Abstract
The purpose of this paper is to discuss the levels of reproduction of daimyo retainers in the last half of Tokugawa period in Japan. Sekiyama (1958) assumes that the samurai class did not reproduce itself. However, this claim has not yet been confirmed because we cannot directly calculate the reproduction rate from genealogies, which are the primary source for the analysis of samurai population. This study approaches this issue through the method of micro-simulation. First I conducted the simulation by positing a stable population at a simple reproduction level (NRR≒1), using a micro-simulation program, called CAMSIM, which had been developed in the Cambridge Group for the History of Population and Social Structure. Most values of parameters needed for the simulation were calculated from the personal histories of daimyo retainers of Uwajima han, which had been submitted to Uwajima han office. Especially, the age-specific mortality rates listed in the life table of west 8 and 9 and north 8 of Coale and Demeny (1983) were selected as the values of the parameters in consideration of infant mortality rate that could be only calculated from the personal histories. The NRR derived from the simulation was 1.02 with west level 8, 1.09 with west 9, and 1.00 with north 9. In this way, the population of more or less simple reproduction level could be simulated. Next I calculated the mean number of sons per married 'family head'. In the simulated population in question, I defined the 'family head' as the surviving eldest son of 'ego' at his death. The number was 2.14 with west 8, 2.11 with west 9, and 2.21 with north 8. The number can be said to be within the range of 2.11-2.21. I also calculated the mean number of sons surviving more than 15 years per 'family head' from the results of the simulation. The number was 1.20 with west 8, 1.31 with west 9, and 1.24 with north 9. The number can be said to be within the range 1.20-1.31. Third I compared them with the equivalents from genealogies of daimyo retainers in eight hans, which are presented in Murakoshi (1991, 1993, 2001). The mean number of sons per married family head of retainers of Uwajima han was 2.13. This number was more or less the same as the number from the result of the simulation (2.11-2.21). Many values of parameters were calculated from the personal histories of retainers of Uwajima han. Therefore I claimed that the result of the simulation should be acceptable. The mean numbers of adult (surviving more than 15 years) sons per family head of retainers of seven hans-Tokushima han, Aizu han, Akita han, Morioka han, Kaga han, Hagi han and Saga han-have been already estimated. The number in Tokushima han was within the range of 1.22-1.26 from the last half of the 18^<th> century to the first half of the 19^<th> century. The number in Aizu han was 1.24 in the last half of the 18^<th> century. The number in Akita han was 1.22 in the last half of the 18^<th> century the number in Morioka han was 1.25 from the first half of the 18^<th> century to the last half of the 18^<th> century. The number in Kaga han was 1.19 in the last half of the 18^<th> century. The number in Hagi han was 1.20 in the first half of the 18^<th> century. The number in Saga han was within the range of 1.21-1.26 from the last half of the 18^<th> century to the first half of the 19^<th> century. In brief, the mean numbers of adult sons in seven hans were within the range of 1.19-1.26. The numbers calculated from the result of the simulation were within the range of 1.20-1.31. The two ranges were found more or less equal. Therefore I concluded that daimyo retainers must have reproduced themselves, rejecting Sekiyama's claim.
