SUMMARY
The Osteological Paradox. An Identification Problem By Hans Christian PetersenIn studies of human skeletal material from medieval cemeteries one of the main objects of study is the health and wellbeing of the population represented by the skeletons. Often comparisons between social, regional or chronological groups are attempted. As studies of this kind try to make inference about the once living people by looking at their skeletal remains, the possibility of reaching wrong conclusions is present. One obvious problem is the possibility that the most healthy will live long enough to develop skeletal lesions specific to a certain disease investigated, whereas the more frail ones die before they develop these lesions. A straightforward – and wrong – conclusion would be that the healthy population is more affected by the specific disease than the less healthy population. However, other more intriguing aspects of the osteological paradox exist, basically leading to the risk of obtaining wrong estimates of the parameters describing the health status of the population(s) in question. With wrong parameter estimates one will also get wrong comparisons of, say, frequencies of individuals affected by different diseases. The parameters are not identifiable. The major problem is the impact of selective mortality, i.e., the fact that the more frail individuals are more prone to die at a certain stage of a developing disease than are the less frail individuals. This means that it is not always possible to identify the right explanation among several possible explanations for an observed pattern of skeletal lesions. Examples are given and suggestions for solving at least some of the problems are presented.