Validation of Exceptional Longevity
|
Validation of Exceptional Longevity Validation of Extreme Longevity Cases in the Past: The French-Canadian Experienceby B. Desjardins |
[ References ]
Age validation of alleged cases of extreme longevity is rendered necessary by a well known fact: there is a general tendency for the ages of the very old to be exaggerated, and for those with a reported high age at death, data is often not reliable.
This is particularly true of the past, for a variety of reasons. Registration of the basic demographic facts did not become systematic and official until the XIXth or XXth century in most countries. People were mostly illiterate, and probably did not relate to a date of birth or an age as people today are expected to. Thus, in the absence of the exact information when a death was registered, it was normal that the age of the deceased be impressionistic, a translation in the case of the aged of someone's perception of "very old" (or "very very old"!). Exactitude in this matter was probably not seen as important, especially when the document from which we gained information about age at death was not written up specifically for this purpose.
The corroboration of the reported age at death in past populations can be done in a variety of ways, depending on the sources available, themselves derived from each administrative, cultural or religious setting. The systematic registration of baptisms, marriages and burials prescribed by the roman-catholic Church since the XVIth and XVIIth centuries offers one of the best sources in this regard. Indeed, not only do the resulting parish registers allow the exact measure of age at death by the matching of people's baptism and burial certificates, but they include the possibility of confirming the identification of the individuals and of verifying the plausibility of measured ages through family reconstitution.
In Canada, parish registers appeared right at the outset of the French colonization early in the XVIIth century and have been well kept and preserved from that time. Their society being homogeneously catholic, French-Canadians have thus been systematically covered for several centuries, allowing them to be used for studies of longevity. We will here present some work that has been done on French-Canadian centenarians or quasi-centenarians; we will then reflect, on the basis of these experiences, on the theory and practice of age validation for the past.
"The question of the human longevity of our times is one of the greatest importance in biological investigations, and the furnishing of a considerable number of well-established facts, and contributing to clear it of the confusion produced by the gratuitous assertions by which it is surrounded, appears to be a service to science".
This declaration appears in the notes presenting a table which summarizes an investigation on centenarianism published along with the results of the 1871 census of Canada (Government of Canada 1878). It is thus with the noblest of intentions that Joseph-Charles Taché, high-ranking civil servant in charge of the census, had decided to carry out an investigation of longevity in Canada which he deemed "a necessary complement to the whole of the second part [relating to population] of the Canadian statistics".
The investigation was carried out as follows. First names were collected of persons reported as having died after reaching the age of a hundred. For this purpose, the schedules of censuses, parts of the parish registers, books, newspapers and magazines were consulted, yielding a total of 421 reputed centenarians belonging to the different provinces and nationalities of Canada, including aborigines. Being aware that reported ages could not be trusted, Taché admitted in a definitive list from these 421 "only individuals whose ages could be proved by authentic documents, examined with a rigorous scrutiny". This had the effect of eliminating outright the majority of names, belonging to people born out of the country or not subject to systematic registration of demographic facts, and conversely of keeping virtually French-Canadians only, whose ages could be ascertained through parish registers. The list had then dwindled to 82 references.
He then proceeded to investigate each of these 82 cases, with the collaboration of the Quebec clergy in general and of Cyprien Tanguay, himself a clergyman, renowned genealogist and author of a comprehensive genealogical dictionary of Quebec families of the XVIIth and XVIIIth centuries (Tanguay 1871-1890). Their efforts yielded a list of nine "true centenarians", among whom appeared portentously Pierre Joubert, alleged to have died at 113 years of age in 1814.
This identification of Pierre Joubert as a super-centenarian was readily accepted and has known a great career: Jeune (1995) reported that his case was the only one believed in reviews by Young in 1905 and Bowermann in 1939. The Guinness book of records once listed him as the oldest person to have lived, and his name is still cited to this day as one of the proven long-livers of mankind (Guinness 1997). This credence probably originated in the official nature of the document in which the case was published, the whole study having never been contested until recently.
Indeed, over a century elapsed before Taché's list of centenarians was systematically verified, by demographer Hubert Charbonneau (1990, 1991). For obvious reasons, he first investigated Pierre Joubert's claim to fame, which proved remarkably easy to verify. As Joubert had married, Charbonneau decided it was logical to look for his wife's burial, something apparently nobody had thought of doing. The certificate, appearing in the 1786 parish register of St-Denis-sur-Richelieu, a village south-east of Montreal where the couple was residing, specified that the deceased was a widow, negating outright the survival to 1814 of her husband. Charbonneau then went back in time in the same register and completed his demonstration by finding the true burial of Pierre Joubert in 1766, "âgé de 67 ans" (he had actually just turned 65). The person deceased in 1814 turned out to be the son of Pierre Joubert, born in 1732.
Charbonneau then proceeded to examine the other eight cases confirmed by the Canadian statisticians, listed here with years of birth and death and age at death:
In the first case, the XIXth century investigators linked a death certificate of 1847 with a baptism in 1738. This baptism actually belonged to a cousin of the same name; however, the correct Rosalie Lizot was born in 1740 and would still have died at age 107. The death certificate is very precise, naming her deceased husband, and nothing at this time has been found to disprove the link between the 1740 birth and the 1847 death. The Jean-Baptiste Poupard born in 1689 actually died in 1730; as the 1793 death certificate gives no other information than name and age, it could be ascribed to many others, as could be François Forcier's for the same reasons. François Giguère-Despins and Pierre-Noël Plante proved to be cases of confusion between namesakes, and Marie-Josette Dupuis's identification was very uncertain. Thérère-Marie Skanaouati was an Indian living on a reserve near Montreal; missionaries linked her death in 1874 to a baptism in 1774 on the basis of very vague information and on the account of an eighty year old kin, far from convincing evidence indeed. Finally, Anne-Charlotte Dumont's case proved to be without ambiguity: a spinster, her death certificate identifies her very precisely with her parents. She had no namesakes, and was known to have lived with her sister in Trois-Rivières where the death is registered; furthermore, her nephew is cited at the burial. She had been a centenarian for 27 days.
So out of the nine "official" centenarians of Taché's investigation, seven were shown by Charbonneau to be false or were discarded as insufficiently ascertained, with another, the highly improbable case of a woman who would have died aged 107 years in 1847, being left open in the absence of information allowing to eliminate it. Taché's 421 references thus yielded one confirmed centenarian, a noblewoman who died in 1842.
Reflecting back, one wonders why did the list of confirmed centenarians, published under the patronage of the Canadian government, end up to be deflated to such a degree, and why did the myth of super-centenarian Pierre Joubert endure such...longevity? The answer certainly lies within the context of the study. Joseph-Charles Taché was a doctor, an author, a journalist and a politician, before becoming a senior official in the Ministry of Agriculture and Statistics. He was definitely a man of great culture and knowledge, and a true scientist. His credibility was enhanced by his notes demonstrating an awareness of the nature of the problems facing his endeavour and listing age validation criteria which would still stand the test today: "Only the names of those were inserted whose identity could be established by the record of birth and the record of death...wanting information being supplied by other records or by ascertained collateral facts, presenting direct and conclusive proof of the identity of the person and of the necessary dates...Whatever did not stand the test of scrutiny and produce an irresistible conviction was set aside...For each case, it is believed that it can be said, with the assurance which the subject allows: there is only one individual who could have been the person designated, and that individual was born at such a date, in such a place, of such father and mother, and died at such a date and at such a place...".
The problem is that the authors of the study did not respect these principles. The identification of the Pierre Joubert who died in 1814 with the one born in 1701 relied on the rarity of the name and the wording of the document, where the deceased was identified in relation to a son instead of his spouse or parents as is usually the case. A link was accepted which produced a 113 year old, disregarding the fact that the death certificate specified he was only (!) 105 years old and that the burial of Joubert's homonymous son had not been found (since it had been attributed to his father!). Likewise, other centenarians were authenticated without sufficient rigour. To their defense, people of that time had not the knowledge we have today of the odds against the ages they were validating. Furthermore, they were certainly influenced by the often repeated idea, by no less than Jules Verne for example (Charbonneau 1991, p. 219), that French-Canadians, "issus d'une race forte", were blessed with exceptional longevity. Taché himself expressed in his notes that "This investigation has the greater value in the important question of longevity, that it refers to a population which, more than any other, perhaps, presents frequent examples of long life". Finally, the necessary verifications that should have been carried out were extremely difficult, given the fact that the registers were dispersed all over a vast territory.
As to why the myth persisted, one may mention the apparent strength of the dossier and the perceived difficulty of disproving it, and the absence of justification for doubting it sprinkled with a touch of national pride in those who were best in the position to mount the challenge. Charbonneau himself was prompted to revise Taché's investigation by the fact he had the necessary information close at hand, thanks to a major project he launched with his colleague Jacques Légaré under the name Programme de recherche en démographie historique (PRDH) at the Université de Montréal: the reconstitution of the Quebec population of the XVIIth and XVIIIth centuries (Légaré 1988).
Many documents, such as marriage contracts and notarial deeds, exist to alleviate any gaps in the excellent series of parish registers in Quebec. The population, after an initial influx of immigrants, grew essentially through natural increase, numbering but 70 000 at the end of the French regime, a century and a half after its foundation. Out-migration, although significant at some periods, was negligible in total. It is the existence of this complete population of manageable size covered by exceptionally good sources which allowed the project to be carried out, through the setting up of a computerized population register containing the biographical and genealogical information on all individuals who lived in the colony. At this time, the "Registre de la population du Québec ancien" (RPQA), name given to the data bank set up by the PRDH, includes all 700 000 births, marriages and burials registered in Quebec before 1800. The 300 000 registered before 1766 are available for analysis in the form of reconstituted families; also linked into the data set are marriages and burials of married persons up to 1800. The PRDH also possesses a complete set of microfilms of Quebec parish registers up to 1876 (Desjardins 1998).
It is this wealth of information which allowed Charbonneau to verify the investigation done in the 1870s. But this work was more age "un-validation" than age validation. Given the information now available, it appeared logical to PRDH members to launch their own investigation of centenarianism (Charbonneau and Desjardins 1990).
Contrary to the predecessors of the XIXth century, it was possible to collect all reported cases of centenarians in the death certificates for the period, given they were all computerized. Among the 210 000 burials registered in Quebec before 1800, 178 mentioned an age at death of 100 or more. The group was made up of about an equal number of immigrants and of Canadian-born persons of European origin, and of a minority of Indians; 58% were men (Table 1). Nearly three out of five of them would have died exactly at age 100 (Table 2). The attraction of round numbers is evident, and boldness was in order, as a tenth of declarations reached or surpassed 110 years, and two made it to 120 years.
With the information from the RPQA, verification of these reported ages was possible for 75 of the 76 Canadian-born and for 50 of the 83 immigrants. In the case of the latter, it must be said that verification relied generally on other declarations of age, at marriage or in a census; this entailed a certain degree of approximation, but no doubt the margin of error for false centenarians was greater than the one linked to age declarations of young adults.
After verification, only one centenarian remained, a Jesuit named Jean Feuville, reported to have been born April 16, 1600 in the south-west of France, who died in Québec City December 8, 1700 at the age of 100 years, 7 months and 22 days. Although the information on Feuville has a precise character, with exact dates of birth and of entering into religion (October 18, 1617), it is second hand. Father Feuville would have arrived at age 49 in Canada, which is the highest age, but by just a little, of all Jesuits who came. Furthermore, the place of birth the same source gives him in Aquitaine is suspicious. It is thus impossible to accept Father Feuville as a confirmed centenarian.
A dozen false centenarians would, however, have reached 95 years of age before the end of the XVIIIth century. The eldest, Élisabeth Chavigny, born in Canada, died in Québec City as early as 1748, aged 99. At the other end, 14 individuals had not reached 80 years of age, and five not even 75; the youngest, an immigrant, might have been only 72 years old. On the average, the alleged centenarians were 88 years old.
One must not be surprised by these results. The list that was verified was drawn up on the basis of ages appearing in the death certificates; as explained in the introduction to this paper, such ages are notoriously imprecise in parish registers of the past, specially in the case of very old people. In most cases, the priest had no way of verifying the age of the deceased, and thus relied on impressions or hearsay, which most often yielded overstatements of age.
The PRDH population register that was used to verify the list is established by using the family reconstitution method (Fleury and Henry 1956). Biographies are drawn up by linking births, marriages and burials on the basis of names, with coherence verified at both the individual and family levels. The method is a classic of historical demography, and has yielded myriads of statistically sound results. But statistical soundness does not preclude some errors at the level of individual cases, which may
|
Table 1. Reported centenarians before 1800 according to sex and origin | ||||||||||||||||||||
* Only Indians settled in missions and converted to catholicism, a small minority, could appear in parish registers | ||||||||||||||||||||
|
Table 2. Reported centenarians before 1800 by sex and age at death | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
happen as some links accepted in family reconstitution are not as sure as others. The most frequent cause of error here is confusion between siblings, because first names were often repeated or replaced within a family, when children died at an early age. The acceptance that Élisabeth Chavigny lived to age 99 in the first half of the XVIIIth century would tend to prove that reaching the age of 100 was possible so long ago, where numbers would be greater, and is thus by no means a trivial fact; so her case was investigated thoroughly to establish with absolute certainty its authenticity.
Élisabeth Chavigny's burial certificate appears in the register of the Hôpital-Général de Québec in the correct chronological order, dated January 24, 1748. It duly records that the body of "Marie Élisabeth Chavigny veuve Landron âgée de cent deux ans commencé" (Landron widow aged 102 years started, which means 101 years old) was buried the day before in the pauper's cemetery.
Élisabeth Chavigny had married Étienne Landron in Québec city the 10th of October 1667, 83 years and three months earlier. The marriage appears in the parish register of Notre-Dame-de-Québec at that date; the spouses had signed a marriage contract in front of notary Gilles Rageot eight days before. From the time of her marriage, Élisabeth Chavigny always resided in the city of Québec, where baptism, marriage and burial registrations show no gap. Her husband was a successful merchant and the couple belonged to the high-bourgeoisie of Nouvelle-France. This has the advantage of giving the couple a lot of visibility in the parish registers as witnesses at religious events and in the notarial documents as active participants in the commercial life of the colony.
Élisabeth Chavigny gave birth to 16 children. Several died at a young age, but one of her daughters reached 88 years. Four of her sons seem to have migrated before adulthood, probably to France to better their education or to take up a military career, common choices for sons of the colonial elite. Her husband, two sons and a daughter died during a terrible smallpox epidemic that hit the St-Lawrence valley in the winter of 1702-1703 (Desjardins 1996). Widowed after 35 years of marriage, she never remarried.
Family sheet of Élisabeth Chavigny's family of origin
But she continued to appear regularly in the documents. She made a donation to her daughters in 1726; in 1730, she notarized a will which she replaced with another seven years later. Finally, in 1738, she distributed her belongings among her children. This would have coincided with her leaving home to go live at the Hôpital-Général, the nursing-home of the time, where she ended her days.
Élisabeth Chavigny, wife of Étienne Landron is thus monitored for 71 years from her marriage, in documents where she is clearly identified through her dead husband or her children. She had a very discriminating name; there was only one other contemporary Élisabeth Chavigny, her niece, born in 1700. There is no possibility of confusion between the two, because the niece died in 1731, with the death certificate naming her husband who remarried six months later, naming in turn his deceased wife. The registers of Québec have no certificate that could be hers between 1738, last date where we know she was still alive and 1748, year of her death. There is no reasonable doubt that the woman who married Étienne Landron in 1667 is really the one who died in 1748. However, her age at the moment of death remains to be verified, which requires the validation of the link made with her baptism.
Élisabeth Chavigny was the daughter of François Chavigny and Éléonore Grandmaison. Their's was the only Chavigny family in the colony; the Landron-Chavigny marriage is explicit and the filiation is confirmed in the marriage contract, where her mother was present with her fourth husband (Chavigny was her second). He was a nobleman with a high post in the Administration; Éléonore Grandmaison was an important Lady of the colonial society and a lot has been written about her.
Married in France, the Chavigny-Grandmaison couple arrived in Canada in 1641, when the colony still numbered but a few families. Having settled a few kilometres outside the walls of Québec, it is mentioned in a journal kept by the Jesuits in 1645 and 1647 that a priest had to go out to Chavigny's house to baptize a daughter. This is what also happened for our Élisabeth, baptized January 31, 1649 at the age of approximately a month and a half, according to the certificate. There is thus a slight imprecision in her date of birth, which can be placed in December 1648, but without the exact day.
Élisabeth had four older sisters which all survived to adulthood (see family sheet); their lives are well documented in the RPQA but anyway, for our purpose here, what matters is to avoid mistaking Élisabeth with a younger sibling. Her parents had only one more child, a male, since we know her father became ill and left the colony in 1651, leaving wife and children behind, to seek medical treatment in the home country. He died at sea, and his widow, Éléonore Grandmaison, remarried in August of 1652.
This sequence is unequivocal: Élisabeth was born in 1648; her mother had a son in 1650, and the interval between the two excludes any birth in between; the father left
in 1651 and died; his widow remarried in 1652. Nowhere is there room for a confusion between siblings which would yield a younger age for Élisabeth Chavigny. She gave birth at regular intervals until she was 44, and one last time when 47, typically after a longer interval. This age at last birth is above average, but well within the acceptable limits. Longevity is often associated with slow aging (late puberty and late menopause); it is easy to accept someone who almost reached the age of 100 remained fertile for a long time. Thus it seems certain a woman born in 1648 celebrated her 99th birthday. And if this can be seen as corroborating evidence, one might note that the other four children of François Chavigny and Éléonore Grandmaison, whose death dates are known (a daughter left the colony and died in France; she was 62 when last heard of) died at the average age of 73, a good 10 years over the normal age of the time.
Although not in such detail as the one above, all ages at death of 90 and over in the RPQA were validated (Charbonneau 1995). One hundred and five nonagenarians were then confirmed from the some 20,000 births on the shores of the St-Lawrence before 1700, or one out of every two hundred persons born. They were almost evenly split between the sexes, but no male lived beyond the age of 95 (Table 3).
|
Table 3. Confirmed nonagenarians among the 20,000 born in Canada before 1700 by sex and age at death | ||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||
Validation of age at death is a priori very simple, relying on unequivocal identification of a person and an age. In the ideal case where age is calculated from linking a birth registration with a death registration, this implies making sure the two certificates relate to the same person, on the basis of the information usually available in such documents: names, names of kin, especially parents and spouse, age, occupation, and places of birth and of residence. Corroboration is always appreciated, either through age declarations, censuses, etc.
How precisely these principles are actualized depends on the context of the case being investigated. Vital events might not be officially registered but some form of registration, often religious, can be just as reliable. In this regard, historical age validation is no different from contemporary age validation: no universal set of rules can be established because no universal sources of information exist.
As we stated before, Roman-Catholic parish registers are usually an excellent source for age validation. As they were used to detect marriage impediments because of kinship or affinity, priests were instructed to identify clearly the persons both through their names and those of their kin. When kept as prescribed, parish registers contain all the material needed for age validation. Michel Fleury and Louis Henry (1956) put in writing the rules to be followed to link up the different certificates to obtain individual and family histories. The resulting "reconstituted families" offer excellent corroboration through the internal coherence of each member's life history and the protection they offer from mistaken identities, between kin in particular, when someone is linked to the baptism of an older brother or sister, who died very young, with the same first name. One then usually proceeds in two steps, as we have shown with the case of Élisabeth Chavigny: first making sure the death certificate is attributed to the right adult, then linking with certainty this adult to his baptism.
But all is not perfect. Migration is the greatest enemy of family reconstitution on the basis of parish registers because, except for the stable, some of the vital events occur outside the range of registration; thus, the study of mortality has never been the strongpoint of family reconstitution studies. French-Canadian data alleviate the problem as the whole area is included, out-migration having been negligible before the XIXth century. Simply by following basic rules of identification and coherence, the PRDH reconstitution offers a quality of data unmatched anywhere else at such a scale.
The statistician is thus perfectly at ease. But one must distinguish the individual case from the whole set of cases. Mistakes were made, often when the priest himself tried to establish a link between different events in time. Mistakes in identification happened, through name-saking, imprecision in the documents, etc. One must then take into account the degree of certainty needed, given one's purpose. As Élisabeth Chavigny was alleged to be the oldest person to have lived in Quebec before 1800, her case was submitted to a severe exercise in validation. The other hundred nonagenarians were individually validated, but to a lesser degree. What matters is that the resulting age at death distribution is certainly statistically valid, even though a couple of individual cases might not be. Once again, it must be said that, in practical terms, age validation, whether historical or contemporary, cannot be dissociated from its context, both in terms of the available sources of information and of the importance of each individual case.
[ Home | Contents | Return to previous page ]
Max-Planck-Gesellschaft 2003