(Part 1)

I’ll focus on the later Edo Period part of the question. It’s a very tricky one, and may not have an entirely satisfactory answer. But I hope my very long exploration of the subject will at least provide you with some information.

First of all, “Illegitimacy” as a concept is not really a good fit with Edo Period society. There’s no legal category of “legitimate” or “illegitimate” until the Meiji period. There was a distinction between children of a man’s official wife,and the acknowledged children of a concubine, with the previous category receiving precedence in inheritance. But children born outside of an official marriage/concubinage relationship, although they existed, weren’t marked as “illegitimate”, in contrast to European records. There’s no direct way to track numbers for these children prior to the Meiji period.

As you say, there’s evidence for plenty of pre-marital sex going on in Edo Period Japan among the peasantry. Anne Walthall sums up this evidence in her chapter: “The Life Cycle of Farm Women in Tokugawa Japan” from Recreating Japanese Women, 1600-1945. (p. 51)

Several historians believe that premarital sex among village girls and boys was common. In Louis Frederic's words, "Boys and girls could have amorous intrigues as it suited them and these did not necessarily lead to marriage."

This statement is footnoted as follows:

Louis Frederic, Daily Life in Japan at the Time of the Samurai, 11851603 (New York: Praeger, 1972), 58. Richard Varner, "The Organized Peasant: The Wakamonogumi in the Edo Period," Monumenta Nipponica 32, no. 4 (1978): 480, asserts that a man had to have many short term sexual encounters before he could think of marriage. Thomas Elsa Jones found that in Namase in Ibaraki prefecture not more than 2 percent of the unmarried young women were virgins, and all married women had had sexual relations before marriage; quoted in Robert J. Smith, "Making Village Women into 'Good Wives and Wise Mothers' in Prewar Japan," Journal of Family History 8, no. 3 (1983): 74. In the villages of Higashi Matsuyama in Saitama prefecture, the sons and daughters of ordinary peasants were comparatively free in their sexual experiences, and for them marriage had the meaning of settling down; see Shishi Hensanka, Higashi Matsuyamashi no rekishi, vol. 2 (Saitamaken, Higashi Matsuyamashi: Higashi Matsuyamashi, 1985), 366

Walthall continues:

Meiji ethnographer Yanagita Kunio further argued that the young men's associations of each village, and not parents, actually controlled the selection of marriage partners by means of night visits. The boys would spend evenings with members of the girls' groups, working, singing, and chatting before pairing off and sleeping together. They would then tell their parents whom they had chosen as a marriage partner.

Evidence from other areas, however, suggests that sexual relations were permitted only after the couple became engaged. In the mountains of Shikoku, young men visited girls in villages as far distant as twenty to thirty kilometers, where they helped in paper making and tried to ingratiate themselves with the girls' families, but they were watched so closely that they seldom managed to become intimate. In other areas, both boys and girls were criticized only if they had sex with more than one partner.

Walthall’s summary is a confirmation of widespread premarital sexual activity among the Japanese peasantry, but it’s also a caution against assumptions of universality. Different regions of Japan had different cultural expectations. That widespread sexual activity could take place within the frame of courtship/engagement, under the survey of parents. In a case where two people are taking steps towards official marriage, pregnancy isn’t going to be the same problem as for a completely unofficial relationship.

What did happen to children of the peasantry conceived outside an official relationship? Despite all this evidence of premarital sex, early 20th century ethnographers in Japan found that people were quite certain that pre-Meiji, there hadn’t been so many of these children. Why? One visitor to Japan in the 1930s, Ella Embree, couldn’t believe this.

Bastards were not uncommon in Suye-mura in the 1930s, but one interpretation of a passage from Ella Embree's journal is that in previous generations fewer illegitimate infants were allowed to live. After describing premarital sexual activity as virtually universal when she was young, a middle-aged informant noted that "in her day ... few babies resulted from such encounters." to which Embree adds "I wonder how correct this information is. considering the number of local bastards in Suye today)" (Smith and Wiswell 1982:130). The Embrees appear to have been unaware that infanticide was commonplace in the nineteenth century. (p. 265, “Conjugal Power in Tokugawa Japan” by G William Skinner in “Sex and Gender Hierarchies” edited Barbara D. Miller).

There’s a historical consensus that abortion and infanticide were commonplace in the Edo Period, although as you’d expect, this varies by place and time. In the article When Abortion Became a Crime: Abortion, Infanticide, and the Law in Early Meiji Japan, Susan L. Burns makes an explicit connection between these practices and illicit pregnancies.

In the 1930s, a group of Japan folklorists endeavored to compile a record of customs related to reproduction in localities throughout Japan. Later published as Nihon san’iku shuzoku shiryo shusei (Collected Materials on Japanese Birth Customs) , their work reveals that local informants believed that a fundamental shift in attitudes towards abortion and infanticide had taken place in the Meiji period. For example, a report from Chiba-ken states that "before Meiji, abortion and mabiki (i.e., infanticide) occurred openly and without shame. ... In fact, people felt ashamed instead if they were poor and still had many children."" Describing another common motive for abortion, an informant in Niigata related that "in the case when the child did not have a Father because (the pregnancy resulted from] an illicit relationship, it would be aborted." Similarly, another informant stated that "in this region, before Meiji, people called an illegitimate child 'a child of immorality'" and thus unmarried women typically sought an abortion." What these fragmented pieces of information suggest is that after 1868, what had been commonplace, indeed socially expedient, acts began to be transformed into objects of sanction and shame.

Burns has put a copy of her article on her webpage: http://home.uchicago.edu/~slburns/page1/page1.html