The Jñānaprasthāna by Kātyāyanīputra has survived in two Chinese translations, one by Gautama Saṃghadeva and the other by Xuanzang. It consists of eight chapters, and the first chapter is divided into eight sections. The Sanskrit fragments from Eastern Turkestan (Xinjiang) which will be the subject of this paper stem from the last section of the first chapter. This section is named Si 思 (cetanā) and its Indic title might have been Cetanāśvāsaka. The name of the chapter is known as Za 雜 (“miscellaneous”) in Chinese. According to the Vibhāṣa, the chapter owes its name to the fact that it deals with the interpretation of miscellaneous terms appearing in sūtras. The Vibhāṣa, commentary on the Jñānaprasthāna, mentions the classification of the verses into vargas in the Udāna by Dharmatrāta as an instance to explain the classification of the Jñānaprasthāna into the eight skandhakas. If we further follow the instance within the Udāna, scil. the Prakīrṇaka-varga, Prakīrṇaka-skandhaka might be conceivable as the Sanskrit name of the chapter in concern. The Sanskrit text reconstructed on the basis of the fragments corresponds to the portion of the Cetanāśvāsaka which can be subdivided into two parts, namely māna and vitarka. The māna part (I–II) deals with the two of the seven kinds of pride (māna), to wit, abhimāna and hīnamāna. The vitarka part (III) deals with the trayaḥ akuśalavitarkāḥ, three kinds of unwholesome thoughts. The relationship between our text and the Chinese versions of the Jñānaprasthāna is not easy to define precisely. Demiéville seems to be of the opinion that the Sanskrit text is more closely related to the version translated by Gautama Saṃghadeva than to the one translated by Xuanzang, a view which cannot be corroborated in full on the basis of our revised and enlarged text. The contrary seems, at least partly, to be the case as we can observe below through the juxtaposition.
