Temples belonging to the esoteric schools of Buddhism in medieval Japan have produced an extremely vast and elaborate religious literature, called the “sacred teachings” (shôgyô). Although such texts have received little attention from historians, the last few years have seen, especially with the research team based in the Shinpukuji temple in Nagoya, a renewed interest in them as well as the publication of many manuscripts. However, the use of such texts as historical sources presents a series of challenges. The “sacred teachings” were mostly composed in Japanese kanbun (classical Chinese read in the Japanese word order), and consist in ritual procedures, doctrinal or canonical commentaries, or records of oral traditions and various events. Their writing style is fragmentary, almost cryptic at times, and they are not designed to be comprehended by the non-initiated. So, in order to understand their contents and to use them as proper historical sources, one has to reconstruct the vast knowledge network they were built upon.
Digital humanities can provide particularly helpful tools to clarify this stream of ideas and to determine patterns in the diffusion of rituals, symbols, or doctrinal interpretations among the monks of the time. They can also help us assess the veracity of traditional claims of authorship, or even evaluate the boundaries between schools or rival branches. For example, in a book published in 2015, Ishii Kôsei was able to prove the authenticity of several texts attributed to Shôtoku Taishi, relying on computer-assisted vocabulary analyses based on text mining and especially n-gram. This presentation aims to build on his research and to devise a method and tools applicable to the study of medieval “sacred teachings.” It will first examine the issue of the authorship of two texts, the Daijingû honji and the Ben’ichisan kuketsu, and assess its implications.
According to their colophons, the Ben’ichisan kuketsu and the Daijingû honji were written around the middle of the 13th century. However they both mention a ritual called the "Ritual combining the Three Worthies (Sanzon gôgyô hô)," a practice that recent scholarship has proved to have been created by Monkan (1278-1357), a monk active mostly during the 14th century. In the medieval esoteric schools, ritual and doctrinal authenticity has always been a major issue, and monks imagined various ways to assess their superiority over rival schools and practices. One of them was for a monk to attribute his own work to an eminent figure from the past. According to the research by Abe Yasurô, Monkan did this with the text called Goyuigô hiketsu, which was attributed, most certainly by Monkan himself, to the Daigoji master Jichiun, active in the early 13th century. So it is quite probable that a similar process was at work with the Daijingû honji and Ben’ichisan kuketsu.
The main hypothesis of this presentation, based on an analysis of the texts’ contents and their historical contexts, is that they may well have been written by Monkan, or at least by a member of the same school of thought, and attributed voluntarily to previous masters in order to assert their authenticity.
I will first create a database of Monkan’s known works, including the Goyuigô hiketsu (whose authorship will surely be proved in the process). The main goal is to use data mining in order to isolate linguistic pattern and analyze their frequency, mostly via python language n-gram scripts, and then to compare such data with the Ben’ichisan kuketsu and the Daijingû honji. However, occurrences of series of Chinese characters do not always have the same relevance, and, as a preliminary step, it will be necessary to organize the data.
Concretely, I will use recent editions of manuscripts and transcribe them into digital data, following the guidelines of the Text Encoding Initiative. Since I will work on a defined number of documents, I will try to create coherent metadata categories for this type of text (categories for “paragraphs” could include canonical citations, ritual procedures, opinions of their author, etc.) and apply them to the data. This will give a clearer picture of the significance of each linguistic pattern in Monkan’s works, and help determine the topics discussed in the text, the type of references, and, most importantly in the writer’s style.
The next step will be to confront this data with the Ben’ichisan kuketsu and Daijingû honji, which will first be analyzed with the same method. The results should give us a clearer picture of the degree of similitude, or maybe the differences, between them and other works by Monkan. If possible, we will also try to identify the author of another similar text, the Shinzô zuzôkan, as recent research by Uchida Keiichi has shown that the name appearing in its colophon might well refer to Monkan. However, this text is not fully available, and we would have to work with partial data.
As a whole, this research will not only help answer the question regarding the authorship of the aforementioned texts, but also shed new light on Monkan’s thought as a whole, and especially his links to the medieval Shintô tradition, which for now remain uncharted territory. On a longer-term basis, this method will also prepare deeper investigations into the transmission process of rituals and ideas in the Middle Ages. Furthermore, if expanded to other texts from a similar period, it can clarify the actual diffusion of doctrinal arguments between rival schools, as Shingon and Tendai, or spiritual lineages inside the same school, allowing us not only to understand the position of a figure such a Monkan, who has the particularity to have been rejected as a heretic after his death, in the history of his school as well as in the history of ideas as a whole, but also to question the traditional divisions inherited from later periods and their influence on our understanding of the situation in medieval Japan.