Textual sources
Our knowledge of chan during the Tang period, with the significant exception of a number of manuscripts found at Dunhuang dating back to the early 9th century, comes mainly from the numerous histories of the chan sects which appeared in sequence from the mid 10th-century until about 1300, and which thus were compiled considerably later than the events which they describe. The biographies and recorded teachings of Shitou, Yaoshan and Yunyan are translated here in their entirety from Wu deng hui yuan (Five Lamps Merged in the Source), compiled by the monks Pu Ji and Hui Ming. This work, first published in 1253, is the most comprehensive of the traditional chan histories; it combines material found in five earlier "transmission of the lamp" histories (chuan deng shi) and also adds new material.
To learn more about Shitou's life and teachings, we have also included some additional stories from Zu tang ji (Collection from the Hall of Ancestors), which appeared in 952, and further stories of Shitou taken from the records of different of his disciples, also found in Wu deng hui yuan, or in Jingde chuan deng lu (Transmission of the Lamp from the Jingde Era), published in 1004. It is clear that some major differences can be seen in the Song histories, not only in terms of the material included, but also in terms of literary style, and thereby probably the intended readership. A good example appears in the Zu tang ji version of Shitou's youth, compared with the account in Wu deng hui yuan. The former, with its usual emphasis on folk Buddhist elements, presents a somewhat miraculous account of Shitou's birth, which is omitted altogether in the more scholarly Wu deng hui yuan.
How can we be sure that what we are reading is the real teaching of these chan masters of early Tang? Obviously we can’t. In future decades it is likely that Chinese scholarship, through textual comparison and linguistic analysis, will be much better able to decide issues of historic authenticity. In the meantime, it seems fair to suggest that some parts of the Song histories provide us with reasonably accurate representations of what these masters originally thought and taught. Despite the constant temptation for later editors to revise or omit, interpret or interpolate, there must have been also operative a conservative tendency to pass along the truth as true to the original as possible, even when issues of sectarian privilege or lineage were involved. Another point worth remembering is that from the standpoint of Buddhist practice, it makes little difference if the texts are genuine or not. It is certainly more significant that generations of Zen Buddhists in China, Korea and Japan have, for about a thousand years, accepted the Song records as "gospel" and have used their teachings as models for their own practice.