Like Mother, Like Daughter

What amounts to proof in the world of science? In the context of the UK legal system, it is common to speak of two standards of evidence. The first, used in criminal cases, is where evidence must establish facts "beyond reasonable doubt." The second, weaker standard used in civil matters, requires evidence to make a case on a "balance of probabilities."

Just over five years ago we blogged about a potential problem with the established longevity data for the world's longest lived person, Jeanne Calment. At that time, it seemed likely that Calment's age at death, long recognised as an outlier, would become subject to even greater scrutiny and evidence gathering, ensuring Calment's status was either more firmly established or clearly overturned. For a while, there seemed to be movement - nine months after that blog, the spat had even reached the BBC News in a way that suggested national politics may have become involved.

However, if a week is a long time in politics (and pandemics), it appears that not even five years is long enough to establish clarity in some vexed areas of science. As of this writing, the only change has been the arrival of new storm clouds. A preprint research paper detailing more highly specific evidence in the case against Jeanne Calment became available in August 2023. This preprint argues that Jeanne's daughter assumed her mothers identity around the time her mother died from tuberculosis.

One strand of evidence is based on a reexamination of tapes made of interviews conducted by the validators of Jeanne Calment's longevity. These raw tapes appeared to contain information not obvious from previously published (and thus edited) transcripts. Although those interviews demonstrated Mme Calment had a remarkable memory for past events, when errors were made, these seemed to consistently reflect the experience of Yvonne rather than her mother. Examples included identifying Yvonne's first communion dress as her own, referring to Fernand Calment as her father rather than her husband, and claiming she was taken to school by a maid who worked for the Calment family when Yvonne was a child.

Another strand is based around signatures, which changed markedly at the time of Yvonne's supposed death. Just a snapshot is shown here, taken from a larger set in the paper. The paper suggests the Yvonne began to impersonate her mother in 1933, with visible changes in the signature that are certainly indicative of this. Some samples are shown below, displaying a pronounced change in the final "t" from 1933 onwards.

The paper also explains that Yvonne's husband appears to have sold his own apartment after the death of his wife, and moved in with Mme Calment, something that makes most natural sense where Yvonne is in fact the survivor. Overall, the research is replete with sound reasons to question the existing story, and finishes with a request for DNA testing to conclude the matter once and for all. Whilst exhuming the body of a national hero seems unlikely to happen, there is certainly enough doubt to wonder what science should do in such a case.

The problem of course, is that this isn't simply a personal or even a national matter. Calment's data point now appears throughout global research into human longevity, potentially distorting models and certainly influencing important conclusions based upon that research. What is or isn't true in the case of Jeanne Calment transcends personal or national pride, and speaks to the integrity of the fundamental data underpinning a field of research. Questioning and correcting such data is important. In 2019, Stephen ushered in the new year discussing research by Saul Justin Newman, research that attributed the previously well-established phenomenon of late life mortality deceleration largely to data error. Another provocative paper by the same researcher has languished in preprint, perhaps for casting similar aspersions towards the concept of longevity "blue-zones".

Science cherishes the ideal of objectivity, but is nevertheless an activity carried out by human beings, and so this ideal is more accurately described as an aspiration. However, when important doubts emerge, arising from multiple independent researchers, and are backed by credible evidence, they ought to be dealt with and resolved as completely as possible. In the absence of certainty, perhaps the best approach is for researchers to set the Calment data point aside. It can no longer be considered established beyond reasonable doubt, and may now be failing even on a balance of probabilities...