A Preliminary Analysis of the Tathāgatagarbhasūtra as Found in the Lang Kanjur
Common throughout the De bźin bśegs pa’i sñiṅ po’i mdo (Tathāgatagarbha sūtra) of the Lang Kanjur are several features which are generally assumed to be archaic, such as the ya btags in all words beginning with m- followed by the vowel i or e (e.g. myi, myed, etc.), the usage of the da drag, the tsheg placed before śad, the mtha’ rten ’a (e.g. dpe’ ), occasionally a reversed gi gu, la(s) (b)stsogs pa for la sogs pa, the omission of genitive particles and, in the verses, the reading ’i instead of yi ( ’i counting as a full syllable).
The version of the sūtra represents the canonical transmission (and not the translation found in the “Kanjur from Bathang”). Stemmatically, the text in the Lang Kanjur is very close to the three Phug brag versions of the sūtra, which have been shown to derive from one and the same archetype. It shares mistakes with this archetype. In other instances it is, however, free of the secondary readings found in all three of the Phug brag versions. In all the cases where Phug brag shares a mistake with the representatives of the Tshal pa-line, the Kanjur version from Dolpo also has this secondary reading. Its use for establishing the stemma of the canonical versions of the De bźin gśegs pa’i sñiṅ po’i mdo is therefore restricted primarily to evaluating the readings of the Phug brag Kanjur in the instances where Phug brag deviates from the Tshal pa-transmission. In all the cases where the Chinese translations of the sūtra could be used to decide on the originality of a reading in the Tibetan, it turned out that whenever the variant in the Lang Kanjur was identical with the one of Tshal pa as against Phug brag, the latter variant was secondary. (Zimmermann, appendix, 104–5)
Notes85. For more details on this paracanonical translation see Zimmermann 1998.
86. See Zimmermann 2002: 173–177.