________________
32
Journal of Indological Studies, No. 18 (2006)
dismissed by Vacaspati who appeals to Uddyotakara's explicit mention of the sutra in the singular as invalidating evidence.30 It should be noted that in this digression both the opponents and Vacaspati appear to regard 1.1.2 as ending with apavargah, the reading which represents the generally accepted text.31 If this reconstruction is accepted, it has to be assumed that Vacaspati actually commented upon a version of 1.1.2 different from that of Jayanta Bhaṭṭa and other commentators. 32
There are some places where Vatsyayana refers to 1.1.2. His mention of the sutra in his commentary on 4.2.1 can be adduced as internal, problematic evidence which speaks against the T version of 1.1.2. In the context of discussing the faults (dosa), which are the three elements causing karma
not available in Thakur's edition according to ST 69,20, since "one complete folio escaped the camera" (cf. Thakur's Preface to the ST). Furthermore, it can suggest that Śrīkantha regarded 1.1.2 as ending with bhavad, but this assumption evidently contradicts the discussion adduced by Vacaspati which Śrīkantha supposedly comments upon. In fact, Śrīkantha's mention of these two yoga-s is placed in the part of his commentary on Udayana's corresponding discussion (cf. LDI(1) f. 42r 6: "une" abbreviating "Udayane"). It is totally unclear how he could keep silent about the possibility of the charge of contradicting the NVTT.
30 Cf. NVTT 63,7-8= EM 157,20-21: tam imam sutravibhāgam amṛṣṇamāno vārttikakṛd aha idam sūtram. ekavacanena bhedam vyavartayati. Vacaspati further adduces as the argument against the theory of yogavibhaga the fault of the "splitting of a statement" in 1.1.2 (cf. vakyabheda). Cf. NVTT 63,8-9 EM 157,21-22: na hi samucchedakramapratipadanenäpavargaparatayaikavakyatve sambhavati vakyabhedo nyayyaḥ. ("The splitting of a statement, indeed, cannot be reasonable, because [sutra 1.1.2] constitutes a single statement inasmuch as it is devoted to [the explanation of] liberation by means of demonstrating the order of the complete destruction [of pain and the others].") In Vacaspati's view, the sutra constitutes a single statement (ekaväkyatva), insofar as 1.1.2 has "liberation" as its main objective to be presented (apavargaparataya). Cf. NVTT 63,8-9 EM 157,21-22; NVTP 108,7-9 EM 173,22-24. For ekavakyata, see Preisendanz 1994: 204-207; Kane 1962: 1297-1298. For vakyabheda, cf. Kane 1962: 1299-1303.
31 For another problematic instance, cf. SDS(BI) 115,8-10: kintu tattvajñānād duḥkha ... bhava iti, which seems to silently quote the corresponding portion of the NBh. There is also an explicit reference to 1.1.2 in the same compendium ascribed to Madhava; cf. SDS(BI) 116,9-11: tatha ca paramarṣam sutram duḥkhajanma... bhavad apavarga iti (ellipsis by me). Abhyankar's edition, however, provides a different reading in both places, namely, the one ending with payad apavargaḥ. Cf. SDS 245,7-9 and 246,16-18.
32NBhüş 72,15-17 (= HJJM(1) f. 16r 7-8), which runs parallel to the introductory part of the NBh on 1.1.2, does not support niḥśreyasadhigamaḥ of the T version: tat khalu vai tattvajñānam kim ātmalābhānantaram eva niḥsreyasam sampadayatiti. ucyate
na, kim tarhi tattvajñānād duḥkhajanmapravṛttidoṣamithyajnananam uttarottarāpāye tadanantarābhāvāds apavargo bhavatiti vakyaśeṣaḥ. (°bhāvād HJJM(1); °pāyād NBhüş, where the editor seems to have corrected the text of 1.1.2 to the common one.) With regard to the question whether Bhāsarvajña regarded 1.1.2 as ending with bhävät or with apavargaḥ, it depends on the interpretation of "the rest of the sentence" (vakyaseṣaḥ). Yogindrānanda, the editor, places a danda after apavargaḥ, which indicates that he regards bhavati as the rest of the sentence. However it is also possible to assume that the words apavargo bhavati are intended as that which is to be supplied. The latter assumption entails that Bhāsarvajña regarded 1.1.2 as ending with bhāvāt. On Bhāsarvajña's discussions on NS 1.1.2, cf. Yamakami 2001: 13-18.
