John Harvard's Journal
Scientific Misconduct, and Its Aftermath
A university investigation found psychology professor Marc Hauser “solely responsible” for eight instances of scientific misconduct, announced Faculty of Arts and Sciences (FAS) dean Michael D. Smith on August 20, 10 days after the Boston Globe broke news of the investigation.
Hauser has taught at Harvard since 1992; he investigates animal cognition as a window into the evolutionary roots of the human mind. “A star in his field and an intellectual celebrity,” as the journal Nature put it, he has written popular works that include Wild Minds: What Animals Really Think (2000) and Moral Minds: How Nature Designed a Universal Sense of Right and Wrong (2006). He has won many teaching awards, including a Harvard College Professorship, and co-taught the popular Core course “Evolution and Human Nature” (nicknamed “Sex”).
Now he is on a yearlong leave of absence; the University has not said whether the leave was previously planned or connected to the misconduct investigation. Hauser has not responded to inquiries from Harvard Magazine. An automatic e-mail reply stated that he is at work on his next book, listed on his curriculum vitae as Evilicious: Explaining Our Evolved Taste for Being Bad; Viking Press has not indicated whether the book is still scheduled. (After news broke of the investigation into Hauser’s lab, the University office of public affairs and communications removed photographs of Hauser from its inventory of available images.)
Hauser was to have taught a psychology course at the Extension School this fall, but on the first day of the session, students who had enrolled were notified by e-mail that the course had been canceled at Hauser’s request, and tuition and fees would be refunded to those who did not wish to take an alternate course. The message contained a statement from Hauser: “Because of the controversy surrounding the investigation, I have decided that the best thing for the students is that I not teach at the Extension School until things conclude with the case. Given my great desire to teach, I look forward to sharing my knowledge of these disciplines in the future.” A course Hauser was scheduled to teach in the spring, “The Moral Sense: From Genes to Law,” has also been canceled. Two graduate students he advised, and two undergraduates whose senior theses he was to advise, have found other mentors. President Drew Faust told the Globe on September 22 that Hauser “may decide he may not wish to come back.” She was apparently not hinting that Hauser had made this decision, but rather was responding to a reporter’s follow-up question after she said there were “uncertainties” surrounding his scheduled return in the fall. (See also Faust’s comments on the Hauser incident and University values, “After the Storm: Presidential Perspectives,” page 55.)
In explaining the misconduct finding, Dean Smith cited three papers with which a faculty investigating committee found “problems.” One, published in the Proceedings of the Royal Society B in 2007, involved rhesus monkeys’ ability to recognize human gestures; a correction published in July states that Hauser and a collaborator returned to the original site to re-run the experiment after it was discovered that video records and field notes for their original work were incomplete. The new experiment reproduced the original findings, the journal reported. A second paper, published in Science in 2007, examined the ability of rhesus monkeys, cotton-top tamarins, and chimpanzees to make inferences about humans’ mental states based on observed actions. Field notes and other information about the rhesus monkeys were missing for this paper also, so the researchers returned to the site to replicate that experiment, too. A Science spokeswoman said peer reviewers were still analyzing the new material.
The third paper, published in 2002 in the journal Cognition, purported to find that cotton-top tamarin monkeys could recognize auditory patterns, or “grammars.” The experimental setup described in the paper exposed one group of monkeys to two different grammars, and another group to two sequences from the same grammar. The paper reported that the monkeys who heard a second sequence from a different grammar paid more attention, indicating that they noticed the difference. Heightened attention to novel stimuli, a facility much studied in human infants, is believed to underpin the capacity for speech.
The Cognition paper was retracted in August, after Hauser notified the journal that the data did not support the reported findings. Editor Gerry Altmann said he subsequently received more details in a letter from Smith: an examination of the experiment’s videotapes found that all the monkeys had been exposed to two different grammars, making it impossible to compare reaction to a familiar versus a novel grammar. Altmann said the letter did not specify how this happened. “It is conceivable that there was in fact no intent to deceive or fabricate, if we assume a whole chain of procedural errors,” he wrote on his blog--if, for example, one group heard the wrong grammar due to computer error or human error, and none of the researchers who scored the monkeys’ responses or checked the results noticed the discrepancy. But Altmann, a psychology professor himself, believes this explanation is improbable: “I am forced to conclude that there was most likely an intention here, using data that appear to have been fabricated, to deceive the field into believing something for which there was in fact no evidence at all.”
Smith’s public statement did not give details on the other five counts of misconduct, but said they related to studies that “either did not result in publications or where the problems were corrected prior to publication.” Neither did Hauser in his statement to the Boston Globe the same day, though he said he was “deeply sorry” and had made “significant mistakes.” “I have learned a great deal from this process and have made many changes in my own approach to research and in my lab’s research practices,” he added. “Research and teaching are my passion. After taking some time off, I look forward to getting back to my work, mindful of what I have learned in this case.”
Harvard’s investigation evidently began in 2007; the Chronicle of Higher Education obtained a statement from a former research assistant in Hauser’s lab who said he and other researchers had approached University authorities with their concerns about practices there. According to the Chronicle, the trouble started with an experiment involving rhesus monkeys’ ability to recognize auditory patterns. Hauser and one research assistant watched the monkeys’ responses to the sounds and “coded” the results, marking whether they seemed to respond to novelty (similar to the study with the cotton-top tamarins). A second research assistant, who was asked to analyze the data, noticed a major discrepancy: the first research assistant’s coding found the monkeys didn’t notice the patterns; Hauser’s coding of the same videotapes indicated that they did.
After Hauser resisted a request to have a third investigator review the data independently, one of the research assistants and a graduate student reviewed the tapes themselves without Hauser’s permission, the Chronicle said.
They each coded the results independently. Their findings concurred with the conclusion that the experiment had failed: The monkeys didn’t appear to react to the change in patterns. They then reviewed Mr. Hauser’s coding and…discovered that what he had written down bore little relation to what they had actually observed on the videotapes. He would, for instance, mark that a monkey had turned its head when the monkey didn’t so much as flinch. It wasn’t simply a case of differing interpretations, they believed: His data were just completely wrong.
Harvard did not announce anything about the affair until the Globe reported the Cognition retraction on August 10. By that point, Nature subsequently noted, gossip about an investigation of Hauser’s lab “had become standard cocktail-hour fare” at conferences of scientists. Other publications reported that, years earlier, SUNY-Albany psychology professor Gordon Gallup Jr. had publicly questioned a Hauser finding that monkeys could recognize their own reflections in a mirror; he reviewed the videotapes from Hauser’s experiment and said he saw no evidence of recognition.
The University still has not said exactly what discipline, if any, has been handed down. According to Smith’s statement, the options for cases of misconduct include involuntary leave; the imposition of additional oversight on the laboratory of the scholar in question; and restrictions on his or her ability to apply for grants and supervise graduate and undergraduate students.
In an interview with Harvard Magazine, President Drew Faust said neither the length of the investigation, nor Smith’s reticence, should be interpreted as an attempt to sweep the incident under the rug. She noted the complexity of the work under review and the need to keep certain details confidential until federal agencies have concluded their own investigations. (Smith said the University was cooperating with inquiries from agencies including the Public Health Service Office of Research Integrity, the National Science Foundation Office of Inspector General, and the U.S. Attorney’s Office for the District of Massachusetts. Spokespeople for all three agencies said they could not confirm their involvement; generally, sanctions imposed by the three agencies range from a requirement for additional oversight to return of grant money, a prohibition against applying for government grants for a set period of time, and criminal indictment.)
Facing so much uncertainty, Hauser’s colleagues in evolutionary biology are left to wonder how much of his work is reliable. He has been a prolific researcher, coauthoring more than 200 papers during a 25-year career. Emory University primatologist Frans de Waal, quoted in Nature, called the affair “disastrous” for the field: “If one prominent person is under suspicion, then everyone comes a little bit under suspicion.” The danger, Hauser’s colleagues have said, is that the wider academic community and the public may start to believe that animal behavior can’t be objectively evaluated. There have already been suggestions that “this kind of work is imprecise and it’s really anybody’s call what the animal is doing and it’s not rigorous,” UCLA primatologist Joan Silk told the Globe. “That’s a big misperception.”