Send message to Author

Paper short abstract

There survive today four textual lineages into which extant texts of the Pillow Book are categorized, but the relationships among these have proven difficult to establish. This paper argues that we should rather accept that from the beginning there have always existed multiple different texts.

Paper long abstract

There exists today general scholarly agreement that all texts of the Pillow Book known to yet survive can be categorized among four discrete textual lineages: the Teika-bon line (or “Sankan-bon”), the Den-Nōin-bon line, the Sakai-bon line, and the Maeda-ke-bon (a unique exemplar as yet without collinear cognates). Of these four, the Teika-bon has traditionally been regarded as offering the superior textual witness, and indeed most modern editions adopt it as their base text, with the predictable consequence that even most research today on the Pillow Book effectively treats the Teika-bon as default.

Particularly neglected has been the single manuscript of the Maeda-ke-bon. Despite being the oldest text of the Pillow Book, and indeed the only one surviving from the Kamakura period, for many years there existed a scholarly consensus dismissing it as a mere composite drawing from the Den-Nōin-bon and Sakai-bon lines, leading the Maeda-ke-bon’s witness to go largely ignored. With such consensus having since been disproven, however, the time is ripe for reevaluation of its value. And indeed, the time for a more general reevaluation of the Pillow Book’s texts is long overdue. The vaunted Teika-bon line, deriving ultimately from a copy in Teika’s own hand, represents after all but a single authority’s recension, and one itself dating no earlier than the Kamakura period (Antei 2/1228). For research on the Pillow Book to nonetheless continue its reliance on the Teika-bon alone can hardly be termed ideal.

At the same time, merely exchanging Teika’s text for another, perhaps for the Maeda-ke-bon, is no more ideal. All such solutions suffer fundamentally from the defect of assuming that there ever existed a single urtext, one that judicious scholarship might someday approach. As I argue, however, the truly judicious approach is rather to simply accept that, in the case of the Pillow Book, from the time of composition there have always existed multiple different texts. Proceeding from such an argument, this paper considers solutions to several connected textual problems, in particular the various ways in which these multiple “originals” might relate to the four distinct text-types that survive today.