The historical demography of Japan to 1700mostly of commoners was the law of the land, most experts agree that the largest number of
troops that could have been raised from a population deemed to range from 5.8 to 6.4 million
was about 110,000. From a reliable source for the mid- fifteenth century, I obtained a range for
the number of soldiers that could have been outfitted around 1450. Scholars estimate that one
warlord (shugo daimyō) on average fielded about 2,825 soldiers, both mounted and on foot. There
were about thirty- seven of these lords, suggesting that all together at the least they could send
104,525 fighters into battle. Some daimyō administered more than one province, and so it can be
argued that the figure of 2,825 fighters might be best applied to the total number of provinces
(sixty or so). The Toki, for instance, ruled only one province (Mino), but provided more than
3,000 troops in 1433. If one multiplies 2,825 by sixty, then the highest total equals 169,500. The
range for troops locally raised for battle is thus 104,525–169,500.
In addition, the central government of the time, known as the Muromachi shogunate, pos-
sessed its own army approximately equivalent to those of ten warlords’ armies (28,250). The
range for the grand total of all fighters that could have been committed to fight in 1450 thus runs
from 132,775 to 197,750. Working by analogy from the figures for fighters and total population
for the early 700s, I calculated a multiplier of 1.21. Applying this multiplier willy- nilly to the
two Muromachi troop numbers, the population comes out to between 7.4 and 11 million, or an
average of 9.2 million. Adding an acceptable total for urbanites, I arrived at an estimate for
Japan’s total population around 1450 of 9.6 million.
Given the many assumptions behind this number, one might easily entertain doubts. Was the
ratio of maximal military forces the same in 730 as in 1450? Did samurai warlords take all their
soldiers with them when they went fighting, or did they leave an unknown number at home?
How well defined was the status of samurai around 1450? Therefore, the estimate of 9.6 million
is not much better than a reasonable guess. Fortunately, Saitō Osamu has inferred a population of
10.5 million for the same date by working backwards from reliable figures of the Tokugawa
shogunate.^39 Most evidence warrants the generalization that the years from about 1360 to 1450
comprised a “Muromachi heyday” during which the Japanese population resumed a fairly strong
growth trajectory.
If the figures for mid- Muromachi Japan seem like guesswork and invite doubt and contro-
versy, the same can also be said for the Warring States’ era (1467–1600), during which many resi-
dents of the archipelago were almost continuously at war. Disease came back with a vengeance,
as a new potentially deadly pathogen (syphilis) arrived with the Europeans in 1512. Famine
trended much worse: four death registers (kakochō) recording the name and date of the deceased
suggest that deaths began to spike between 1430 and 1460, and climbed again during 1490–1520.
Finally, warfare became more and more deadly, both for combatants in ever- larger, better armed
hosts and for innocent bystanders caught up amongst the hungry armies as they mowed down
and stole crops, burned villages, and kidnapped commoners to resupply their forces.^40 Taken
together, these extraordinary mortality factors suggest that the late fifteenth and sixteenth cen-
turies were a time of little or no growth.
The death registers merit special attention because they reinforce the concept of the “spring
hungers,” a time in the late spring and summer when the grain from the previous year had been
exhausted and before the new harvest in the fall. Consider these figures from the Hondoji death
register including month- by-month mortality numbers from 1395 to 1600 for Shimōsa province
in the Kanto.
As Table 16.2 shows, the populace of this Kanto province was more likely to die during the
first half of the year, particularly during the first, third, fourth, and sixth months of any given
year. Once the harvest had come in during the autumn, however, one’s chances of survival
improved markedly. Other surviving death registers exhibit a similar trend. In the event, it seems