Abstract:
A deficiency of children in burial sites is defined as the difference between a supposed child mortality for a time period and the real number of skeletons of children found in burial sites after archaeological excavation. Such a deficiency of children has been claimed for early medieval times. A child mortality of about fifty percent has been assumed for these times, but few skeletons of children have been found in excavations of medieval graveyards. This thesis shows that there is no evidence for the claim of high child mortality in medieval times. In addition to an inconsistent definition of the age group of children, the hypothesis of a fifty percent child mortality is based on examinations in paleoanthropology. But in paleoanthropology it is supposed that the fraction of children present in burial places in medieval times should be around fifty percent. The child mortality rate is, however, not the same as the proportion of children expected in burial sites. The hypothesis of high child mortality in medieval times is therefore based on a misunderstanding of results in paleoanthropology. The assumption that children should make up fifty percent of burials is based on demographic life tables and the interpretation of a single burial site, regarded as representative. The use of life tables, however, is increasingly criticized because life tables are based on a idealized demographic model and do not mirror real life. On the basis of an alternative probabilistic model for mortality, this thesis simulates the evolution of burial site composition using the Monte Carlo method. The analysis examines how several factors influence the relative frequency of children and adults in burial site. In the simulations, the development of burial sites shows a typical pattern. There is a period in which the proportion of children and adults fluctuates strongly and then levels off to a constant ratio. This pattern is influenced by several factors. For example, fertility has a stronger influence on the ratio of adults and children than mortality. The level of life expectancy related to the life tables, however, has only a small influence on mortality. On the other hand, the composition of the starting population has a great influence on the time to equilibrium and the proportion of age groups, and the standard deviation between different simulations is extremely large. The proportion of children in burial sites is therefore influenced by many factors. These results lead to the conclusions that it is imposslible to infer anything about child mortality on the basis of the relative frequency of children in burial sites.