scientist Sign a Statement, Petition, or Letter of Support?
It is understandable that MIT’s Department of Brain and Cognitive Sciences should want to defend its status in the scientific community and come to the defense of a beloved colleague. But what of the two hundred signers of the letter to the New York Times? The “International Community of Scientists”? Each of these individuals lent their professional identities to a claim of journalistic bias when few—if any—of them could have known the facts. In an interview for this article, Luke Dittrich reported that only one of the signers had an advanced copy of his book. In preparation for the book, Dittrich interviewed four people who later signed the statement. He recalled discussing the general question of informed consent with all four, but “I don’t think I delved into the three specific ‘allegations’ that MIT made.”
So here is what we know:
- All of the signers responded very quickly—in all but one case based on the New York Times Magazine article alone.
- The signatures were obtained before the specific claims of bias were made public by MIT.
- Once MIT identified the “allegations” it became clear that few if any of the signers could have had relevant knowledge of the facts. For example, it took an additional week and a half for MIT to investigate the H.M. files and come to the conclusion that they had been “maintained and not destroyed”—a claim that would be difficult to validate.
Scientists have a unique and important role in the public dialogue. They are trusted—or should be—to have special skills of analysis achieved after long study and practice. There are many important social issues to which scientists can and should contribute. Unfortunately, when it comes to evolution, climate change, and the benefits of vaccination, scientists are too often ignored.
But if scientists are going to maintain their credibility, they should not squander their authority by weighing in on subjects outside their circle of knowledge. Many of us—myself included—have at times made public statements that go beyond the data, but if we want to maintain the influence of science in society, we need to make every effort to stick to the subjects we know best.
After spending a week thinking about this case, I cannot help but wonder what might have happened if MIT had just let the issue slide. Dittrich’s book is interesting and provocative, but the “allegations” that MIT saw in the New York Times article are undoubtedly a much bigger deal inside the field of neuroscience than outside. Had Professor Corkin’s supporters kept their powder dry, they might have done a better job of protecting her honor. The Dittrich article and book would have caused whispers within the scientific community, but by returning fire MIT ensured that the case would be picked up by the press and that more copies of Dittrich’s book would be sold. MIT may ultimately have come to understand this point. Neither Dr. DiCarlo, the head of the Department of Brain and Cognitive Sciences, nor the department media office returned emails requesting comment for this article.
Not responding to the New York Times would also have saved over two hundred scientists from having to endorse a dubious statement.
Finally, I strongly recommend Dittrich’s book, Patient H.M.: A Story of Memory, Madness, and Family Secrets. It paints a vivid picture of the history of psychosurgery and the treatment of the mentally ill from the mid-nineteenth to the late twentieth centuries, and it provides an evocative and detailed account of the world’s most famous human research participant. In addition, Dittrich has a number of personal connections to the story. For example, it was Dittrich’s grandfather, William Beecher Scoville, who performed the tragic operation in 1953 that turned Henry Molaison into Patient H.M.