Age Validation of Reported Centenarians before 1900 in Denmark
by A. Skytthe, G. Hauge, and B. Jeune
[ References ]
The reports of extreme ages in previous centuries have not been questioned seriously until the middle of the last century. In 1873 Thoms introduced some guidelines to be followed when the age of an alleged centenarian was to be investigated (Thoms 1873). Later Bowerman (1939) examined cases of extreme longevity. Six cases were considered as having lived more than 110 years, among them the Hon. Katherine Plunket, who is still accepted as a supercentenarian (see Thatcher in this monograph). Only one man was accepted by Bowerman as a supercentenarian, Pierre Joubert aged 113 years, but Charbonneau (1990) revealed in a study based on genealogical records that it was a case of mistaken identity, whereby the deaths of father and son, both having the same name, were confused. Bowerman also examined some well known cases of outstanding longevity from earlier centuries and found similarities between several cases: Resident in distant localities, poor in vital statistics, illiterate, no evidence of date of birth but oral testimony of the old person or interested relatives. These findings inspired Bowerman to conclude that "..faking of longevity is undoubtedly a favourite indoors sport among illiterate people".
The guidelines proposed by Thoms are very useful in establishing a framework for the validation of alleged centenarians. However the degree and detail of validation depends upon the number of centenarians and their ages to be validated. Furthermore one has to take into account the availability of sources of individual information in different time periods. Few attempts have been made to validate the extreme ages in a population on the basis of individual cases, more often the validation of old age reports has been made from examination of the age and mortality structure (Kannisto 1994).
In this chapter we will discuss some of the problems confronting the validation of extreme ages in previous centuries in Denmark, which has a long tradition for the registration of demographic events like births, marriages and deaths. We will focus on alleged centenarians, who died in the nineteenth century in Denmark, but also take a look at alleged centenarians from earlier centuries.
The starting point of any age verification is a given age at a given date. In case of deceased centenarians this will usually be the death registration at date of death, an age and a name of the person that died. The ultimate goal will be to find the correct date of birth for this person.
However it is very unusual that a date of birth can be found directly. Often it is necessary to estimate a date of birth indirectly from several sources. Further information will be needed depending on the degree of certainty that is wanted. Subsequent information will include marriage, children, confirmation - all events that are registered in the parish registers.
The earliest available parish register in Denmark goes back to the 16th century, but few are left from that period. In 1645 the King issued a decree, by which the clergy was requested to register all births, baptisms, marriages and burials occurring in the parishes. Thus from the middle of the 17th century most parishes registered these demographic events in their parish registers. However there were few rules as to what should be written in the parish register and how. Some had almost the form of a diary, where baptisms, marriages and burials were mentioned as they happened, whereas others were kept very neatly (Ørberg 1970).
Until 1812 only one copy of the register was maintained in each parish. Due to fires or war some of the parish registers were lost, and this was a problem for the administration, because the parish registers were the fundamental instrument in setting up tax payment lists, enrolment lists and censuses. Therefore a decree was issued in 1812 requiring two copies of the parish register to be kept separately. At the same time preprinted books were introduced in order to standardize the management of the parish registers. During the next couple of years the changes were implemented, and in 1820 all parishes had changed to the new type of registers. The basic layout of the parish register has not changed since, and it now consists of seven sections: A section for each sex for births, confirmations, and deaths, and finally a section for marriages.
No obligation has existed for the clergy to mention the place of birth of people being married or having died. From the middle of the 19th century they were requested to write the place of birth, when children confirmed their baptism at the age of 14, but this is of little value if age validation of centenarians in the 19th century is your task.
This brings us to another valuable tool in the validation process, namely census lists, and here the censuses from 1845 and onwards are especially valuable. For the first time the census in 1845 included a question on place of birth. The accuracy of the answer depends to some extent on how close to the counting place the person was born; persons born in the same parish as the counting parish can be regarded as reliable, while persons born in another distant county or even another country often only give the name of the county or country.
The first census in Denmark was held in 1769, but the names of people were not stated as well as some special groups of the population were not included (e.g. soldiers). In the next two censuses, in 1787 and 1801, the names and ages of all citizens were stated. It was the responsibility of the local vicar in a parish that every person in his parish was recorded, and at the census day (a Sunday) he called for a representative of each family or household to appear at the vicar's house to inform of the members of the household. The next census was in 1834, and since 1840 a census was held every five or ten years. From 1834 to 1860 local enumeration commissionaires performed the counting by walking from house to house, writing down every person in each household. From 1870 until the latest "paper-based" census a hundred years later the census lists were filled in by the house owners themselves.
As mentioned above the censuses are very valuable tools in the age verification. Not only is place of birth mentioned from 1845 and onwards, but if the person remains in the same parish, it will be possible to follow that person through several decades, thereby establishing a temporal relation in the life course of the person. By comparing the ages given in different censuses a qualified judgement on the alleged age, and also an estimate of the date of birth, can be given.
The tools mentioned so far are all on a population basis. And they are to be considered as the basic tools in the verification of the age and identity of a person, since it is mandatory for a child to be registered in the birth register when born, and likewise a death has to be registered (the exception to that will be deaths occurring in war time or persons that just disappeared on land or sea).
For selected groups in the population some additional sources may exist. List of members of a trade (e.g. carpenters, blacksmiths, butchers) may help tracing persons through time and thereby increase/decrease probability of an alleged age. The genealogy of members of the aristocracy is often well studied and descriptions can be found in the archives. Teachers and vicars are professions that have a long tradition for literacy, and records are kept on almost every member of their profession. Probate court registers may be helpful, if you want to perform a detailed description of a persons lifetime, and want to find relatives, e.g. children. Regional or local administrative archives may give information about the person. One example is the protocols from the Public Assistance Committee in each parish. Since the beginning of the 19th century each parish had a Public Assistance Committee that provided poor relief to old and poor people who had no relatives to take care of them. These protocols may be very helpful because many of the alleged centenarians were poor and had to rely on help from the community in which they lived.
On a regional or local basis supplementary registers or indexes may exist. In Copenhagen separate lists of deaths were kept. In towns where newspapers were published obituaries and death advertisements of people that lived in the town or nearby can be helpful, and several lists of deaths have been compiled and published (e.g. Richter 1901-05).
In the burial registers before 1800 it was common not to mention a deceased woman by name, but by her relation to the head of the family, e.g. as the wife or widow of her husband. Since the name is often the only way to identify a person, the first step is to find the marriage of the couple. In other situations, e.g. when you try to do a family reconstitution, and it is desirable to make a list of all children of a couple, the problem may be that dead children may not be mentioned by their name.
Another complicating factor is the widespread use of common names. In many areas very few names were used and that makes it very difficult to make the correct link between two registrations of a person. In a study of the family structure in the 18th century in Denmark Johansen (1975) found that names like Ane, Karen and Maren among women and Christen, Hans, Jens, Niels and Peder among men each accounted for more than 10 percent of the names. Moreover the same person may be referenced with different surnames: Genealogic/patronymic (as the son or daughter of the father), occupational (as a blacksmith), or geographical (as coming from some town or island).
Some confusion may arise from the age statements in various sources. Until the census in 1870 the age of a person was supposed to be given in terms of commenced years, e.g. an age report of 90 years actually means that the person is 89 years old. But in the burial registers the age at death is often given as completed years of life.
The degree of verification has to be considered in relation to the importance of the case being verified. A low level verification may include collecting little additional information to the death registration, like one or two census entries for the person if not the birth or baptism of the person can be found. A medium level verification may in addition to the above include all census entries before the death of the person, and for males an entry in the enrolment lists. Finally a high level verification will have to include a total family reconstitution with marriages of the person, all brothers and sisters and their deaths in order to rule out any name-saking or other mistakes of identity. Thus it has to be considered - before you engage yourself in an age verification process - that the time spent may be quite high. The time invested has to be balanced by the importance of the subject: Verification of an extremely high age, say 110, for a single person may take several days or weeks of search in the archives, but can be justified by the exceptional age.
The Danish Centenarian Register contains reported centenarians in Denmark, in total about 4,500 Danish centenarians, hereof about 3,500 since 1960. The historical part contains centenarians that have died at an alleged age of 100 years or more, while the contemporary part contains all living centenarians in Denmark, almost 500. Since this chapter concerns age verification of centenarians in centuries before the 20th century, only the historical part of the Centenarian Register dealing with pre-twentieth century centenarians will be treated.
Number of reported centenarians in 1775-1899. Numbers are taken from the official death statistics
From the official statistics an estimate of the number of centenarians is available since the end of the 18th century, where the number of deaths in the age group at 100 years and above was reported by the clergy in all parishes and collected by the central administration. Table 1 shows the number of reported centenarian deaths in Denmark from 1775 to 1899 (Johansen 1975; Danmarks Statistik 1905). However, not all alleged
centenarians are included in the historical part of the Centenarian Register. In order to be included it is required that the person is identified by name, date and place of death. Until now (August 1998) a total of 197 of the 788 reported centenarians from 1775 to 1899 have been identified by name and place of death, mainly in the period from 1840 to 1868. The primary source for identification has been the original yearly reports on numbers of births, marriages and deaths from the parishes to the Statistical Bureau. They have been used for locating the parishes which reported the death of centenarians, followed by a search in the parish registers in order to identify the centenarian. In some cases it was not possible to find a centenarian in the expected parish, indicating some reporting errors in the official statistics (see below). Unfortunately the reports were only archived until 1868, thereafter the number of identified centenarians is sparse because no other available sources are at hand on a national scale.
Number of dead centenarians found in parish records on Funen from the period 1644-1899. Each bar represents the counts of a decade.
At the regional archive of Funen indexes of deaths and marriages have been compiled for almost each parish on the island on the basis of the parish registers, covering the period from the earliest available parish register (about 1650-1700) to 1891. The indexes made it possible to do a thorough search for reported centenarians covering the period of available parish registeres in a region of Denmark. Therefore the indexes were searched for reported centenarians. If an index did not exist for a parish the parish register itself was searched. A total of 273 alleged centenarians were identified with 135 before 1775 and 138 after 1775. The number of centenarians identified from 1644-1899 is shown in Figure 1.
Generally a low level of verification was performed. For the 246 reported centenarians before 1840 from Funen available information on the alleged centenarians was collected from the parish registers together with some additional information on spouses or children, when available. Whenever possible the census lists were searched, i.e. when the centenarian had died after 1787.
Based on additional information the probable age at death was estimated, and the alleged centenarian was assigned to one of two groups, indicating the probable status of the alleged centenarian:
|A)||The estimated age is 100 years or more, meaning that the alleged centenarian cannot be rejected as a true centenarian, and|
|B)||The estimated age is below 100 years, meaning that the alleged centenarian can be rejected as a true centenarian.|
When more than one additional source was available, higher priority was given to the age statement closest to the birth of the person, assuming that the younger a person is the better he can give his correct age.
As it may be expected from the description above of tools for validation, only a few of the early centenarians can be verified. In only 46 cases of the 246 identified centenarians from Funen additional sources of information have been found, mostly a single entry in a census list. For the remaining 200 cases it was impossible to find additional information and therefore no age validation could be attempted. The result is shown in Table 2.
Result of validation of centenarians from Funen 1644-1840
Evidence for the 14 reported centenarians from Funen, 1787 to 1839, that cannot be rejected as centenarians
In Table 3 the available information is given for each of the 14 cases that could not be rejected as centenarians. Only one additional source of information was found for 10 cases and two additional sources of information for the remaining 4 cases. This implies that the quality of the validation is not very good, and that several of the 14 alleged centenarians may be rejected as true centenarians given the fact that the census took place a few years before the death of the alleged centenarian.
A 109 year old man, who died in 1823, is among the 14 centenarians that could not be refuted. The available information consists of notes from the Public Assistance Committee's protocol, where he is mentioned every third year from 1806 to 1823 and each year his stated age is compatible with the numbers of years passed since the last note. But the ages given cannot be regarded as independent information, and therefore the notes have to be counted as one additional source. Furthermore he could not be found in the 1801 census, where he was supposed to be 85. We clearly need additional information from an earlier period in order to really accept his age claim, and the same goes for several others in Table 3.
Reported age at death among 246 centenarians from Funen 1644-1840
Looking at the age structure of the 246 reported centenarians from Funen (Table 4) there seems to be an over-representation of the age of 100. Although we do not know the number of persons that died at age 99, the number of persons at age 101 is lower than expected using the heaping indicator proposed by Kannisto in the monograph (see the chapter of Kannisto). In spite of this age heaping even the 105+/100+ ratio is to high (17.7), indicating overstatement of extreme ages. It is very likely that this just represents the notion of a very old person: When no precise age was known, and the person was very, very old, he was assumed to have reached the remarkable age of 100 years or even much older - an early form of the centenarian cult as Peter Laslett names it.
The age of reported centenarians in the period 1840 - 1899 was checked mainly with the help of parish registers and censuses. Special attention was drawn to the centenarians that were reported to have lived to the age of 105 or more. Any indication of birth or place of birth was pursued, but often it was impossible to find the birth registration due to non-existing parish registers.
Among the 192 centenarians reported in the period 1840-1899 (Table 1) two of the alleged centenarians could not be found in the expected parish, instead two persons aged 95 and 96 at death respectively were found indicating reporting errors from the parishes. Furthermore additional seven alleged centenarians have been found in parishes where no centenarians were reported in the death statistics.
Result of validation of Danish centenarians 1840-1899. The two persons aged 95 and 96 (see text) are included in the group of rejected centenarians
In Table 5 the result of the validation process is shown. Of the 199 centenarians that have been reported in the period 1840-1899 a total of 118 have been identified with name and date of death. It follows immediately that in the period 1840-69, where nearly all centenarians are identified, only about half of the alleged centenarians were actually centenarians based on the estimated age at death. More men, 17 out of 32 validated, are rejected as centenarians than women, 29 out of 65 validated.
The number of rejected centenarians in Table 5 has to be regarded as a minimum. In most cases, information from only one or two additional sources supports the age claim, and with more available information from a higher number of additional sources there will be a tendency to reject the person as a true centenarian. Furthermore, in the case of a single additional piece of information like a census registration, the census could have taken place just a few years before the death of the alleged centenarian. The falling number of not refuted centenarians from the 1840's to the 1860's indicates that a thorough validation of centenarians in the 1840's may not be feasible with the available sources.
Due to lack of identification sources of centenarians after 1868 only few centenarians from 1870 to 1899 have yet been identified and subsequently validated. The above validation can therefore only give a hint of the pattern of age at death for alleged centenarians in the three last decades of the century, but the validity of the age statements seems to increase towards the end of the century.
In order to exemplify the age validation considerable effort was devoted to the validation of two reported extreme ages. The first is the former sergeant Bartholomæus Albrecht, who died in 1841 at the age of 118½ years. The second person is Mads Pedersen Ribe, who died in 1864 at the age of 109 years. In each of these cases more than one week was used in search of additional information supporting - or disproving - the remarkable high age. A third example represents the normal effort devoted to an age verification. This is the 102 year old Niels Thorsen, who died in 1846.The time spent on verifying his age was approximately one day.
On the 17th February 1841 Bartholomæus Albrecht died in Ubberud, Odense county, at the age of 118½ years. At his time he was a well known figure in Odense, and his death was reported in several local newspapers with long obituaries mentioning his career as a soldier and former sergeant in the Danish army. He was said to have enjoyed extraordinary good health and strength. Presumably he would carry a burden of several liispund to Odense (a distance of 10 km) rather than have the goods transported by truck. At his birthday he several times visited the royal Prince Christian, the later King Christian VIII, who for some years resided in Odense as governor. The Prince also arranged for a portrait to be painted of the old sergeant (painted by Bærentzen in 1838). The story goes that the Prince offered a room to Albrecht in Odense, so that he could avoid travelling 10 km from his residence to Odense, but Albrecht refused the offer because he would not let his young wife stay alone in the house. So he walked 20 km every day during the making of his portrait! (Bloch 1899).
According to notes in the parish register and in census lists, he was born near Nürnberg in Germany in 1722 and served in the war between Prussia and Austria in 1756-63. There are different reports on when he came to Denmark, both 1757 and 1762 have been mentioned. He served as a soldier in several regiments until he finally received his pension and settled in Ubberud parish.
Being born in another country, and also being a soldier, makes it almost impossible to verify his age claim with a birth registration. In this way he resembles many of the alleged super-centenarians from earlier centuries as mentioned in the chapter on Luxdorph, for example Carolicopsky and Drakenberg. But he lived until 1841, which is 70 years later than Drakenberg, thus enabling us to find some additional information.
First of all, four censuses were carried out, all in which he has been found. Secondly, according to the probate court protocol, he was married twice, and both marriages have been confirmed by entries in the parish registers. In the obituaries it was mentioned, that he had 12 children, of which 5 were still living. In the probate court protocol 5 children are also mentioned, although the residence of two of them was not known. By searching various parish registers seven children of his first marriage have been found.
With the help of census lists and parish registers a chronological description of his life can be made (Table 6). It is by no means complete - more information and details are to be found, but it represent what has been possible to collect within a week's search in the archives.
From the available information it is very difficult to dismiss Bartholomæus Albrecht as a true centenarian. On the other hand it is most unlikely that he really attained the remarkable age of 118 years. If the age stated in the census in 1801 is to be taken as valid, he still would have reached the very unlikely age of 114 years. More reasonable is the age stated in 1787, 50 years, which leads to an age estimate of 104 years at death.
Available information about Bartholomæus Albrecht, who died at the alleged age of 118 years, in Ubberud, Odense County, Funen in 1841
Although no evidence can be brought forward for the following hypothesis, a possible explanation of his apparently extraordinary good health together with the stories of his walks to Odense could be that it was his father who was born in 1722 and had served as a soldier in the Austrian army. Albrecht might have been born around 1740, and then joining the Danish army about 1760, he gave his fathers date of birth as his own in order to get advantages in the Danish army.
The case of Mads Pedersen Ribe was briefly mentioned in Skytthe & Jeune (1995). At that time only a modest age verification had taken place.
On April 24, 1864 the death of a man called Mads Pedersen Ribe was registered in a countryside parish, Rorslev, near Middelfart on Funen. His age was stated as 109 years, and in justifying this remarkably high age the vicar added in the burial register "Born in Ribe 24 June 1755". Given such an exact date and place of birth implies that the vicar must have had some additional information, either from some documents or possibly obtained by himself: The vicar may have tried to verify the age of Mads Pedersen Ribe.
The name Ribe implies the place from where Mads Pedersen came. Ribe is an old town in western Jutland, and looking in the birth register from Ribe Cathedral we actually find an entry saying "... 24 June 1755 Mads, son of Peder Madsen, soldier". No other information is given, not even the name of the mother.
In the probate register we find no additional information - he was a pauper and left no possessions, and no information is given about heirs. A remarkable thing is that we have not been able to find him in the 1860 census from Rorslev parish - a presumably 105-year old man ought to be easily found by looking in the age column alone. On the basis of the above information it was not possible neither to refute nor to confirm the status of Mads Pedersen Ribe as a centenarian.
A more thorough verification has taken place since 1995. First of all we found an entry in the 1850 census list of a neighbouring parish, Asperup, with the following statement: "Mads Peder Ribe, 83 years, widowed, day labourer, receives some benefits from the Public Assistance Committee. Place of birth Ribe". Another entry was found in the 1840 census list from Rorslev parish, the poorhouse, reading: "Mads Pedersen, 85 years, widowed, staying in turn at different families in the parish". Of the two entries the 1850 entry seems to match the deceased Mads Pedersen Ribe better than the 1840 entry because of a closer likeliness of the name. The information about receiving public assistance lead to the protocols from the Public Assistance Committee. According to these protocols Mads Pedersen Ribe had received poor relief at least since 1852: Each year he received clothing and some years money for tobacco and help in case of illness.
Being widowed implies that he had been married. Searching the burial register of Rorslev we found the death of "Anne Pedersdatter, 54 years, wife of Mads Pedersen in the poorhouse in Blanke" dated 10 September 1831. Their wedding took place 23 May 1817 in Rorslev: "Smallholder Mads Peter Pedersen Ribe, 45 years old, Blanke, and maid Anne Pedersdatter, 39 years old, serving at Holsegaard". Immediately after the wedding they had a son, Anders, born 21 June 1817, and he confirmed his baptism in 1832. Some confusion obviously existed because his confirmation is noted both in Rorslev parish and Asperup parish, and the parents' names are slightly different in the two parishes: In Rorslev the parents were listed as: "Mads Peter Ribe and Ane Pedersdatter from poorhouse in Roerslev. Note: He (Anders Madsen, AS) belongs to Asperup" while in Asperup they are listed as: " smallholder Mads Peter Pedersen and wife Anne Pedersdatter from Blanke, but .. is for farmer Jens Christiansen in Baaring", meaning that Anders was serving at Jens Christiansen's at that time.
Thus there seems to be no doubt that Mads Pedersen Ribe became widowed in 1831, and that he was married in 1817, presumably at the age of 45 years. However going further back in the parish register of Rorslev reveals further information about Mads Ribe. The marriage to Anne Pedersdatter was his second marriage. He was widowed a half year earlier: 26 November 1816 Ane Marie Jespersdatter died, 54 years, wife of smallholder Mads Peter Pedersen of Blanke. Their marriage entry is found in Rorslev 22 February 1806: "Mads Pedersen from Veilbye and widow Ane Marie Jespersdatter from Blanke" - it was his first and her third marriage.
The findings mentioned so far indicate that Mads Pedersen Ribe died at a younger age than 109 years, but not how much younger. However, a pretty close estimate of his correct age arises from the entries in the enrolment list. Knowing that he was married in Rorslev in 1817 we managed to find his entry in the enrolment list from Rorslev in 1817, and here his age was stated as 41 years old. Since there are links to earlier entries also we could follow him back to the enrolment list from Veilby in 1801, where his entry (list L140/381) reads: "Son of Peder Madsen, Mads; place of birth: Ribe; Age 26 years; Place of residence: Gamborg; Acc to Resolution of 14.7.1801 from the Chancellery hired as a land soldier for 8 years by Ole Svendsen (son of Svend Sørensen) Place of birth: Kustrup Age 18 years who pays 30 Rdl.". In the National Archive we also found a note concerning the capability of Mads Pedersen Ribe to serve as a soldier. The note says that he is capable to serve, but not obliged to serve, indicating that he was born in a town and not in a rural area, since only men from the rural areas were obliged to serve as soldiers.
From 1801 to 1820, when Mads Pedersen was erased form the enrolment list, we can follow him every three years in various enrolment lists, consistently with an age corresponding to being 25 or 26 years old in 1801.
Thus it seems very likely that Mads Pedersen Ribe was born in 1775 or 1776 in a town, probably Ribe, and that he died at the age of 89, not even close to being a true centenarian. We can only guess how he could be regarded as 109 years old, but one could imagine a confusion between two persons both named Mads Pedersen as indicated by the former mentioned two entries in census lists from 1840 and 1850.
According to the parish register of Fjellerup parish, Vends District, Odense County, a man by the name of Niels Thorsen died 6th April 1846 at the age of 102 years. However, reading in the probate court protocol the credibility of the age statement is severely damaged, because on 6th April 1846 the probate court protocol for Vends District reports the death of Niels Thorsen, aged 100 years.
To our luck he lived almost all his life in the same parish thus making it possible to make a complete list of ages reported at the censuses in the period from 1787 to 1845 (Table 7). Also shown in Table 7 are the corresponding ages of his wife, Karen Rasmusdatter, with whom he was married 2nd October 1784. It is remarkable that a few years are added to the age of Niels Thorsen in every census and that there are errors also in the age of his wife.
Again we have to dismiss the claim of Niels Thorsen to be a true centenarian - from the available information he probably died at the age of approximately 85 years.
Age reported in censuses for Niels Thorsen and his wife
The differences between age statements at different periods are well known from other historical demographic studies. Johansen (1975) investigated a sample of the Danish population in the 18th century and with the help of family reconstitution methods the ages in the 1787 and 1801 censuses were compared for the identified persons. Although differences were found on the individual level, no trend in either age exaggeration or the opposite in different age groups was found on the population level. The sample, however, consisted mainly of people below the age of 80, and age exaggeration is expected to be a problem especially among old people. A study of the validity of old ages above 80 in a region would be of great value in order to illuminate whether or not age exaggeration is a problem.
One conclusion to be drawn from this study is that it is very difficult to attain a reliable and valid description of the oldest-old mortality in previous centuries. The age at death of old people should not be used without precaution, and without a thorough and careful investigations of the age at death of the old population false and incorrect conclusions may be drawn from such studies.
We thank Walther Limkilde for his great enthusiasm and help during the identification and validation process. Grethe Banggaard supplied valuable information about Bartholomæus Albrecht. We are also grateful for the help from numerous genealogists, who shared their knowledge of centenarians which they have come across during their private studies.
[ Home | Contents | Return to previous page ]