Last week it came to light that famous author and neurologist Oliver Wolf Sacks had fabricated patient stories in his popular books. Steven Pinker, in a viral series of tweets, called it a bombshell, asking, “Why did The New Yorker, which perpetuates the myth that they employ an army of meticulous fact-checkers, pollute our understanding of mind and brain by publishing these fabrications for decades?”
Sacks was a writer for The New Yorker until his death in 2015. He wrote many books documenting his patients, in 1973 Awakenings and in 1985 The Man Who Mistook His Wife for a Hat, the latter which led to his widespread fame. He also had a successful academic career. According to Wiki, “In July 2007 he joined the faculty of Columbia University Medical Center as a professor of neurology and psychiatry.” It was also an open secret Sacks was possibly a fabulist. British academic and disability rights activist Tom Shakespeare described him as “the man who mistook his patients for a literary career,” although this does not imply that what he wrote was factually wrong.
To answer Pinker’s question as to how no one noticed or why fact-checkers were misled, there are two factors. First, columns and books are not subject to peer review, unlike academic articles. Fact-checking by a magazine or a newspaper does not measure up to the scrutiny of academic peer review. Obviously, fact checkers are not peers, who by definition are published experts of their area of expertise.
Peer review can be thought of as proactive, whereas with books, the ‘review’ comes after the publication. Book publishers can and do publish incorrect or dubious stuff all the time, and usually no one notices or cares. It’s only when enough people complain or the transgression is especially egregious (e.g. Jonah Lehrer, who inserted non-existent Bob Dylan quote in his best-seller Imagine: How Creativity Works,) may a publisher conduct an internal investigation, or a university may investigate if the author is affiliated with one. Peer review is also the slowest thing ever. It’s countless emails with editors and months, if not years, of waiting for referee reports and editorial decisions. It’s understandable why an author would just want to skip all of that and write a book, and also books have much broader appeal.
But this does not excuse deception. Given the extent of his fabrications, on that basis alone Sacks could have been fired, as was the case with Harvard tenured professor Francesca Gino, who was terminated in 2025 for academic misconduct. Books are not exempt. In 2002, Emory University professor Michael A. Bellesiles resigned for fabricating historical data in his 2000 best-seller Arming America, which in 2001 had been awarded the prestigious Bancroft Prize and was presumed correct. An independent investigation later showed he had fabricated quotes and other data, which eventually led to to the rescindment of its award and the author’s resignation. Sacks, similar to Bellesiles, had leveraged his academic credentials to publish stuff that was knowingly wrong, subsequently profiting from sales and a Hollywood adaptation of one of this books.
Second, alluding my post, “Efficient markets and the replication crisis” on how it took 13 years for Mike Israetels PhD ‘thesis’ to finally be reviewed, upon which it was shown to be of no academic value, a similar pattern may have been seen with Dr. Sacks. This explains, to answer Pinker’s question, why The New Yorker didn’t do enough, if any, fact-checking.
I posit that the so-called “replication crisis” exists in part because large swaths of published work are deemed unworthy of close scrutiny. If no one checks the veracity of work, then this is usually an implicit admission the work isn’t important. In the context of peer review, submitted papers deemed to be unimportant are ‘desk rejected’ at the editor’s discretion, meaning the paper is not sent out for refereeing, so as to not waste resources reviewing trivial or wrong work. But this screening process does not exist for books, in which the most important criteria for publication is commercial success, not correctness.
Peer review is time-consuming, even to debunk lies. If the implications of something being correct are inconsequential, then no one may be bothered to review it at all. It’s not just about research confirming peers’ biases, but that the research doesn’t matter to begin with. I think this was the case with Sack’s writings. He told interesting stories, similar to Gladwell, who Steve Sailer noted the similarities between the authors, and that was good enough.
So why was Gladwell debunked, but Sacks wasn’t? Because Gladwell poked the hornet’s nest of IQ. In this case, invoking the ire of Pinker, Sailer and others who rightfully refuted his claims in Outliers about IQ (such as IQ not mattering above 120) as wrong. IQ is really important due to the social implications, such as the role IQ may play in poverty, educational achievement, among other important qualities, and any broad claims will be investigated thoroughly. The left goes over every IQ study with the fine-toothed comb. No data are left unturned, no twin study not called into doubt. The left spent 3 decades trying to debunk The Bell Curve, and failed. Did Sacks’ work receive anywhere as close scrutiny as Murray and Herrnstein? Certainly not.
Same for the purported link between abortion and crime, popularized in the 2005 best-seller Freakonomics, which too was swiftly challenged by academics and lay people alike , such as, again, Steve Sailer, who called the findings into doubt. The abortion-crime link is still in the air, but given how abortion and crime are such hot-button issues, and when combined, it’s little wonder the findings were so thoroughly interrogated. Or back to Bellesiles, guns are a huge issue in America, hence why his book was reviewed for correctness immediately after its publication.
To bring this back, journalists have been playing fast and loose with the truth for as long as the profession has existed. In 2011, shortly after his suicide, Jonathan Franzen disclosed David foster Wallace fabricated some of his non-fiction originally published from ’94-’96 for Harper’s, and compiled in his ’97 book A Supposedly Fun Thing I’ll Never Do Again.
This went unnoticed for so long, some 15-20 years, because there was nothing to be gained from a rigorous fact-checking if the dialogue Wallace had overheard on the cruise ship was actually spoken. Or if an errant baton had actually spun out of control and hit an audience member in the bleachers as he claimed during his visit to the state fair. It’s conceivable it happened, and that level of plausibility is usually the only burden of proof that must be met.
Could Sacks’ patient actually have mistaken his wife for a hat? Given that it’s a well documented phenomenon for people to, say, see the likeness of religious figures in food items (technically called pareidolia), it’s not inconceivable that at least one man in the history of humanity did just that. Hallucinations are real; why can’t it be a hat. Again, it’s not like anyone can check this. If the man were interviewed and recanted the story, then he could then claim the man was ‘cured’.