Introduction

This article examines the story of the Buddha's first return to his home city Kapilavastu after his awakening, as found in the fifth chapter of the Buddhist work known as the Book of Zambasta. This is possibly the oldest extant Buddhist text in Khotanese, an Eastern Middle Iranian language once spoken and written in Central Asia, in part of what is now the north-western Chinese region of Xinjiang.

My purpose here is to investigate the fifth chapter in the context of the reception of the Mahāyāna in Khotan. In particular, I explore how an appreciation of the mode of transmission of the text, through the analysis of its structure and contents and in comparison with a number of possible sources and parallels, can contribute to the reconstruction of the first documentable stages in the spread of the Buddhadharma to Khotan.

Chapter five is found as such in the only virtually complete manuscript of the Book of Zambasta, stemming from the eighth century. It was not included in ms. T III S 16, the earliest witness of this work, datable on palaeographic grounds to the fifth/sixth centuries, since the single surviving folio of this earliest witness, which contains stanzas belonging to chapter nine, enables us to calculate that the manuscript only contained part of the collection starting from chapter seven, that is, counting backwards, the first folio of the manuscript must have contained the beginning of chapter seven. Theoretically, this earliest witness could have been either a selection of the work for independent circulation or an initial nucleus which was subsequently expanded. Thus, in theory, there are two possibilities: either chapter five was part of the Book of Zambasta from the very beginning or, if the work developed as a collection through a number of unattested intermediate stages of formation (or if it were to become a collection later, as the result of a single redactional event), it became part of such collection after the fifth century, in which case, as a consequence, the chronology of its redaction would remain uncertain. However, the possibility that the work did not exist from the outset in its present shape is less likely, given the character of the Book of Zambasta as a collection of individual texts distinctively designed and intercommunicating as one unitary work in terms of both content and structure.