¹ See Sāṅkhyasaptativṛtti (Esther Abraham Solomon (ed.), Sāṁkhya-Saptati-Vṛtti (V₁) (Ahmedabad, 1973), 1:8–2:4). This story is also found in the Suvarṇasaptati (Taishō Shinshū Daizōkyō 大正新脩大藏經, vol. 54, no. 2137, 1245a08–18); Sāṅkhyavṛtti (E. A. Solomon (ed.), Sāṁkhya-Vṛtti (V₂) (Ahmedabad, 1973), 1:5–19); Māṭharavṛtti (Vishnu Prasad Sharma and Sri Satkarisarma Vangiya (eds), Sāṃkhyakārikā of Śrīmad Īśvarakṛṣṇa with the Māṭharavṛtti of Māṭharācārya and the Jayamaṅgalā of Śrī Śaṅkara (Varanasi, 1970), 1:24–2:14).

² Sāṅkhyakārikā 2, ‘dṛṣṭavad ānuśravikaḥ sa hy aviśuddhikṣayātiśayayuktaḥ/ tadviparītaḥ śreyān vyaktāvyaktajñavijñānāt//’. The Sanskrit text of the Sāṅkhyakārikā is from Albrecht Wezler and Shujun Motegi (eds), Yuktidīpikā: The Most Significant Commentary on the Sāṃkhyakārikā (Stuttgart, 1998), pp. 278–285. The translation is from Larson, Gerald James, Classical Sāṃkhya: An Interpretation of its History and Meaning (Delhi, 1979), p. 256Google Scholar, with modifications.

³ For examples of these sources, see Hyoung Seok Ham, ‘Buddhist Critiques of the Veda and Vedic Sacrifice: A Study of Bhāviveka's Mīmāṃsā Chapter of the Madhyamakahṛdayakārikā and Tarkajvālā’, (unpublished PhD dissertation, University of Michigan, 2016), pp. 127–136.

⁴ Cf. the Sāṅkhyakārikā 1, ‘Because of the torment of the threefold suffering, (there arises) the desire to know the means of counteracting it. If (it is said that) this (desire—i.e., inquiry) is useless because perceptible (means of removal are available), (we say) no, since perceptible means are not final or abiding’ (duḥkhatrayābhighātāj jijñāsā tadabhighātake hetau/ dṛṣṭe sāpārthā cen naikāntātyantato’bhāvāt//). The translation is from Larson, Classical Sāṃkhya, p. 255.

⁵ The Sāṅkhyakārikā 4 lists ‘trustworthy testimony’ (āptavacana) as one of the three sources of valid knowledge and then glosses it in the next verse as ‘what is heard from trustworthy beings’ (āptaśruti). In the Sāṅkhyakārikā 6, it is declared that this means of valid knowledge covers objects that lie beyond the reach of perception and inference. For detailed discussions on the commentaries’ interpretations of these verses, see Ołena Łucyszyna [Olena Lutsyshyna], ‘Classical Sāṁkhya on the authorship of the Vedas’, Journal of Indian Philosophy 40.4 (2012), pp. 453–467 and Łucyszyna, O., ‘The scope of the Pramāṇas in classical and postclassical Sāṃkhya’, Asian Philosophy 32.1 (2022), pp. 33–51CrossRefGoogle Scholar.

⁶ Łucyszyna, ‘Classical Sāṁkhya on the authorship of the Vedas’.

⁷ The Jayamaṅgalā (circa 700 ce or later) is omitted from this grouping because it does not mention the Veda as a case of trustworthy testimony. However, I concur with Łucyszyna, when she says: ‘It is unlikely that J[aya]M[aṅgala] denies the authority of the Veda, though nowhere in this text is it said directly that the Vedas are an authoritative source of knowledge’ (Łucyszyna, Ołena, ‘Classical Sāṁkhya on the relationship between the Vedic revelation (śruti) and its own doctrine’, Studia Religiologica 50.4 (2017), p. 314Google Scholar).

⁸ For the dating of these commentaries, see Larson, Gerald James and Bhattacharya, Ram Shankar, Sāmkhya: A Dualist Tradition in Indian Philosophy. Encyclopedia of Indian Philosophies (New Jersey, 1987), pp. 15–16Google Scholar.

⁹ Frauwallner suggests ‘around 550 A.D.’ as the possible date of the Yuktidīpikā without providing any evidence (Frauwallner, Erich, History of Indian Philosophy (Delhi, 1973), Vol. 1, p. 226Google Scholar). Halbfass proposes placing the Yuktidīpikā (YD) after Kumārila (600–660 ce) since ‘there is no conclusive evidence for Frauwallner's suggestion’ and the YD on Sāṅkhyakārikā 2 ‘seems to be a response to the Ślokavārttika [of Kumārila]’ (Wilhelm Halbfass, Tradition and Reflection: Explorations in Indian Thought (Albany, 1991), p. 94). In the introduction to their critical edition of the YD, Wezler and Motegi provide the basic framework to determine the date of the text. What is certain is that the YD quotes Dignāga (480–540 ce) several times but not Dharmakīrti (600–660 ce). Despite the Yuktidīpikā's silence on Dharmakīrti, they assign ‘circa 680–720 A.D.’ as the ‘lower limit’ for the text, that is, terminus post quem (Wezler and Motegi, Yuktidīpikā, p. xxviii). This judgement is based on the fact that the YD quotes the Kāśikāvṛtti (680–700 ce) on Pāṇini's sūtra whose date is solid. Wezler and Motegi acknowledge Halbfass's observation, but Kumārila's possible presence in the text plays no role in dating the YD; however, the YD's seeming reference to Kumārila is compatible with their dating. Against the editors’ warning that ‘the quotation from the Kāśikā cannot, however, be simply done away with by assuming that it is but a later addition to the text’ (ibid.), Mejor puts forward the thesis that the YD predates the Kāśikā. Referring to the parallelism observable between the YD and the Jayamaṅgalā (another later commentary on the Sāṅkhyakārikā), Mejor suggests ‘perhaps J[aya]M[aṅgalā] has preserved the reading closer to the original reading of the YD which was only later replaced by a lucid explanation taken from the Kāśikā’ and, if so, then ‘the earlier date for the YD, i.e., circa 550 C.E., is secured’ (Marek Mejor, ‘Some observations on the date of the Yukti-dīpikā (apropos of the new edition)’, in Essays in Indian Philosophy, Religion and Literature, (eds) P. Balcerowicz and M. Mejor (Delhi, 2004), p. 414). By pointing out another quotation from the Kāśikā, Mejor even proposes the possibility that it is Kāśikā, not the YD, which quotes another's words; and this turns the date of Kāśikā into the ‘upper limit’ (terminus ante quem) of the YD (ibid., pp. 415–416). Bronkhorst also endorses the date of 550 ce but for a different reason. Having demonstrated that there is a case for the Kāśikā drawing upon the source(s) that it shares with texts older than itself, such as the Brahmasūtra, Bronkhorst concludes that the Kāśikā-quotation in the YD ‘may conceivably have been taken from an earlier commentary in the Pāṇinian tradition’ and agrees ‘that the date proposed by Frauwallner, circa 550 C.E., is, if not secured, at least possible or even probable’ (Bronkhorst, Johannes, ‘More on the sources of the Kāśikā’, in Problems in Vedic and Sanskrit Literature, (ed.) Deshpande, M. R. (Delhi, 2004), pp. 52–53Google Scholar). However, both Mejor and Bronkhorst do not seriously consider Halbfass's observation that the YD on the Sāṅkhyakārikā 2 introduces the opponent's argument, which resembles that of Kumārila (seventh century). Given diverging scholarly opinions on the date of the YD, I tentatively consider it to have been produced between the sixth and eighth centuries.

¹⁰ Łucyszyna, ‘Classical Sāṁkhya on the Authorship of the Vedas’, pp. 461–462.

¹¹ Ibid., pp. 460–461.

¹² The edition of the Yuktidīpikā used in this study is Wezler and Motegi, Yuktidīpikā. When I specify a location of the text, I will abbreviate the text as ‘YD_WM’. The YD on Sāṅkhyakārikā 2 is at YD_WM 29:16–55:1.

¹³ Halbfass, Tradition and Reflection, p. 93. The section of the Ślokavārttika to which Halbfass refers is grouped under the heading of ‘10.4 aṃśadvayasthahiṃsāsamarthanam (Justification of animal sacrifice, etc.)’ in Kataoka, Kei, Kumārila on Truth, Omniscience, and Killing (Wien, 2011), Vol. 1, pp. 52–58CrossRefGoogle Scholar and Vol. 2, pp. 481–513.

¹⁴ Kumārila lists the examples of having sexual intercourse with one's own guru's wife (gurustrīgamana) and drinking liquor (surāpāna) that respectively give pleasure and neither pain nor pleasure to the receivers of the actions (i.e. the guru's wife and liquor) but do give pain to the agent (i.e. the student and the drinker). This is to refute the claim attributed to Sāṅkhya that the nature of the fruit of an action befalling the agent in the future depends on the effect of that action on the receiver in the present. See Ślokavārttika, codanā, 236cd–237ab (Kataoka, Kumārila on Truth, Vol. 1, p. 53 and Vol. 2, p. 488). For an analysis of the Sāṅkhyas’ claim introduced in the Ślokavārttika (codanā, 235cd–236ab) and the similarity it bears to the Buddhists’ arguments, see Ham, Hyoung Seok, ‘On a Bhāviveka-Sāṅkhya alliance against ritual killing: explaining two nearly identical syllogisms held by Bhāviveka and the Sāṅkhyas respectively against the Lokāyatas and the Mīmāṃsakas’, South Asian Classical Studies 13 (2018), pp. 359–375Google Scholar. The opponent in the YD (YD_WM 33:7–11), on the other hand, mentions the example of ‘sexual intercourse with the guru's wife’ (gurubhāryāgamana), but not the example of drinking liquor. Instead, there is another example of a teacher, according to the Sāṅkhyas’ reasoning, ‘who would obtain an undesirable fruit’ (aniṣṭaphalasambandhaḥ syāt) for having a young boy (māṇavaka) engage in righteous actions such as ‘being celibate’ (brahmacarya), ‘repeated recitation of the Veda’ (svādhyāyābhyāsa), and ‘living on alms’ (bhaikṣa). Houben, based on this case, observes that ‘the treatment of the topic in the YD's pūrvapakṣa seems not really dependent on Kumārila's discussion’ (Jan E. M. Houben, ‘To kill or not to kill the sacrificial animal (yajña-paśu)?: arguments and perspectives in Brahminical ethical philosophy’, in Violence Denied: Violence, Non-Violence and the Rationalization of Violence in South Asian Cultural History, (eds) Jan Houben and Karel van Kooij (Leiden/Boston/Köln, 1999), p. 150). Nevertheless, he acknowledges that the opponent of the YD is a Mīmāṃsaka when he states that ‘the YD gives a pūrvapakṣa with sophisticated Mīmāṃsā-like arguments which presupposes, if not Kumārila's Śl[oka]V[ārttika], in any case other Mīmāṁsā-texts apart from Jaimini's Sūtra and Śabara's Bhāṣya’ (ibid., p. 151).

¹⁵ YD_WM 32:1–8, ‘tatprāmāṇyānabhyupagamād adoṣa iti cet. …etac cāyuktam. kasmāt? abhyupagamavirodhāt. ‘dṛṣṭam anumānam āptavacanaṃ ca’ iti prāmāṇyatrayam abhyupagataṃ bhavadbhiḥ. idānīṃ vedasyāptavacanatve saty aprāmāṇyaṃ bruvataḥ svamatavyāghātaḥ. tasmād ayuktam etat.’

¹⁶ YD_WM 34:9-12, ‘ucyate. na, abhiprāyānavabodhāt. citram api bahv etad abhidhīyamānaṃ¹ nābhiprāyaṃ ² spṛśatīty upekṣyate. kiṃ kāraṇam? [yasmān] na vayaṃ vedasya prāmāṇyaṃ ³ pratyācakṣmahe. no khalv api brūmaḥ śāstracoditāyāṃ hiṃsāyāṃ pravartamānasyāniṣṭaphalasambandho bhavati.’ [¹ Emended from abhidhīyamāno. ² One manuscript (abbreviated as ‘Dkha’ in YD_WM) attests a better reading: abhidhīyamānaṃ no nābhiprāyaṃ. ³ Emended from vedasyaprāmāṇyaṃ.]

¹⁷ See YD_WM 34:22–35:8.

¹⁸ For example, in an often-quoted sentence from the Tantravārttika, Kumārila lists Sāṅkhya's texts along with those of the Buddhists and Jainas and characterises them as ‘not endorsed by those who know the Veda’ (trayīvidbhir na parigṛhītāni), ‘contrary to the Veda’ (trayīviparīta-), ‘unconnected with [the Veda]’ (-asambaddha-), and ‘teaching about different matters [from dharma, that is,] mostly about livelihood though being infused with the fragrance of a bit of wholesome teachings that accord with śruti and smṛti such as non-violence, honesty, self-control, giving, and compassion’ (ahiṃsāsatyavacanadamadānadayādiśrutismṛtisaṃvādistokārthagandhavāsitajīvikāprāyārthāntaropadeśīni). See Subbāśāstrī (ed.), Śrīmajjaiminipraṇīte Mīmāṃsādarśane ādita ārabhya dvitīyādhyāyaprathamapādāntaḥ prathamo bhāgaḥ (Poona, 1929), 194:8–13, and Kunio Harikai (ed.), ‘Sanskrit text of the Tantravārttika: Adhyāya 1, Pāda 3, Adhikaraṇa 1~3 collated with five manuscripts’, Annual Report of Medical Anthropology and Humanity 3 (2008), 42:4–13.

¹⁹ The opponent supposes that the Sāṅkhyas—those who acknowledge the Vedic authority—perceive killing as impure based on the Veda's instruction. The YD implicitly concurs with this by providing no refutation. Cf. YD_WM 32:15–18, ‘Suppose that you are also given a question on this matter: “How do you ascertain that killing is impure because it destroys the cherished bodies of living beings?” Then [you] should certainly answer: “Based on the Veda”’ (yadi caitasminn arthe bhavān api paryanuyujyeta katham idaṃ niścīyate yad uta prāṇiṇām iṣṭaśarīravyāpādanād aviśuddhir hiṃseti, avaśyam abhidhānīyaṃ śāstrata iti.).

²⁰ The YD_WM (39:18–19) characterises the Veda (āmnāya) as ‘not preceded by human intellect, independent, and working for the sake of what is ultimate for human’ (apuruṣabuddhipūrvakaḥ svatantraḥ puruṣaniḥśreyasārthaṃ pravartamānaḥ). This is the reiteration of the Mīmāṃsaka opponent's claim that the YD introduces at YD_WM 32:13–14 (apuruṣabuddhipūrvakas tv āmnāyaḥ svatantraḥ puruṣaniḥśreyasārthaṃ pravartate).

²¹ YD_WM 43:5–9, ‘śabdasāmarthyān nityatvam iti cet. syād etat. śabdapramāṇakā vayam. yac chabda āha tad asmākaṃ pramāṇam. sa cāsya hetor amṛtatvam āha tarati mṛtyuṃ tarati pāpmānam ityādi. tasmād anicchatāpy etad avaśyam abhyupagantavyam. anabhyupagame vā pratijñāhānir vedaḥ pramāṇam iti.’

²² This paragraph is based on the YD's commentary (YD_WM 42:12–45:12) on the word ‘destruction’ (kṣaya).

²³ Cf. Patrick Olivelle (trans.), The Early Upaniṣads: Annotated Text and Translation (Oxford, 1998), pp. 236–237.

²⁴ This is the Taittirīya Saṃhitā 2.1.1.4. See Keith, Arthur Berriedale, The Veda of the Black Yajus School Entitled Taittiriya Sanhita (Cambridge, 1914), part 1, pp. 133–134Google Scholar.

²⁵ This paragraph is based on the YD's commentary (YD_WM 45:13–47:3) on the word atiśaya (‘relative superiority’).

²⁶ YD_WM 46:(5), ‘It is because even [in the case of a person] who merely mutters a prayer, [his] body certainly attains the status of being a part of [that ritual]’ (japamātram api hi kurvato ’vaśyaṃ śarīram aṅgabhāvam eti).

²⁷ Cf. Atharvaveda Saṃhitā 11.8.32, ‘Therefore, indeed, one who knows man (puruṣa) thinks “this is brahman”; for all deities are seated in him, as cows in a cow-stall’ (tasmād vai vidvān puruṣam idaṃ brahmeti manyate/ sarvā hy asmin devatā gāvo goṣṭha ivāsate//); the translation is from Whitney, William Dwight, Atharva-Veda Saṁhitā: Translated With a Critical and Exegetical Commentary (Cambridge, 1905), part 2, p. 651Google Scholar. For the Sanskrit text, see R. Roth and W. D. Whitney (eds), Atharva Veda Sanhita (Berlin, 1855), 261:11–12. Note that pāda d of the Atharvaveda Saṃhitā 11.8.32 is different from the YD's verse. However, Whitney (ibid.) notes that one manuscript has a reading of ‘śārire ’dhi samāhitāḥ,’ as it is quoted in the YD.

²⁸ Bronkhorst investigates the Vedāntins such as Śaṅkara who ‘present themselves as Mīmāṃsakas’ (Johannes Bronkhorst, ‘Vedānta as Mīmāṃsā’, in Mīmāṃsā and Vedānta: Interaction and Continuity, (ed.) J. Bronkhorst (Delhi, 2007), pp. 33–34). The author of the YD, unlike those Vedāntins, does not present Sāṅkhya philosophy as ‘Mīmāṃsā at heart’ (ibid.) but it can be called Mīmāṃsa in a limited sense as long as he does not deny the Mīmāṃsā mode of argumentation and the ideological presuppositions that it makes.

²⁹ Olivelle, Patrick, The Āśrama System: The History and Hermeneutics of a Religious Institution (Oxford, 1993), p. 99Google Scholar.

³⁰ A long digressive section on the legitimacy of renunciation begins with the opponent's remarks quoted below. The section ends at YD_WM 42:4 where the YD moves on to provide its comments on the second fault of the Vedic sacrifice, namely, ‘destruction’ (kṣaya).

³¹ YD_WM 35:9, ‘sannyāsānupapattiḥ aviyogaśravaṇāt.’

³² YM_WM 31:1-2, ‘Question. Then, what is this revelation? Answer. Mantra and brāhmaṇa’ (āha. kaḥ punar ayam anuśravaḥ? ucyate. mantrabrāhmaṇam.). Cf. Mīmāṃsāsūtra 2.1.32–33: ‘The name “mantra” is applied to those texts that are expressive of the said (assertion, of things connected with prescribed acts). To the rest (of the Veda) the name “brāhmaṇa” (is applied)’ (taccodakeṣu mantrākhyā// śeṣe brāhmaṇaśabdaḥ//). The translation is from Ganganatha Jha (trans.), Shabara-Bhāṣya (Baroda, 1933), Vol. 1, pp. 202 and 204. For the Sanskrit text, see Subbāśāstrī, Śrīmajjaiminipraṇīte Mīmāṃsādarśane, 434:3 and 436:2.

³³ Cf. Mīmāṃsāsūtra 1.1.2, ‘Dharma is a beneficial action defined by an injunction’ (codanālakṣaṇo ’rtho dharmaḥ//). For the Sanskrit text, see Frauwallner, Erich, Materialien zur ältesten Erkenntnislehre der Karmamīmāṃsā (Wien/Graz/Köln, 1968), 16:8Google Scholar.

³⁴ For discussions on arthavāda and nāmadheya in the Mīmāṃsā literature, see Arthur Berriedale Keith, The Karmamīmāṃsā (Calcutta, 1921), pp. 79ff. and Jha, Ganganatha, Pūrvamīmāṃsā in its Sources (Varanasi 1964), pp. 159ffGoogle Scholar.

³⁵ Cf. Mīmāṃsāsūtra 1.2.7, ‘Being construed along with injunction, they would serve the purpose of commending those injunctions’ (vidhinā tv ekavākyatvāt stutyarthena vidhīnāṃ syuḥ//). The text and translation are from Kunio Harikai, ‘Mīmāṃsaka theory of Gauṇa or metaphor from Śabarasvāmin to Kumārilabhaṭṭa’, in Dieux, génies, anges et démons dans les cultures orientales & florilegium indiae orientalis Jean-Marie Verpoorten in honorem, (eds) Christophe Vielle, Christian Cannuyer and Dylan Esler (Bruxelles, 2017), p. 279.

³⁶ See the relevant discussion at YD_WM 38:8–39:3.

³⁷ YD_WM 39:17–19, ‘yad dhi kartavyatayā neṣṭaṃ tad apuruṣabuddhipūrvakaḥ svatantraḥ puruṣaniḥśreyasārthaṃ pravartamāna āmnāyaḥ kim iti prarocayet?’

³⁸ Cf. YD_WM 40:7–9, ‘Furthermore, [renunciation is established] because [your argument] is uncertain. It is not certain that only what is enjoined is to be done. Likewise, the followers of Śabara recite the following. ‘Here, [in case of the following sentence, that is,] “Going to the village is glorious for you,” eulogy alone, without an injunction, makes Devadatta like to go to the village’ (kiṃ cānyat. anekāntāt. na cāyam ekānto yad vihitam eva kartavyam. tathā ca śābarāḥ paṭhanti grāmagamanaṃ bhavataḥ śobhanam ity atrāntareṇa vidhiṃ stutir eva devadattaṃ grāmagamanāya prarocayatīti).

³⁹ YD_WM 40:16-41:2. Kumārila also considers diverse branches of Vedic transmission lineage and people's carelessness as reasons why we cannot find the root Vedic texts in case of some smṛti passages (Kiyotaka Yoshimizu, ‘Kumārila's criticism of Buddhism as a religious movement in his views on the sources of dharma', Acta Asiatica 108 (2015), p. 46). Cf. Tantravārttika (Subbāśāstrī, Śrīmajjaiminipraṇīte Mīmāṃsādarśane, 164:18–19 and Harikai, ‘Sanskrit text of the Tantravārttika: Adhyāya 1, Pāda 3, Adhikaraṇa 1~3’, 9:26–27): ‘The source [Vedic text] of the smṛti is not found [in some cases] because Vedic branches are scattered [over the world], human beings are careless [in their search for the root text], and [the matter at stake] is [mentioned] in various chapters’ (śākhānāṃ viprakīrṇatvāt puruṣāṇām pramādataḥ/ nānāprakaraṇasthatvāt smṛter mūlaṃ na dṛśyate//). The translation is adapted from Kei Kataoka, ‘Transmission of scripture: exegetical problems for Kumārila and Dharmakīrti’, in Scriptural Authority, Reason and Action, (eds) Vincent Eltschinger and Helmut Krasser (Wien, 2013), p. 252.

⁴⁰ Olivelle, The Āśrama System, pp. 84–85.

⁴¹ The translation of this passage is adapted from Olivelle, The Early Upaniṣads, pp. 279 and 281.

⁴² Cf. Olivelle, The Early Upaniṣads, p. 437.

⁴³ YD_WM 52:15-23, ‘vidhisadbhāvāt kriyāprādhānyam iti cet, na, uktatvāt. katham etat? nāsti vidhikṛto viśeṣaḥ. upetya vā tatrāpi tadutpatteḥ. asti hi jñānasyāpi vidhāyakaṃ śāstram. katham? evaṃ hy āha. ya ātmāpahatapāpmā vijighatso vipipāso vijaro vimṛtyur viśokaḥ <satya>saṃkalpaḥ so ’nveṣṭavyaḥ sa jijñāsitavyaḥ. sa sarvāṃś ca kāmān avāpnoti sarvāṃś ca lokān yas tam ātmānam anuvidya vijānātīti prajāpater vacanaṃ śrūyate <iti>. punar apy āha. dve vidye veditavye parā caivāparā ca yā <iti>. tasmād vidhisadbhāvāt kriyāprādhānyam iti svapakṣānurāgamātram¹ etat’ [¹ Emended from svapakṣānūrāgamātram].

⁴⁴ As Yoshimizu points out, Kumārila changes his opinion on whether the Upaniṣadic injunctions have their unique objective of liberation or not, most likely referring to the same Chāndogya Upaniṣad passage quoted in the YD. See Kiyotaka Yoshimizu, ‘Kumārila's reevaluation of the sacrifice and the Veda from a Vedānta perspective’, in Mīmāṃsā and Vedānta: Interaction and Continuity, (ed.) J. Bronkhorst (Delhi, 2007), pp. 235–236.

In the Ślokavārttika, Kumārila considers that the Upaniṣadic injunction for the knowledge of ātman serves the purpose of ritual actions. Cf. Ślokavārttika (Sambandhākṣepaparihāra) vv. 103–104, ‘[The Upaniṣadic injunction] that ātman should be known is not enjoined for the sake of liberation. It [only] indicates that the knowledge of ātman is a cause of engagement in a ritual action. As long as it is accepted that [the knowledge of ātman] is for the sake of other [objectives, that is, rituals], the description of its result must be a eulogy; [thus, the result of the knowledge of ātman] is no other than the result [of ritual actions] such as heaven’ (ātmā jñātavya ity etan mokṣārthaṃ na ca coditam/ karmapravṛttihetutvam ātmajñānasya lakṣyate// vijñāte cāsya pārārthye yāpi nāma phalaśrutiḥ/ sārthavādo bhaved eva na svargādeḥ phalāntaram//). The translation of verse 104 is adapted from Yoshimizu, ‘Kumārila's reevaluation of the sacrifice’, p. 235, fn. 109.

In his Tantravārttika, however, Kumārila, referring to the Chāndogya Upaniṣad 8.15.1, confirms the distinctiveness of the Upaniṣadic injunctions. For example, he states: ‘this [Upaniṣad] passage that declares the result of the state of attaining the supreme self (paramātman) characterised by the non-return [to this world] is not eulogy (arthavāda) since [the knowledge of ātman] does not belong to [any ritual] context and is not inherently related with a[ny] sacrificial performance’ (apunarāvṛttyātmakaparamātmaprāptyavasthāphalavacanam aprakaraṇagatatvena anaikāntikakratusaṃbandhāc ca na … arthavādatvam; Subbāśāstrī, Śrīmajjaiminipraṇīte Mīmāṃsādarśane, 288:15-17 and Kunio Harikai (ed.), ‘Sanskrit text of the Tantravārttika: Adhyāya 1, Pāda 3, Adhikaraṇa 9 Vyākaraṇa Adhikaraṇa collated with five manuscripts’, South Asian Classical Studies 6 (2011), pp. 16–19).

⁴⁵ The translation is from Larson, Classical Sāṃkhya, p. 256.

⁴⁶ With the exception of the Tattvakaumudī, all the classical commentaries on the Sāṅkhyakārikā understand ‘tad-’ as referring to the visible and Vedic means discussed respectively in Sāṅkhyakārikā 1 and 2. For example, see Suvarṇasaptati 1245b27–29, ‘Question. If then, which means is superior? Answer. ‘What is opposite to these two is superior.’ The two means are, namely, 1. what is taught by the medical science and 2. what is taught by the Veda. The [superior] means, which is opposite to these two means, is attained by the investigation (jijñāsā)’ (外曰: 若爾何因爲勝? 答曰: 翻此二因勝. 謂二因者. 一醫方所說, 二皮陀所說. 翻此兩因欲知所得因); see also Sāṅkhyavṛtti (6:16–17), Sāṅkhyasaptativṛtti (9:12–13), Gauḍapādabhāṣya (Colebrooke, H. T. and Wilson, H. H. (eds and trans), The Sankhya káriká, or, Memorial Verses on the Sánkhya Philosophy (Oxford, 1837), 3:14–15Google Scholar), Jayamaṅgalā (Sarma and Vangiya, Sāṃkhyakārikā of Śrīmad Īśvarakṛṣṇa, 67:22–23), and Māṭharavṛtti (6:27–28). The Tattvakaumudī takes ‘tad-’ to refer to ‘revelational means’ (ānuśravika) alone by glossing the pronoun as ‘tasmād ānuśravikāt’ (Srinivasa Ayya Srinivasan (ed.), Vācaspatimiśras Tattvakaumudī: Ein Beitrag zur Textkritik bei kontaminierter Überlieferung (Hamburg, 1967), 74:28–29).

⁴⁷ See the marginal note (YD_WM 47:(3)) on the word ‘superior’ (śreyas): ‘Since the visible means is far removed [from the Veda], even consideration of it is not appropriate here. With this intention, the author of the [Sāṅkhya]kārikā stated “better,” not “best’’’ (dṛṣṭasya tu dūrāpāstatvād atra gaṇanaiva na yuktety āśayena kārikākāraḥ śreyān ity abhyadhān na tu śreṣṭhaḥ).

⁴⁸ Cf. YD_WM 47:10–13, ‘tad ity anena karmavidhiniṣpāditasya svargaprāptilakṣaṇasya phalasyābhisambandhaḥ. tasmād viparītaḥ śuddho ’kṣayo niratiśaya ity arthaḥ. ko 'sāv ity ucyate mokṣaḥ śreyān. etad uktaṃ bhavati. ubhāv apy etau praśasyau svargāpavargāv āmnāyavihitatvāt, mokṣas tu praśasyataraḥ.’

⁴⁹ See YD_WM 50–52 for these quotations.

⁵⁰ YD_WM 54:5, ‘paraṃ rahasyaṃ vedānām avasāneṣu paṭhyate/’.

⁵¹ The YD, in fact, uses the term ‘āśramāntara’ (another āśrama). It is the marginal note that glosses that expression as ‘sannyāsāśrama’ (YD_WM 53:(1)). Note that the original meaning of āśrama refers to the residence and lifestyle of a particular type of householder and, thus, ‘another āśrama’ basically means ‘āśrama other than that of householders’. Olivelle states that ‘the compound āśramāntara is used with the meaning “other than a householder” in the Mahābhārata and even by medieval authors such as Śaṃkara and Kumārila’ (Olivelle, The Āśrama System, p. 23).

⁵² Instead of ‘stutiḥ’ adopted in YD_WM 54:11, I took the reading of ‘stutā’ listed as a variant in the manuscripts.

⁵³ YD_WM 54:11–16, ‘kaivalyaprāptihetutvād yā vedavihitā stutā/ praśastā yājñavalkyādyair viśiṣṭais tattvaniścayāt// seyaṃ viṣayarāgāndhair viparītārthavādibhiḥ/ vidyā kanyeva paṇḍāya dīyamānā na śobhate// tasmād rāgānugair uktāṃ kuhetupṛtanām imām/ apohya matimān yuktyā hy āśramād āśramaṃ vrajet//’.

⁵⁴ Although it is beyond the scope of this study, there is a verse attributed to Pañcaśikha that claims the possibility of liberation in any of āśrama. It says: ‘One who knows the 25 principles will be emancipated, whichever stage of life he may dwell in, whether he may have twisted hair, or a shaved head or knotted hair. There is no doubt about this’ (pañcaviṃśatitattvajño yatra tatrāśrame vaset/ jaṭī muṇḍī śikhī vāpi mucyate nātra saṃśayaḥ//). The text and translation are from Motegi, Shujun, ‘The teachings of Pañcaśikha in the Mokṣadharma’, Asiatische Studien 53.3 (1999), p. 513Google Scholar. This verse must have been famous among the Sāṅkhyas since it or its variants is quoted in most of the commentaries on the Sāṅkhyakārikā, namely, Suvarṇasaptati, Sāṅkhyavṛtti, Sāṅkhyasaptativṛtti, Gauḍapādabhāṣya, Jayamaṅgalā, Māṭharavṛtti. Curiously enough, it is not quoted in the YD and the Tattvakaumudī. Those commentaries that quote the verse, however, pay attention only to the fact that the verse is speaking of the 25 principles (tattva) and neglects the implication of this on the relationship between āśrama and liberation. But Pañcaśikha seems to be related to the doctrine that liberation is possible in any āśrama considering the contents of the Mahābhārata, Book 12, Chapter 308 where the king Janaka, who is the disciple of Pañcaśikha, debates with Sulabhā under the framing question, ‘concerning whether there has ever existed a man who attained emancipation without giving up the position of king’ (Motegi, ‘The teachings of Pañcaśikha’, p. 518). Therefore, there must have been diverging opinions over the necessary relationship between āśrama (especially, sannyāsa) and liberation among the Sāṅkhya thinkers and it may not be a coincidence that the YD and the Tattvakaumudī do not quote this Pañcaśikha's verse.

⁵⁵ Motegi, ‘The teachings of Pañcaśikha’, pp. 534–535, fn. 80.

⁵⁶ Eltschinger, Vincent, Buddhist Epistemology as Apologetics: Studies on the History, Self-understanding and Dogmatic Foundations of Late Indian Buddhist Philosophy (Wien, 2014), p. 95CrossRefGoogle Scholar.

⁵⁷ See Larson and Bhattacarya, Sāmkhya, pp. 15–16.

⁵⁸ The Tattvakaumudī considers even the ‘impurity’ (aviśuddhi) of Vedic sacrifice mentioned in Sāṅkhyakārikā 2 to be caused by a Vedic prohibition, namely, ‘do not kill any living being’ (mā hiṃsyāt sarvā bhūtāni). Thus, Vācaspati‘s discussion is not whether Sāṅkhya recognises the authority of the Vedic mandate to kill an animal in ritual contexts; rather, it discusses the contradiction between the Vedic prohibition (‘do not kill’ (mā hiṃsyāt)) and the Vedic injunction (‘One should slaughter an animal dedicated to Agni and Soma’ (agnīṣomīyaṃ paśum ālabheta)). Seeing no contradiction between the two (na…kaścid virodho ’sti), Vācaspati concludes that one act of ritual killing causes two—one positive and one negative—results. Cf. Tattvakaumudī 74:9–10, ‘[There is no contradiction] because [ritual killing] will increase sin for the human [agent] but [at the same time] will help [the completion of] ritual’ (sā hi puruṣasya doṣam āvakṣyati kratoś copakariṣyatīti). However, having quoted a passage attributed to Pañcaśikha that speaks of how the demerit mixed with the merit from performing a sacrifice is removable and endurable (svalpaḥ saṅkaraḥ saparihāraḥ sapratyavamarṣa iti), Vācaspati (72:23–74:2) explains that such a negative mixture can be removed by expiatory rites (prāyaścitta) and, even if one neglects to perform them, the pain caused by animal slaughter is bearable to ‘someone who has already plunged into the great lake of nectar’ (-sudhāmahāhradāvagāhin).

⁵⁹ Olivelle, The Āśrama System, p. 7.