Early Modern Demographics: The Isle of Thanet 1560–1640. A Cohort Study
The author provides a fascinating demographic analysis of the Isle of Thanet which uses baptism, marriage and burial registers of the seven parishes from a sample of 8,000 records. The primary data provides an historical insight to the population of the Isle of Thanet and raises interesting questions in relation to the family reconstitution in these local Kent communities.
‘Nasty, brutish and short.’ Such was Hobbes’ famous definition of human life in his 1651 book Leviathan.[fn1] It is a phrase often repeated even though he applied it only to those living in wartime, but is it true? To a large extent, this cannot be known for the first two factors are subjective. From our perspective, people may have lived in squalor with few of the possessions or opportunities which we associate with happiness, but it is unlikely that many endured lives devoid of a single smile or laugh or hug and the ability to appreciate the beauty of a spring morning or music is not dependant on material income. The British, after all, are renowned for making the best of things and always looking on the bright side of life. The question of whether lives were short is, however, quantifiable, and it is the purpose of this paper to look at the evidence of what became of almost eight thousand people born on the Isle of Thanet between 1560 and 1619.
The late twentieth century saw the development of a technique known as family reconstitution. This involved copying out the names of people baptised from a parish register and then going through the marriage and burial registers of the same church to try and discover what became of them before analysing the resulting data to draw conclusions about things like mortality and nuptiality. The technique worked best in the pre-Restoration period before the proliferation of non-conformist churches for which records often did not survive but it was not widely used because it took thousands of hours to reconstitute a single parish. To this date, less than fifty parishes across England have been fully analysed and nobody has ever attempted a group restitution such as is presented here. Moreover, restricting the study to the use of parish registers reduced the number of people who could be traced. There might be three men of the same name in the parish but which one died in each year could be impossible to discover without the assistance of other information such as wills or memorial inscriptions. This led to an enhanced technique known as total reconstitution which employed these sources together with taxation records, marriage licences, apprenticeship documents and anything [pg136]else which might shed light on the names in the parish register. The results were fuller but such studies were even rarer and it remained true that only a minority of people could be traced. Few people spent their entire life in their parish of birth and given many of those who did so died as infants, data gathered on life expectancy was inevitably skewed.
For this study, the total reconstitution technique has been used for all seven parishes on the Isle of Thanet. Most people who moved at this time, travelled a relatively short distance so this enables considerably more individuals to be traced. The Isle of Thanet has been chosen because it is a separate geographical entity and it has excellent records. Not only are the parish registers of the main five churches virtually intact, but two parishes have comprehensive rate assessments and details of payments to the poor. There are also various muster rolls and various shipping surveys to be found in the National Archives which provide details about residence and occupation. It was not uncommon in the early seventeenth century for couples from Thanet to travel to Canterbury to get married before returning to the island to settle down so there are many relevant marriage licences. The fact that around a third of the adult men who died left a will has also proved beneficial. By combining all this data and looking at the records of other churches across the county,[fn2] it has proved possible to ascertain the age of 73 per cent of brides at marriage, 67 per cent of grooms, 81 per cent of those being buried and to trace 87 per cent of those baptised through to death or at least adulthood. This last is double what would have been achieved with family reconstitution alone. Using that method, just 43 per cent would be traced and over a third of those would have died in infancy.
Not only have efforts been made to trace everyone baptised on the island from 1560 to 1619, but a similar reverse exercise has been applied to burial registers from 1560 to 1699 where ages have been added. In the early period there are clearly many gaps for Birchington is the only parish with records back to 1538. In some cases, it is impossible to know the exact age because the birth took place before registration began. In 1647, the clerk at St Peter’s recorded that Margaret Shonk was a widow of ‘near a hundred years old.’ It might be thought that they exaggerated but Margaret had been married in the church in 1573, seventy-four years earlier, so she must have been well into her nineties. Entries of complete outsiders are excluded both for reasons of impracticality in tracing them and because they would skew the figures. The maritime parishes all regularly recorded the death of strangers in shipwrecks as well as names of soldiers and sailors who had perished from the scars of battle or resultant infection at particular times such as the Armada and the wars against the Dutch and Irish.
For all the marriage entries from 1560 to 1640, efforts have been made to enter the age and place of birth of each party, the number of children born, the number who died in infancy, the birth intervals, and the dates of death plus, if applicable, dates of remarriage, for both partners.
The result is a considerable body of data which casts light not only on people’s lives in this period but on the opportunities open to them and the particular issues faced by different parishes. For example, nearly a third of burials in St Peter’s were of people aged over fifty compared to less than a fifth in Minster, and is calculated from baptism records. Although an inability to work due to accident or sickness can affect people of any age, it is more common as people get older. It [pg137]is likely, therefore, that St Peter’s faced more demands on its purse in supporting the frail elderly than Minster. To reiterate both parishes have very full data, and Minster was plagued by malaria, the coastal towns were not. Life expectancy in the marshes was lower, hence less elderly to tend to in Minster.
Table 1. Average number of baptisms, marriages and burials per church 1560-1639.
[tb][th]Parish|Birchington Bap|Birchington Ma|Birchington Bur|Minster Bap|Minster Ma|Minster Bur|Monkton Bap|Monkton Ma|Monkton Bur|St John Bap|St John Ma|St John Bur|St Laurence Bap|St Laurence Ma|St Laurence Bur|St Nicholas Bap|St Nicholas Ma|St Nicholas Bur|St Peter Bap|St Peter Ma|St Peter Bur[/th]
[tr]1560s|13.4|3.7|13.1|14.1|3.4|16.5|1.5|1.0|7.5|25.7|8.0|23.5|22.9|6.8|19.1|6.8|4.0|6.0|17.1|5.0|12.9[/tr]
[tr]1570s|15.4|4.0|13.0|15.0|6.8|14.7|3.5|1.0|7.5|25.7|8.0|23.5|22.9|6.8|19.1|6.8|4.0|6.0|17.1|5.0|12.9[/tr]
[tr]1580s|19.5|6.3|12.2|17.6|6.7|18.3|3.0|2.0|5.5|18.5|5.0|18.7|29.4|9.9|25.3|6.6|2.2|4.4|22.8|7.2|18.6[/tr]
[tr]1590s|18.1|6.6|15.4|18.7|9.2|20.7|3.0|1.0|4.0|26.2|4.1|15.6|32.6|8.4|20.6|7.8|5.3|5.5|30.3|6.7|13.8[/tr]
[tr]1600s|18.8|5.2|11.8|17.3|7.7|19.8|no data|2.8|3.0|32.1|7.0|17.4|30.1|9.7|24.5|5.5|2.0|5.3|29.1|7.0|15.2[/tr]
[tr]1610s|19.1|6.2|20.6|19.0|7.3|22.6|5.3|no data|no data|37.6|9.5|32.3|35.1|9.1|23.8|13.0|0.3|1.8|26.0|7.5|18.0[/tr]
[tr]1620s|18.5|5.1|19.3|19.3|4.5|20.6|5.9|1.2|3.2|39.7|13.5|32.7|38.8|12.2|25.4|13.0|1.0|6.0|28.1|7.7|18.9[/tr]
[tr]1630s|21.2|6.4|22.1|18.0|5.2|17.8|6.8|2.6|6.1|46.2|12.7|38.7|43.1|10.5|32.8|13.3|4.7|14.9|29.9|6.0|21.9[/tr]
[/tb]
[pg138]
Key: Bap – Baptisms; Ma – Marriages; Bur – Burials
Population
The Isle of Thanet consists of seven parishes. The three maritime ones are St Laurence (Ramsgate), St John (Margate) and St Peter (Broadstairs). The four rural ones are Birchington, Minster, St Nicholas-at-Wade and Monkton. The distinction is not absolute for the maritime parishes in this period were predominantly rural too in terms of land use but the existence of substantial numbers of sailors meant there were key differences in occupational structure, economy and diet. At the end of Elizabeth’s reign, the average number of entries per year and per parish is outlined in Table 1.
None of the communities were urban and in places like Monkton, the birth of a child or the celebration of a marriage would have been a significant and newsworthy event. Yet the spread of events does enable some estimate to be made of population size. Accurate measurement of population in the era before census returns has always been problematic but the most common method has been multiplying the average number of baptisms in a decade by thirty or thirty-five.[fn3] Using this method[fn4], the approximate populations would be:
Table 2. Estimated population by parish 1570 to 1639.
[tb]
[th]Decade|Birchington|Minster|Monkton|St John|St Laurence|St Nicholas|St Peter[/th]
[tr]1570s|501|488|130|601|956|335|741[/tr]
[tr]1600s|611|562|163|1222|1141|423|845[/tr]
[tr]1630s|689|585|218|1609|1573|480|1229[/tr]
[/tb]
For the demographer, another tool to use is the life table. This requires knowledge of the ages of everyone buried during the course of the year.[fn5] This information does not exist for every year but taking three for which it does, it is possible to compare results with those created by multiplying baptisms. In 1615, the life table shows 5278 while multiplying shows 5346. In 1642, the life table shows 5762 while multiplying shows 5996. In 1675, the life table shows 5445 while multiplying shows 6992. The discrepancy in this last set of figures is because in the 1670s, population growth was rampant with baptisms regularly being more than double the number of burials in the coastal parishes. Given that after the Restoration there were also nonconformist churches in the area, the number of births would have been higher than suggested by the Anglican registers.
The life table also enables a population structure to be created which highlights the youthful nature of early modern society (Chart 1).[pg139][pg140]
[fg]jpg|Chart 1. Population pyramid early modern Thanet.|Image[/fg]
Life expectancy
The ages at death are known of 5312 of the 7901 people born on the island during the period 1560-1619. However, a further 579 males and 938 females are known to have survived to adulthood because they married and it is important to incorporate these in order to prevent childhood deaths appearing disproportionately high. If these individuals whose dates of death are unknown are added to those known in the same ratio, it will be seen that people at birth had the following probability of dying in each age range (Table 3).[fn6]
Table 3. Chance of dying in each age by gender based on the ages at death of all those in the birth cohort.
[tb][th]Age range|Male (1 in x)|Female (1 in x)[/th]
[tr]First year of life|5|6[/tr]
[tr]Age 1 to 4|17|15[/tr]
[tr]Age 5 to 9|31|29[/tr]
[tr]Age 10 to 19|21|24[/tr]
[tr]Age 20 to 29|12|8[/tr]
[tr]Age 30 to 39|10|7[/tr]
[tr]Age 40 to 49|7|8[/tr]
[tr]Age 50 to 59|8|9[/tr]
[tr]Age 60 to 69|9|10[/tr]
[tr]Age 70 to 79|13|14[/tr]
[tr]Age 80 to 89|36|33[/tr]
[tr]Age 90 to 99|433|281[/tr]
[/tb]
Infant Mortality
That many children died in infancy is well known but the rates disguise a whole range of differences. The place where a couple lived had a major impact on the prospects of their child surviving to adulthood. Children born in Minster were at significantly higher risk of early death than those born in the coastal parishes of Ramsgate, Broadstairs and Margate and males everywhere were more susceptible than females. The high death rate at Minster almost certainly was due to it being in low lying marshland. Whether the much lower rate on the coast was due to the bracing sea air or the higher fish consumption, which must be assumed due to the major industry of mackerel and herring fishing, is unknown. A diet high in omega-3 has proven protective qualities for mothers and babies.[fn7]
The occupation of the father appears to have had some impact on infant mortality. Over fifty-six per cent of sailors and fishermen, yeomen and gentlemen never lost a child compared to forty per cent of labourers and servants, fifty-one per cent of husbandmen and forty-eight per cent of craftsmen. Overall, of the 795 couples who had children, sixty per cent never lost any of any their babies in infancy; twenty-[pg141]eight per cent lost one, eight per cent lost two and three and a half per cent lost three or more. One in ten families lost all their children.
The age of the mother seems to have had little effect, although mothers aged fifteen to nineteen lost more (23.6 per cent of 104 births). Overall, the age of the mother is known in 2691 cases and from twenty to forty-five, the infant death rate was between 16.1 and 17.1 per cent. Infant deaths were markedly lower in the summer than in the winter and since many infant deaths occurred in the first weeks of life this meant that those born in the spring or summer were more likely to survive than those born in the autumn or winter.[fn8] Sadly, the number of children born in winter was substantially higher than those born in the summer showing the busiest period for conceptions was late April to July with the lowest being October and November.[fn9]
Between 1560 and 1629, a total of 1339 infants died. Of these, 56.2 per cent died within the first month of life. 200 infants were buried on the day of their baptism and 68 a day later. A further 295 (22 per cent) died in their second and third months of life. Without exact causes of death, something no medical practitioner of the time could have supplied anyway, it is impossible to classify which deaths were exogenous and which endogenous although the fact that some three quarters of infant deaths occurred in the first three months of life does suggest endogenous factors were significant.
Regarding multiple births, there was one set of triplets, all of whom died before baptism. There were 140 sets of twins and of those children, 40 per cent died in infancy. In 28.5 per cent of cases, both twins died, 48.6 per cent both twins survived and 22.9 per cent just one survived. Ordinarily during this period, children would be breast-fed for the first two years of their lives.[fn10] It is therefore unsurprising to find that of those children who lost their mother during this time, 39.6 per cent themselves died. Although it is impossible to confirm causes of death, 63 mothers were buried within two weeks of their child’s baptism suggesting that maternal mortality as a result of childbirth complications was one in 123 deliveries. Doubtless some other women would have died of related complications beyond this time period (Table 4).
Table 4. Infant mortality by parish.[fn22]
[tb][th]Parish|Males born 1560–1619|Buried in first year of life % (M)|Females born 1560–1619|Buried in first year of life % (F)[/th]
[tr]Birchington|530|20.94|493|16.23[/tr]
[tr]Minster|517|25.53|529|18.71[/tr]
[tr]St John|996|16.37|934|13.06[/tr]
[tr]St Laurence|931|15.90|989|11.32[/tr]
[tr]St Peter|846|12.53|820|11.95[/tr]
[/tb]
Parish registers naturally record dates of baptism and not dates of birth but the interval between the two events was likely to have been extremely small since the Prayer Book urged that children be presented on the first Sunday after birth.[fn11] [pg142]That people were inclined to follow the Prayer Book regulation is indicated by the fact that almost half (46 per cent) of all baptisms took place on a Sunday with just over a fifth (23 per cent) taking place on Wednesdays. Children who were baptised on other days were more likely to die in infancy suggesting that these were often emergency baptisms: almost twice as many of those baptised on a Monday or Tuesday died as those baptised on a Sunday or Wednesday. The rates for those baptised on a Thursday or Friday were more than fifty per cent higher and a third higher on Saturday. The prevalence of Wednesday, which exists in parishes of both traditional and puritan sympathies and throughout the entire period, may suggest that there was pressure to have children baptised as soon as possible, perhaps reflecting a belief held by some that those who died before baptism would not go to heaven.[fn12]
Nuptiality
Getting married was a regular rite of passage for people born in this period. The lives of over four thousand people have been traced from cradle to adulthood and the vast majority of those wed. Table 5 shows the number of those born 1560 to 1619 who are known to have died at thirty or more years of age but who had not married.
Table 5. Proportion of adults who never married.
[tb][th]|Number|Never married|%[/th]
[tr]Male|2207|246|11.15[/tr]
[tr]Female|2222|80|3.60[/tr]
[tr]Total|4429|326|7.36[/tr]
[/tb]
The average age of getting married remained constant throughout the period being 27.5 for men and 24.9 for women. October and November were the most popular months for weddings and Monday the favourite day. Two thirds of weddings were a first marriage for both parties and one in fourteen involved both parties being widowed. One in eight weddings involved a bachelor marrying a widow and one in seven saw a widower marry a spinster.
Although the wedding service called for the bride’s father to give her away, only 46 per cent of brides had their father alive at the time they got married (Table 6). It is unsurprising to find that of those born in Monkton, only around one in twenty grew up to marry in the village. With such a small population, choice of spouse and employment opportunities were scarce.
Amongst male islanders who married, 29.1 per cent married in the place of their birth and 35.6 per cent elsewhere on the Isle of Thanet. The remaining 35.2 per cent married further afield and only one in eight of these returned to their place of birth to die suggesting that they had moved away for work purposes and met a girl there. Marriage registers did not record the place of residence prior to the nineteenth century. Marriage licences did provide that information and they show many Thanet men married on the mainland, particularly sailors who were only home briefly for whom the three weeks of banns were inapplicable. The majority [pg143]of those who married off the island did so in the next parish, such as people from St Nicholas marrying in Chislet which bordered the parish and was separated only by the Wantsum. Although more than a name would be needed to associate someone with a match further afield or in cases of common names, a marriage does involve two people and a number of register entries. Consider, for example, Richard Meakin who was born and died in Margate. When he was twenty-six, he started having children there with a wife named Christine. They had several children and then she died and was buried as his wife. They did not marry in Margate but there is a marriage entry for Richard Meakin and Christine Frances at Chislet two years before the first child was born. There is no other Meakin in the Chislet registers and Christine is a rare name at this period. Indeed, the only Christine Meakin buried anywhere in the county in this century is Richard’s wife in Thanet. Given the proximity of Chislet to Margate and the dates, it seems certain that this wedding belongs to the Richard and Christine Meakin who died on the island. Similarly, links can be made when siblings all appear in another area indicating the family moved.
Table 6. Numbers of first time brides and grooms having parents alive at the time of their marriage.
[tb] [th]|Both parents alive|Father only alive|Mother only alive|Neither parent alive[/th] [tr]Bride|274|226|231|357[/tr] [tr]As %|25.2|20.8|21.2|32.8[/tr] [tr]Groom|214|223|189|370[/tr] [tr]As %|21.5|22.4|19.0|37.1[/tr] [/tb]
Amongst females, 34.7 per cent married in their parish of birth, 38.3 per cent elsewhere on the island and 27 per cent further afield. Less than one in twenty of those who married away returned to their birth parish. In the two major fishing areas of St Laurence and St Peter, almost three quarters of men married on the island whereas those born in the villages were much more likely to move away with over forty per cent marrying on the mainland. A similar pattern applies to women indicating the health of the maritime economy at this period.
Regarding the duration of marriages, just over a quarter lasted between ten and nineteen years (26.8 per cent), almost a fifth lasted less than four years (18.5 per cent) with a similar number lasting twenty to twenty-nine years (18.1 per cent). Just one in fifty couples could expect to celebrate their golden wedding anniversary together. Overall, a third of marriages lasted less than ten years.
The majority of those who were widowed remarried either by choice or economic necessity though unsurprisingly, those widowed in later life did so to a lesser degree. Three quarters of men widowed before the age of forty who had children remarried and ninety-four per cent of those who were childless but women of the same age with children clearly found it more difficult to find a new husband with only half of them securing another match. Widows over the age of thirty were significantly less likely to remarry whether they had children or not (Table 7).
[pg144]
Table 7. Remarriage interval of those widowed 1560 to 1639 whose ages and dates of death are known.
[tb][th]Age widowed|20–24|25–29|30–34|35–39|40–44|45–49|50+[/th]
[tr]Male – Up to 3 months|7.69|25.00|16.36|25.00|15.94|16.98|3.77[/tr]
[tr]Male – 4 to 12 months|53.85|22.92|30.91|31.58|30.43|11.32|6.92[/tr]
[tr]Male – 13 to 24 months|23.08|16.67|18.18|9.21|7.25|7.55|4.40[/tr]
[tr]Male – 25 to 36 months|0.00|2.08|0.00|2.63|4.35|1.89|3.14[/tr]
[tr]Male – Over 37 months|0.00|12.50|9.09|6.58|8.70|3.77|2.52[/tr]
[tr]Male – Never remarried|15.38|20.83|25.45|25.00|33.33|58.49|79.25[/tr]
[tr]Male – No. cases|13|48|55|76|69|53|159[/tr]
[tr]Female – Up to 3 months|0.00|9.09|6.67|3.90|14.47|1.75|0.40[/tr]
[tr]Female – 4 to 12 months|15.38|11.36|13.33|12.99|1.32|3.51|1.20[/tr]
[tr]Female – 13 to 24 months|30.77|13.64|2.22|9.09|5.26|5.26|0.40[/tr]
[tr]Female – 25 to 36 months|7.69|9.09|2.22|3.90|3.95|1.75|1.61[/tr]
[tr]Female – Over 37 months|15.38|15.91|24.44|10.39|2.63|5.26|1.20[/tr]
[tr]Female – Never remarried|30.77|40.91|51.11|59.74|72.37|82.46|95.18[/tr]
[tr]Female – No. cases|13|44|45|77|76|57|249[/tr]
[/tb]
Fertility
A total of 1024 completed marriages took place between 1560 and 1619 where the bride was aged between fifteen and thirty-nine. Of these, 23.8 per cent did not have any living children at all.[fn13] Only marriages completed on the island were included, and whilst it is possible that some couples married on Thanet, moved somewhere else for twenty odd years and then came back to die. It is likely that much childlessness was caused by maternal malnutrition and miscarriages. Dairy consumption would have been very low on the island and only women in the maritime parishes would have had enough fish. Pregnancies which ended in a miscarriage or stillbirth were not recorded. Of brides who were aged thirty or more, over a third were childless compared to just a fifth of those who were younger. Amongst the other couples, 17.3 per cent had one child, 13.5 per cent had two, 11.2 per cent had three, 9.9 per cent had four, and 25.6 per cent had five or more. The interval between marriage and first child was between nine and twenty-four months in almost two thirds of cases. One in nine brides was pregnant at the time she married. Table 8 shows how fertility varied between the age groups.
The major decline in fertility for those over thirty is a reminder that women in this period typically entered menopause in their mid-thirties (Table 8).[fn14] When Henry VIII expressed his concern about his wives’ fertility, he was not being paranoid or unreasonable but reflecting a common experience. Average intervals between [pg145]first and second children was twenty-eight months. Gaps thereafter increased from thirty onwards. After a child was lost in infancy, the gap was smaller, generally between fifty and twenty months.
Table 8. Fertility by maternal age from marriage to first birth, by age at marriage.
[tb][th]Age of bride|15–19|%|20–24|%|25–29|%|30–34|%|35–39|%|40–44|%[/th]
[tr]Had no children|41|23.3|74|18.9|55|20.9|44|36.1|15|31.9|13|76.5[/tr]
[tr]Less than 9 months|21|11.9|43|11.0|24|9.1|18|14.8|6|12.8|1|5.9[/tr]
[tr]9 to 12 months|33|4.2|108|27.6|75|28.5|26|21.3|13|27.7|1|5.9[/tr]
[tr]13 to 24 months|50|6.4|102|26.1|60|22.8|24|19.7|7|14.9|2|11.8[/tr]
[tr]25 to 36 months|16|2.1|43|11.0|29|11.0|5|4.1|4|8.5|0|[/tr]
[tr]37 months and over|15|1.9|21|5.37|20|7.6|5|4.1|2|4.3|0|[/tr]
[tr]Totals|176||391||263||122||47||17||[/tr]
[/tb]
Mortality
In a normal year, deaths were spread fairly evenly. By season they were: 19.1 per cent summer (June to August), 25 per cent in the autumn (September to November), 28.2 per cent in the winter (December to February) and 28.9 per cent in the spring (March to May). Within that there was some variation. Teenage boys were twice as likely to die in the early summer, possibly indicating a tendency to take to the water or engage in boisterous or dangerous games. For both genders, adults over the age of fifty were more likely to succumb in the winter.
There were some years in which crisis mortality could be observed, defined here as years in which the death rate was at least more than half as high as the average of that recorded in the previous ten years. However, in none of the years in question was the situation repeated across all the parishes indicating that some parishes were more successful than others in securing food in times of dearth or in containing infection. The fact that population density was around one person per five acres must have assisted in preventing the spread of disease.
At Birchington, the worst years were 1570, 1637 and 1661. Minster was affected in 1570 and 1597; St John’s in 1588, 1607, 1615, 1625, 1661, 1669 and 1690; St Laurence in 1607, 1625, 1638 and 1690; St Peter’s in 1638 and 1661. The Birchington plague outbreak of 1637 has been analysed previously and is unique as a clear example of the sickness.[fn15] The pattern elsewhere is much more [pg146]problematic. At St John’s, there was an infection which arose in August 1690 and the dead include four married couples, all in their prime of life with one partner dying a few days before the other in each case. There was no other clustering by family suggesting it was not plague. With no rate books, it is impossible to know if the couples involved lived close to one another or shared a water source.[fn16] At St Laurence in the same year, the crisis months were October and November but only one family lost two members.
In 1615 and 1669 at St John’s, over a third of the deaths occurred in children less than two years old. In 1615, there were no cases of families losing more than one child but in 1669, six families lost two or more suggesting a contagious disease such as measles or smallpox. The winter of 1607 to 1608 was recorded as being one of the coldest and the first three months of 1615 were said to have seen freezing temperatures throughout. Burial rates were notably higher at Birchington, St John and St Nicholas.
Migration
Most of those who moved away from Thanet, either as adults in search of work or as children taken when their parents moved, remained within the East Kent area. For females, the most popular areas were Sandwich (19 per cent), Chislet (12 per cent), Canterbury (10 per cent), Ash (7 per cent) and Deal (6 per cent). Amongst men, the favoured destinations were Canterbury and Chislet (both 13 per cent), Sandwich (12 per cent), Whitstable (8 per cent) and Faversham (6 per cent). Three times as many men as women ventured beyond the county. Of the 18 per cent who moved within the wider boundaries of the county, three quarters went toward Ashford and the Weald, just over a sixth to the Medway and the remainder toward Sheppey. Nobody chose to go to Romney Marsh.
The Missing
Just over one thousand of those baptised on the island have proved untraceable: the difference between male and female is less than one per cent. Those from Monkton and St Nicholas have the highest failure rates at just over a quarter but given the gaps in their early records, this is not surprising. At St Peter and St Laurence, the untraced rates are 10.4 and 10.9 per cent respectively, Minster is 12.5, Birchington 13.7 and St John, which appears to have had a more transient population due to trade, was 14.2 per cent. The biggest difference was that one in eight of those born before 1609 proved untraceable compared to one in six of those born 1610 to 1619. One reason for this increase is almost certainly the disruption caused by the Civil War. No marriages are recorded at St Nicholas or Monkton from 1646 to 1655, years in which some of those born in that period might have expected to marry. That represents the loss of maybe thirty weddings assuming that those churches would have seen the same rate of services as in the previous decade. Across the other five churches whose registers are intact, there is also a reduction in recorded weddings of thirty per cent. Burials too were affected. None were recorded at Monkton where around thirty-five would have been expected, just forty-seven at St Nicholas in place of the expected hundred and fifty, and across the other five [pg147]parishes there was a shortfall of some two hundred and twenty burials compared to what would have been expected.
There were other reasons for people proving untraceable. In almost forty per cent of cases, people had names which were sufficiently common that there could be no certainty which was which due to a lack of supplementary documentation such as wills. A further one in twenty had surnames which were liable to a variety of spellings on the part of parish clerks which offered considerable potential for mis-identification, for example, Dodd/Dadd/Dade/Dods, Little/Ittle/Liddall or Christian names which could be misread or miscopied, such as the male William for the female Wilman. A fifth were part of families which moved away when the children were small. This in itself should not be a problem given many families moved and their children could be traced in another area, but when none of the children could be traced and no burial record could be found for either parent, the possibility is that they moved further afield. Those who left the county would be impossible to find unless they had an extremely unusual name or there was some other evidence of their departure. Several of the Bachelor family emigrated to Massachusetts and the will of John Montstephen refers to one of his sons having gone to Devon with the sad comment by the testator that he has no idea if the boy is still alive.[fn17] It may be presumed that some individuals would have taken a boat from Margate to London and stayed there.
Twenty-three children were fully orphaned before the age of ten so may have been sent to live with relatives or godparents and ten were bastards who might have employed another surname. The number of illegitimate children was very low but the majority of those proved untraceable. That still leaves just over three hundred people who vanished into thin air.
Conclusions
The above is hopefully very interesting and the analysis is accurate based on what has been found. The question remains, are the source materials accurate? At one level, with the exception of Monkton and St Nicholas which have known gaps in the earlier part of the period, the answer appears to be yes. If entries are counted by year and month, there are no obvious gaps or suspiciously low totals. Aside from an odd page at Minster, the registers are in good condition and fully legible and there are transcripts covering the bulk of the period against which registers can be checked and corrected if applicable. Using such tests, demographic historians have tended to assume that registration was good and that human error accounts for barely 2 per cent of entries.[fn18] Yet there are other checks which can be applied such as instances of families having two or more children of the same name with no burial for the earlier one being recorded. It was, of course, possible for a family to have two children of the same name alive simultaneously if they so choose and there are some famous examples in history of that happening,[fn19] but it was surely very rare and not a single will has been found for any of those under review where they have referenced two living children of the same name. If a second child is given the same name without any burial being recorded of the former, it seems more likely that the register is wrong than the parents were simply unimaginative in their choices. Similarly, a check can be made on marriage registers to see how [pg148]many men and women remarry without their previous spouse having apparently been buried. Changes of status can also be spotted sometimes in rate assessments, poor law payments and burials. Another possible check is of people whose wills have been made and proved whose residence is given as a particular parish but for whom there is no corresponding burial. Using wills too, it is possible to check for the existence of wives for whom no marriage entry appears to exist or additional children who show no sign of being baptised.
In all such cases, there are possible explanations for the omission. A person might have died away from home. Thanet is a seafaring area and a number of the men will have died at sea and hence there was no body to bury. A person might have been baptised with one name but used another all their life. There are also the more fanciful suggestions such as the person being murdered or committing suicide or that men were selling their wives in order to obtain what they hoped would be a better replacement.
Yet it is simply not credible that every omission could fall into these categories. In this study, 126 children were found who almost certainly died before another child of the same name was born, 48 people had their wills proved without apparently being buried and 42 men and 47 women remarried without any sign of their previous spouse having died. Of these nineteen of the children probably died in periods or parishes where there was a gap in registration and seven of those whose wills were proved were mariners.
The only conclusion which can be drawn is that registers are not totally accurate and this is not an issue for the Isle of Thanet alone but everywhere. A total of 1517 people married away from Thanet, mostly in East Kent, and do not appear to have been buried anywhere in Kent. They surely did not all cross the county border and die elsewhere and they are undoubtedly no longer alive. There are literally hundreds of people in every parish who were clearly born but whose fate is totally unknown. Equally, there are people listed in every burial register whose birth cannot be traced.
So, what reasons do exist to explain the situation? Human error is probably the most likely. Registers were kept in triple locked chests which required the presence of the vicar and both churchwardens to open, which meant that they were usually written up at the end of the week when all three men were together. In the event of sickness or bad weather, the opportunity for meeting would be less and lead to longer intervals between completion. Memories could be hazy. You can imagine the vicar and wardens sitting round with their piece of paper having discussions about what name had been given to a particular child or what the Christian name was of the woman they had always simply known as Mistress Barber. In most cases, the people involved would have done their best but in busy periods, it may have been difficult. Very few people were literate, and sometimes the wardens themselves could not write. If the vicar was ill or bad weather prevented the meeting taking place, it is easy to see how mistakes would occur. None of those involved would have suspected their work would be subject to detailed examination over four hundred years later.
Another possibility is that the clergy simply did not know the details. Although the law required everyone to attend their parish church, contemporaries regularly reported that this was not the case although prosecutions were rare.[fn20] [pg149]
It should also be remembered that prior to 1598, records were not required to be kept in bound registers. A sheaf of scraps of paper could easily be blown over, knocked down, nibbled by mice with odd pages picked up and put in pockets. Many registers, such as that of St Laurence, were only created at the end of Elizabeth’s reign with entries being copied in from loose papers. Entries could have been missed or misread.
A final possible reason for omissions could relate to cost. There was no nationally required charge for entering details in a parish register at this time but an Elizabethan parish in long Melford, Suffolk noted that when the deceased was a pauper, a charge of 2d was made to the overseers to record the burial. If the overseers did not pay, no entry was made.[fn21] The same system could have applied in Thanet and elsewhere in Kent. Although baptisms themselves were free because they were a sacrament, marriages and burials were not. The eighth rubric after Holy Communion in the Book of Common Prayer required all parishioners to ensure their accounts are settled with the minister at or by Easter each year so could it be that entries were not recorded if money for any reason was still outstanding?
Nonetheless, it is important to note that almost all the recognised omissions relate to the period before 1598 and that they represent barely one per cent of entries. Record keeping improved once bound registers were in use. The records are not perfect but, whilst the problems may be frustrating for genealogists trying to trace a particular family, they are perfectly adequate for demographic analysis and they enable a window to be opened on the lives of our ancestors.
Notes
[fn]1|Leviathan, i. xiii. 9.[/fn]
[fn]2|Cowper, M. (ed.) Canterbury Marriage licences vol.1 1568-1618, vol. 2 1619-1660 (1892), Canterbury Probate Records Database at https://wills.canterbury-cathedral.org/. Scanned copies of almost every parish register in the county are available at www.findmypast.co.uk.[/fn]
[fn]3|Galley, C, The Demography of Early Modern Towns (Liverpool, 1998), 47.[/fn]
[fn]4|The multipier is 32.5.[/fn]
[fn]5|For further details see Rowland, Donald, Demographic Methods and Concepts (Oxford, 2003); United Nations Manual X: Indirect techniques for Demographic estimation (New York, 1983); Bacaër, N,, A Short History of Mathematical Population Dynamics (2011); Pollard, A H et al ‘Estimating Demographic measures from Incomplete data’ in Demographix Techniques (Sydney 1990). For the accuracy of this method compared to back projection and simple reconstitution see Alter, George, ‘The Evolution of Models in Historical demography’, European Journal of Interdisciplinary History, 2020 Winter; 50(3): 325–362. In simple terms, the deaths at each age are entered by gender (ndx) The number of deaths is then subtracted from a base figure of 100 to generate the number of survivors (lx). NLx is calculated by dividing (lx+1) + lx by two and the total population represents the sum of L.[/fn]
[fn]6|For example, 20.3 percent of adult men died in their forties of those whose deaths are known. There is no reason to assume that those who moved away died in a radically different pattern of ages so 20.3 per cent of the unknown 579 men have been added (177) to the original 346.[/fn]
[fn]7|Coletta JM, Bell SJ, Roman AS, ‘Omega-3 Fatty acids and pregnancy’, Reviews of Obstetrics and Gynaecology, vol.3, no.4 (2010) pp. 163-71.[/fn]
[fn]8|Amongst infant burials 1600-39, 60.2 per cent occurred between October and March. Amongst adult burials in the same period, the seasonal difference was less with 54 per cent occurring October to March and 46 per cent April to September.[/fn]
[fn]9|The second busiest time for conceptions was January to early March.[/fn][pg150]
[fn]10|The two years was advocated by Avicenna and widely quoted in the period, for example Sharp, Jane The Midwives Book (1671) p. 367.[/fn]
[fn]11|Booty, John E (ed.), The Book of Common Prayer, 1559 (Washington, 1976) p. 277.[/fn]
[fn]12|Cressey, David, Birth, Marriage & Death: Ritual, Religion, and the Life Cycle in Tudor and Stuart England (Oxford, OUP, 1997), pp. 114-117.[/fn]
[fn]13|It is impossible to know how many conceptions ended in miscarriages or stillbirths as registers. The number of couples without any living children also suggests a significant level of malnutrition amongst women, see Frisch, Rose, ‘Population, Food Intake and Fertility’ in Science vol. 199 (1978) pp. 22-30.[/fn]
[fn]14|Graham, Anne, ‘The Medicalisation of Menopause in Early Modern English Medical and Popular literature’, Postgraduate Journal of Medical Humanities vol 6 (2020) pp. 58-79; Amundsen, Darrel W., Diers, Carol Jean, ‘The Age of Menopause in Medieval Europe’, Human Biology, vol. 45, No. 4 (December 1973), pp. 605-612.[/fn]
[fn]15|Bolton, M., ‘The Experience of Plague in East Kent 1636-1638’, Local Population Studies, 96, pp. 9-27.[/fn]
[fn]16|Mary Dobson noted the prevalence of typhus in the south-west in 1690 and suspected it may have been found further east, Dobson, M.J., ‘A Chronology of Epidemic disease and Mortality in Southern England 1601-1800’, Historical Geography Research Series, no.19. November 1987, 66.[/fn]
[fn]17|PRC/32/42/226a.[/fn]
[fn]18|Galley, Demography, 63. He suggests the only omissions are likely to be children who died before baptism.[/fn]
[fn]19|For example, the second Duke of Norfolk had two sons named Thomas and two daughters named Elizabeth, one each by his first wife and the other by his second.[/fn]
[fn]20|Collinson, Patrick, The Religion of Protestants (1982), pp. 203-209, 215. Some would have been excommunicated and unable to attend.[/fn]
[fn]21|Boothman, Lyn, On the Accuracy of a Late Sixteenth Century Parish Register, Local Population Studies, 49 (1992), pp. 64-65.[/fn]
[fn]22|Numbers from Monkton and St Nicholas-at-Wade are too small to be representative.[/fn][pg151]