This essay describes the aims of our research project and its basic methodology, and discusses its sources.

As is well-known, there is much that still remains to be known about Sakyamuni's life and how the Saṃgha was formed during his lifetime. The sutras that give biographical information about the Buddha record only a limited amount about his life, while research into the organization of the Saṃgha and how it was managed has hardly progressed at all.

Nevertheless, this does not mean we are bereft of materials that can throw light on these subjects, since there is an enormous quantity of early Buddhist scriptures remaining, which undeniably record deeds performed by Sakyamuni and events concerning the Saṃgha. However, though the scriptures record details such as where and to whom they were spoken, and what their subject was, they do not give specifics about "when", simply saying "on one occasion". As a result, we are hampered when trying to arrange these materials chronologically in order to compile a life of the Buddha and a formative history of the Saṃgha. If we try to put together the facts of the Buddha's life by doing no more than editing those items whose chronology is clear, such as the events immediately after his enlightenment and those surrounding his death, we will be left with a large swathe of the Buddha's activities disregarded. The same reservations can be applied to the Saṃgha. Even with the Vinaya Pitaka as the indispensable documentary source for information about the Saṃgha, we cannot be sure of understanding this information correctly. As a result, a great amount of research still needs to be carried out on the Saṃgha itself and how it came into being.

This does not mean, however, that these scriptural sources do not provide promising leads. For example, in the Sâmaññaphala-sutta, Ajātasattu appears as the "King" of the kingdom of Magadha, but in many other suttas he is called "Prince". In fact, materials describing the expulsion of Devadattu actually include the scene of Ajātasattu's ascension from prince to king. Thus we can confidently say that the events of the Sâmaññaphala-sutta come after Devadattu's expulsion, and those in suttas that title him "Prince" come before. Similarly, when the Jetavana vihāra was established by Anāthapiṇḍada, King Pasenajit of Kosala had not yet become a follower of the Buddha. Thus we can say that suttas set at Jetavana and those in which King Pasenajit appears as a strong Buddhist devotee must date from after this event. Again, precept-giving rituals transformed, from ordination by invitation and ordination through the Triple refuge to ordination in the presence of ten monks (five in some localities). The ordination of women was also allowed. These changes indicate the development of the various regulations governing the Saṃgha as well as describing its formative history.
Thus, though individual suttas among the vast mass of early Buddhist scriptures do not tell us directly exactly when they are set, they contain accounts that, it is hoped, can be further clarified as research progresses, and certain information that may at least be placed in chronological order. We can liken it to a situation where we are trying to piece together a diary whose dates have been lost, or to piece together evidence like a detective in a crime novel. The information in the sutras and Vinaya has to be carefully gathered together, analyzed, organized, studied, and arranged in a time-series in order to create a chronological index of the early scriptural sources. We can then attempt to write the biography of the Buddha and conduct research into the formative history of the Saṃgha.

Nevertheless, there is the possibility that among the information recorded in the early scriptures there have slipped in traditions that took shape in later times and events belonging to the points of time when those scriptures were written down. Consequently an event recorded in the early scriptures is not necessarily historical fact. In order to establish what actually is historical fact, both in the Buddha's life and in the formative history of the Saṃgha, we must obtain supporting evidence. This, though, is now a virtually impossible task, and means that inevitably there is another aspect to our research. So while our immediate concern is to reconstruct what kind of image the compilers of the early scriptures presented concerning the Buddha's life and the formative process of the Saṃgha, we must also look at whether or not these images correspond to historical fact. This is why our research uses only Buddhist works and does not, as a matter of principle, employ literature from either Jainism or Hinduism.

We define "early Buddhist scriptures" as the five Pali Nikāyas and the Pali Vinaya, together with their Chinese translations, the four Āgamas and Vinayas such as the Sifen lü (Vinaya in Four Divisions). We exclude the Vimānavatthu, Petavatthu, Apadāna, Buddhavaṃsa and Cariyāpiṭaka from the Khuddakanikāya, which are clearly of more recent composition with many sections coloured by legend, as well as the Genben shuoyiqie youbu lü, the Vinaya of the Mūlasarvāstivādin School. Though even the older of the early scriptures have newer sections, according to common understanding, we do not make a distinction here and treat them all the same. In particular, we place weight on passages that appear both in the Pali works and in the Chinese translations and treat them as Level One sources, since we can assume that they represent the common denominator in the contemporary image of the Buddha's life and the formative history of the Saṃgha. In other words, our treatment will not change according to text but according to the quality of the information the text contains.

Level Two texts are those in which the information given in the Pali scriptures is not the same as in the Chinese translations. The latter represent the transmissions of a number of different Buddhist sects, while the former belong to just one, the Theravadins. Thus it can be considered that the Pali texts present a single image and contain comparatively few contradictions. Level Three texts consist of information transmitted only in Chinese translation.

We do not of course exclude works like the Vimānavatthu and the Mūlasarvāstivādin Vinaya as materials for our research, but treat them as Level Four texts, together with works like the Buddha biographies and commentaries on the Pitaka such as the Pali-language Theravadin aṭṭhakathā. Naturally, since they were written using the early Pali and Chinese scriptures as sources, they contain much of the same information we find in Levels One and Two. Thus if their information is the same as that in Levels One and Two, they can be considered to be of an identical level. On the other hand, descriptions that appear there and nowhere else must be assigned to Level Four.

The relationship between the levels is shown in the following diagram.

A basic feature of our research is to make the vast amount of information in the early scriptures available as computer data. Five researchers and more than ten assistants have spent seven years to reach the present stage when at last research papers can be written. In fact our research has been made possible simply because the computer has been such a close ally, allowing us to extract information at will in an instant from the vast amount of data, and to analyze and sort it in different ways.
We hope to succeed in our aims based on the methodology outlined above and on our source material.