リスク学と確率論の狭間で

抄録

Although the mathematical bases of risk are not entirely derived from the probability theory, the two have a delicate and unique relationship. At the origin of the science of risk, insurance companies flourished in the 17^th and 18^th centuries, where the calculation of insurance premium was based on probabilistic concepts, especially those developed by Pascal. Distinguished mathematicians such as Huygens, Bernoulli, and Bayes all belonged to this era. The probability theory was often used for risk analysis with relatively simple structures such as chemicals and toxicology, and this tendency is still dominant today. However, as risk targets expanded to include complex and non-linear phenomena such as environmental and social risks, the conventional, simple probability theory proved inadequate. In particular, probabilistic tools cannot be used to assess the complex mechanisms of low-level radiation risk. Moreover, huge earthquakes and nuclear power plant meltdown incidents that have a low frequency of occurrence but devastating repercussions cannot be approached with the probability theory. The appropriate mathematical model for such risks has not yet been developed.