Classical S ā ṁ khya on the Authorship of the Vedas

The question as to whether the Vedas have an author is the topic of vivid polemics in Indian philosophy. The aim of this paper is to reconstruct the classical Sa¯m˙khya view on the authorship of the Vedas. The research is based chieﬂy on the commentaries to the S ā ṁ khyak ā rik ā deﬁnition of authoritative verbal testimony given by the classical Sa¯m˙khya writers, for these fragments provide the main evidence (both direct and indirect) for the reconstruction of this view. The textual analysis presented in this paper leads to the following conclusion. According to most classical Sa¯m˙khya commentaries, the Vedas have no author. Two commentators state directly that the Vedas have no author, and four commentators allude to the authorlessness of the Vedas. Only one commentator seems to hold the opposite view, stating that all the authoritative utterances are based on perception or inference of imperceptible objects by authoritative persons, from which it follows that the Vedas too have an author or authors.

The conception of Īśvara as an author of the Vedas is not present in earlier Nyāya works, such as the Nyāyasūtras, Nyāyabhāṣya and Uddyotakara's (550-610 CE) Nyāyavārttika. This conception is distinctly formulated in the Nyāyamañjarī of Jayanta Bhaṫṫa (840-900 CE) (Nyāyamañjarī. Summary prepared by K.H. Potter, J.V. Bhattacharya, and U. Arya. In : Potter 1977, pp. 371, 377-378). As regards early Vaiśeṡika, this conception is not found in the Vaiśeṣikasūtras. It is stated in the Vaiśeṣikasūtravṛtti of Candrānanda (the eighth or the ninth century CE), which is probably the oldest surviving commentary on the Vaiśeṣikasūtras. According to this commentary, the Vedas has been uttered by Maheśvara (see Bronkhorst 1996, p. 288). The conception of Īśvara as an author of the Vedas has become an established view of Nyāya and Vaiśeṡika, presented in many texts, for example, Vyomavatī of Vyomaśiva (900-960 CE) (Vyomavatī. Summary prepared by V. Varadachari. In : Potter 1977, pp. 429, 447), Nyāyavārttikatātparyaṭīkā (see II, 1, 68) of Vācaspati Miśra (900-980 CE), Ātmatattvaviveka (Ātmatattvaviveka. Summary prepared by V. Varadachari. In : Potter 1977, pp. 555-556), Nyāyakusumāñjali (Nyāyakusumāñjali. Summary prepared by K.H. Potter and J.V. Bhattacharya. In: Potter 1977, pp. 569, 588) of Udayana (1050-1100 CE), and Tarka-saṁgraha and Tarka-dīpikā (see 62) of Annambhaṫṫa (the seventeenth century CE). For the dates of Candrānanda and Annambhaṫṫa I consulted Matilal (1977, pp. 74-75, 107). For the dates of all the other writers of Nyāya and Vaiśeṡika mentioned in this paper see Potter (1977, pp. 9-12). 2 The dates of all the Sāṁ khya and Yoga texts mentioned in this paper are given according to Larson and Bhattacharya (1987, pp. 15-18, 19-22). It is necessary to point out the limitations of this research. This paper is based chiefly on those fragments of the classical Sāṁ khya commentaries where the Sāṁ khya writers explain the 'definition' of authoritative verbal testimony (āptavacana) given in the Sāṁkhyakārikā (SK). Sāṁ khya accepted three pramāṇas: perception (dṛṣṭa), inference (anumāna), and authoritative verbal testimony (āptavacana, literally, 'authoritative utterance')-see kārikās 4-6 where Īśvarakṙṡṅa presents the Sāṁ khya teaching on the sources of valid knowledge (pramāṇa). The commentaries to the SK 'definition' of āpta-vacana (Īśvarakṙṡṅa 'defines' it as āptaśruti-"authoritative śruti"; 3 see SK 5) contain the main evidence (both direct and indirect) for reconstruction of the Sāṁ khya view on the authorship of the Vedas. All the Sāṁ khya writers interpreting this 'definition' tried to establish sources of authoritative utterances, i.e. valid sentences being the cause of valid knowledge which is called authoritative verbal testimony. The Vedas were considered to be the main source of authoritative utterances. Besides the commentaries to the definition of authoritative verbal testimony, the direct evidence presented in TK 2 has been considered. Thus all the direct evidence contained in the classical Sāṁ khya texts, which is scarce (we find it in the three fragments: YD 5, TK 5, and TK 2), seems to have been taken into account.
The SK provides no evidence for reconstruction of the view on the authorship of the Vedas, so we concentrate on the SK commentaries, particularly on their definitions of authoritative verbal testimony (i.e. their interpretations of the SK definition of āptavacana). I shall cite these definitions or their parts containing the material for our analysis. I shall not provide a comprehensive analysis of these definitions; only those aspects will be considered which are important for clarifying the view on the authorship of the Vedas.

The Commentary Translated into Chinese by Paramārtha
This is the definition of authoritative verbal testimony given in this commentary, together with the version of the SK definition we find here. Translation of P 5 from Takakusu (La Sāṁ khyakārikā étudiée à la lumière de sa version chinoise, 1904): A teaching of a saintly person is called sacred authority. 5 … A teaching of a saintly person etc. For example, the four Vedas uttered by the god Brahmā and the Dharmaśāstra of the king Manu. 6 Translation of P 5 from the Sanskrit reconstruction of N. Aiyaswami Sastri (Suvarṇasaptati Śāstra, 1944): The word of authority (āpta-śruti) is called authoritative verbal testimony (āpta-vacana). The word of authority is called authoritative verbal testimony-for example (yathā), 7 that what is uttered by Brahmā and by Manu, that is, the four Vedas and the Treatise on Dharma (dharma-śāstra). 8 According to this commentary, authoritative verbal testimony is the word of an authoritative person. So, authoritative statements are the statements uttered by authoritative persons. The sources of authoritative statements include the Vedas uttered by the god Brahmā and the Treatise on Dharma uttered by Manu. Manu is Manu Svayaṁ bhuva ('self-existent'), the first Manu of our kalpa, 9 to whom the Hindu tradition ascribes Manusmṛti (also called Mānavadharmaśāstra or Manusaṁhitā), the most authoritative of the dharmaśāstras, treatises on dharma. He is the son of Brahmā. By the Treatise on Dharma (dharmaśāstra used in singular number) Manu-smṛti is probably meant. In P all the authoritative statements are associated with certain authoritative persons, but it is not said whether these persons are the authors of these statements. The Vedas too are associated with the certain person, i.e. the god Brahmā who uttered them, but it is not clear if the god Brahmā is the author of the Vedas. 5 I use italics to distinguish the text of the kārikā. It is necessary to note that there is an English translation from Takakusu's rendering, prepared by S.S. Suryanarayanan, who is known also as S.S. Suryanarayana Sastri (The Sāṁ khya Kārikā Studied in the Light of Its Chinese Version, 1932Version, , 1933, and my translation is very similar to it. 7 Yathā can be also translated as 'namely'. 8 āpta-śrutir āpta-vacanam ucyate ‖ … āpta-śrutir āpta-vacanam ucyate iti | yathā brahmaṇā manunā ca uktāś catvāro vedā dharmaśāstrañ ca ‖ 9 Kalpa is the day of Brahmā in Hindu cosmology. After it the night of Brahmā comes, i.e. the dissolution (pralaya) of the created world, after which the next day of Brahmā begins. Every kalpa consists of 14 manvantaras, and every manvantara ('the age of a Manu') has its own Manu, its own seven great seers, its own Indra and other gods.

The Sāṁkhyavṛtti and the Sāṁkhyasaptativṛtti
The definitions of authoritative verbal testimony given in these commentaries are very similar. The definition of the SVṙ (5) is as follows: This is the definition (lakṣaṇa) of authoritative verbal testimony (āpta-vacana). It is said here [i.e. in the SK]: "The revelation of authorities (āpta-śruti) and the word of authorities (āpta-vacana) [constitute āpta-vacana-authoritative verbal testimony]". Here the name 'āpta' [is applied to] the teachers (ācārya): Hari, Hara, Hiraṅyagarbha, etc. The 'revelation' (śruti) of these authorities is the Vedas. That is the meaning [of āptaśruti being the first part of the definition of authoritative verbal testimony]. The 'authorities' (āpta) which are the authors (kāra) of the treatises on dharma (dharma-śāstra) are Manu, etc., [and] 'the word of authorities' (āpta-vacana) is 'the word' (vacana) of these authorities-that [is the meaning of āptavacana being the second part of the definition of authoritative verbal testimony]. 10 The definition we find in the SSVṙ 5 is practically the same: "What is the definition (lakṣaṇa) of authoritative verbal testimony?" On this account it is said [in the SK]: "The revelation of authorities (āpta-śruti) [and] the word of authorities (āpta-vacana) [constitute āpta-vacana-authoritative verbal testimony]". There the name 'āpta' [is applied to] Hari, Hara, Hiraṅyagarbha, etc. 'The revelation of authorities' (āpta-śruti) is the 'revelation' (śruti) of these 'authorities' (āpta). The 'revelation of authorities' (āpta-śruti) is the Vedas. That is the meaning [of āptaśruti being the first part of the definition of authoritative verbal testimony]. 'Authorities' (āpta) are the authors (kartṛ) of the treatises on dharma (dharma-śāstra): Manu, etc., [and] 'the word of authorities' (āpta-vacana) is 'the word' (vacana) of these authorities-that [is the meaning of āptavacana being the second part of the definition of authoritative verbal testimony]. 11 In the SVṙ and SSVṙ āptaśruti and āptavacana are interpreted as the two parts of the definition of āptavacana. This reading seems to contradict the intention of the author of the SK. It is clear from kārikā 4 that Īśvarakṙṡṅa treats āptavacana as the defined notion (definiendum), and not as the defining expression (definiens). In kārikā 4 Īśvarakṙṡṅa introduces the sources of valid knowledge to be defined in kārikā 5, and one of them is āptavacana.
According to the SVṙ and SSVṙ, authoritative verbal testimony is "the revelation of authorities (āpta-śruti) and the word of authorities (āpta-vacana)". The revelation 10 āpta-vacanasya lakṣaṇam iti | atrôcyate āpta-śrutir āpta-vacanaṁ ca | atra āptā nāmâcāryāḥ harihara-hiraṇyagarbhâdayaḥ teṣāṁ āptānāṁ śrutiḥ veda ity arthaḥ | manv-ādayo dharma-śāstra-kārāḥ āptāḥ teṣām āptānāṁ vacanam āpta-vacanam iti | The Sanskrit text of the SVṙ and SSVṙ is given by me together with the editor's corrections and insertions of E. Solomon and without marking these emendations. 11 āpta-vacanasya kiṁ lakṣaṇam ity atrôcyate āpta-śrutir āpta-vacanam | tatra āptā nāma hari-harahiraṇyagarbhâdayas teṣāṁ āptānāṁ śrutir āpta-śrutiḥ | āpta-śrutir veda ity arthaḥ | manv-ādayo dharmaśāstrāṇāṁ karttāra āptās teṣām āptānāṁ vacanam āpta-vacanam iti | of authorities is the Vedas, which is the revelation of such authorities as Hari, Hara, and Hiraṅyagarbha. The word of authorities is dharmaśāstras, composed by Manu and other authorities. In the SVṙ and SSVṙ, as in P, all the authoritative statements are associated with certain authoritative persons. Hari, Hara, and Hiraṅyagarbha ('a golden embryo') are the epithets respectively of Viṡṅu, Ś iva, and Brahmā, the main Hindu gods, which were often understood as the three aspects of one divine being from which the universe and the Vedas originate. Did these gods (1) compose (create) the Vedas (being their authors), (2) discover them, (3) remember them from the previous cycle of the existence of the world, 12 or (4) give them out spontaneously? In the SVṙ and SSVṙ we do not find a direct answer to this question. None of these possibilities can be excluded.
As regards the last possibility, in the SVṙ and SSVṙ the Vedas could be comprehended as issuing spontaneously from the mouths of Hari, Hara, and Hiraṅyagarbha or from one divine being of which these gods are the forms. Such a view on the origin of the Vedas could also be peculiar to the P: we cannot exclude that in the P the Vedas were understood as issuing spontaneously from the mouth of Brahmā. According to the Bṛhadāraṇyakopaniṣad (see II, 4, 10), the Vedas go out of "the great being" (mahad bhūtam) together with its breath. It is said here that the Vedas are breathed out (niḥśvasita) by this being. Later this idea was developed by the famous Advaita Vedāntin Ś aṅkara (flourished at the beginning of the eighth century CE). 13 Ś aṅkara says that the Vedas arise from Brahman spontaneouslylike a breath which issues from a man spontaneously, without any special, conscious effort. In his commentary to the Brahmasūtra, Ś aṅkara calls Brahman, being the source (yoni) of the Vedas, omniscient and omnipotent, from which it is clear that it is Īśvara, the subtlest manifest form of Brahman, by whom the Vedas are "breathed out" (see Śaṅkara's Brahmasūtrabhāṣya I, 1, 3 and Bṛhadāraṇyakopaniṣadbhāṣya II, 4, 10). Unlike Advaita Vedānta, the classical Sāṁ khya accepts neither the existence of God (Īśvara), nor the theory that from the "highest standpoint" (pāramārthika) Brahman is the true essence of everything and the only real existent. I suppose that in Sāṁ khya the view that the Vedas spontaneously go out of Brahmā or other divine beings could have the following shape: the Vedas arise from prakṛti (primordial creative matter) spontaneously, but prakṛti "creates" them by the agency of the creator god Brahmā or other divinities which appear at the beginning of a new cycle of the existence of the world. It is worth noticing that according to the postclassical Sāṁkhyasūtras and Sāṁkhyasūtravṛtti, which like the classical Sāṁ khya deny the existence of God, 14 the authorless Vedas evolve from prakṛti spontaneously, though there is no mention of the role of Brahmā or like beings in these texts (see Sāṁkhyasūtras and Sāṁkhyasūtravṛtti V,[46][47][48][49][50].
In the SVṙ and SSVṙ definitions of authoritative verbal testimony cited above, we find the allusion that Hari, Hara, and Hiraṅyagarbha are not the authors of the Vedas.
This indirect evidence consists of the following: distinguishing the authorities which uttered the Vedas from the authorities to which we owe dharmaśāstras, both commentators call the authors (kāra, kartṛ; in the SVṙ the commentator uses the word kāra, and in the SSVṙ-the word kartṛ) only those latter authorities.

MV 5:
[This is the definition of authoritative verbal testimony presented in the SK:] "And (tu) authoritative verbal testimony (āptavacana) is authorities (āpta) [and] revelation (śruti)". This is the third source of valid knowledge. Authorities are the teachers (ācārya): Brahmā, etc., [and] revelation is the Vedas-these are the two kinds of authoritative verbal testimony…. Thus has been explained the authoritative verbal testimony. 16 Gauḋapāda and Māṫhara interpret āpta-śrutiḥ as authorities (āptāḥ) and revelation (śrutiḥ). They apparently treat āpta-śrutiḥ as a dvandva compoundnotwithstanding the fact that in the SK this word is in the feminine singular, i.e. the form in which dvandva compounds do not occur (in our case if āpta-śruti were a dvandva compound, it would have the plural form āpta-śrutayaḥ). Authorities include Brahmā and other teachers, and revelation is the Vedas. Unlike in the P, SVṙ, and SSVṙ, where all the authoritative utterances are associated with certain authoritative persons, in the GB and MV authorities and revelation represent two separate sources of authoritative utterances. In that way Gauḋapāda and Māṫhara probably wanted to say that the revelation (śruti), i.e. the Vedas, was not composed (created) by authoritative persons, that is, it does not have an author.

The Yuktidīpikā and the Sāṁkhyatattvakaumudī
Both these commentaries, though very different, state directly that the Vedas have no author, treating the authorlessness of the Vedas as the argument for their unquestionable authoritativeness. Below I shall cite those fragments of the YD's and the TK's commentaries to the definition of authoritative verbal testimony which are relevant to the issue of the authorship of the Vedas. YD 5: [It is said in the SK:] "Authoritative verbal testimony (āpta-vacana) is authoritative revelation (āpta-śruti) [and the word of authorities (āpta-śruti)] only (tu)". … Śruti is the revelation (śravaṇa). Authoritative revelation (āpta-śruti) is that revelation (śrutiḥ) which is authoritative (āptā). Or thus [is the meaning of the compound āpta-śruti]. … Āpta-śruti is the word (śruti) [that comes] from authorities (āpta). Āpta-śruti is both the authoritative revelation (āpta-śruti) and the word of authorities (āpta-śruti). [According to the grammatical rule which sounds:] "out of [two or more words] having the same form…", [out of two similar words āpta-śruti and āpta-śruti] one [only] remains (eka-śeṣa). 17 There through mentioning the first āpta-śruti the following is maintained: the Vedas (āmnāya) not created by the intellect of [some] puruṣa 18 (a-puruṣabuddhi-pūrvaka), which are independent (svatantra) and inducing to the highest aim of a human, are the source of valid knowledge that cannot be doubted. Through the second [āpta-śruti] it is established that the word (vacas) of the smṛtis 19 composed by Manu, etc., of the Vedāṅgas, treatises on logic (tarka), Itihāsas, Purāṅas, and of the learned persons versed in various arts, which are not faulty-minded, is the source of valid knowledge. 20 According to the YD, authoritative verbal testimony (āpta-vacana) embraces authoritative revelation (āpta-śruti) and the word of authorities (āpta-śruti). In order to ground his interpretation of the SK definition of authoritative verbal testimony, the author of the YD employs the linguistic trick, treating āptaśrutiḥ as eka-śeṣa (literally, 'one [only] remaining [of two or more stems]', see footnote 17) which stands for the two words: āptaśrutiḥ and āptaśrutiḥ, i.e. 'authoritative revelation' 17 The author of the YD refers to the grammatical rule formulated in sūtra I, 2, 64 of the Aṣṭādhyāyī of Pāṅini. The sūtra is the following: "Out of [two or more words] having the same form one [only] remains (eka-śeṣa), if they have the same inflexion" (sarūpāṇām eka-śeṣa eka-vibhaktau ‖ The Aṣṭādhyāyī of Pāṇini 2000: 133-134). According to this rule, if two or more similar words which have the same ending (these are the words that can be joined by the copulative conjunction 'and') occur, only one of these words remains. The word that remains has its own meaning and the meaning of the omitted word (or words). The grammatical form of the remaining word will be dual or plural (depending on that how many things it stands for). 18 By puruṣa the author means a being which is animated by puruṣa (ātman) that makes this being conscious. 19 In this context the word smṛti, which occurs here in the plural form (smṛtīnām), means dharmaśāstras. 20 āpta-śrutir āpta-vacanaṁ tu ‖ … śravaṇaṁ śrutiḥ | āptā câsau śrutiś ca āpta-śrutiḥ | athavā … | āptebhyaḥ śrutir āpta-śrutir | āpta-śrutiś câpta-śrutiś câpta-śrutiḥ | sarūpāṇām ity eka-śeṣaḥ | tatra pūrveṇâpta-śruti-grahaṇenêdaṁ pratipādayati | apuruṣa-buddhi-pūrvaka āmnāyaḥ sva-tantraḥ puruṣa-niḥśreyasârthaṁ pravartamāno niḥsaṁśayaṁ pramāṇam iti | dvitīyena manv-ādi-nibandhānāṁ ca smṛtīnāṁ vedāṅga-tarkêtihāsa-purāṇānāṁ śiṣṭānāṁ nānā-śilpâbhiyuktānāṁ câduṣṭa-manasāṁ yad vacas tat pramāṇam ity etat siddhaṁ bhavati | (Yuktidīpikā 1998, p. 87, v. 1-12.) The YD gives more extensive explanations of most of the kārikās than other commentaries of the classical Sāṁ khya, that is why in my references to the YD, besides the number of the kārikā, I indicate the page and the verse number of the edition of the YD. and 'the word of authorities'. Āptaśrutiḥ occurring in the SK cannot be interpreted as eka-śeṣa-for the reason that it is used in the singular form. If it were eka-śeṣa meaning 'āpta-śruti and āpta-śruti', it would have the dual form āpta-śrutī. However, the interpretation of āpta-śrutiḥ as both authoritative revelation and the word of authorities is not impossible, for we cannot exclude that in the SK the word āpta-śruti is used in a double sense.
It is easy to notice the following similarity between the YD and the two commentaries we analyzed above. In the YD, like in the GB and MV, authoritative utterances originate from authorities and from the revelation (i.e. the Vedas) which is not associated with any authorities. Treating authoritative revelation and the word of authoritative persons as two separate sources of authoritative utterances, the author of the YD distinguishes in that way the utterances composed by certain authors from the utterances of the authorless Vedas. Moreover, the YD presents the direct evidence in the question of the authorship of the Vedas: the Vedas are said to be "not created by the intellect of [some] puruṣa" (a-puruṣa-buddhi-pūrvaka), i.e. by the intellect of some conscious being. It means that the Vedas have no author.
So, in the YD it is asserted directly that the Vedas have no author. It is very likely that in the YD, like in Mīmāṁ sā, the authorlessness of the Vedas represents the argument for their unquestionable authoritativeness. In the same sentence of the YD where the Vedas are said to be without an author we find also other characteristics applied to the Vedas by the Mīmāṁ sakas, namely, their being independent (svatantra), their being "inducing to the highest aim of a human" (puruṣaniḥśreyasârtham pravartamānaḥ), and their possessing the validity that even cannot be doubted. By the independence of the Vedas both in Mīmāṁ sā and Sāṁ khya there can be understood, first, their independence from any author, second, independence of their scope, or functional sphere, that also means their independence from other sources of valid knowledge. These two independences-independence from any author and independence of the scope-represent in Mīmāṁ sā two main arguments for the unquestionable authoritativeness of the Vedas. Calling the Vedas "the source of valid knowledge that cannot be doubted" (niḥsaṁśayaṁ pramāṇam), the author of the YD asserts this unquestionable authoritativeness of the Vedas. According to Mīmāṁ sā, validity of the Vedic utterances can never be doubted for the following reasons: first, they have no author that can speak untruth, second, they can never be refuted because of the independence of their scope (in other words, they are unfalsifiable by other pramāṇas, for imperceptible things which constitute the domain of authoritative verbal testimony are incognizable by other pramāṇas)-see Ś abara's (350-400 CE) 21 Śābarabhāṣya I, 1, 2; I, 1, 5. Taking into account the fact that the context of the fragment of the YD where the Vedas are said to be without an author is influenced by Mīmāṁ sā, and the fact that the authorlessness of the Vedas is connected with their undoubtable reliability, we can conclude that it is very likely that in the YD, like in Mīmāṁ sā, the authorlessness of the Vedas is regarded as an argument for their unquestionable authoritativeness.
The commentary to the definition of authoritative verbal testimony given in TK 5 is also influenced by Mīmāṁ sā. I shall now cite the relevant passage of this commentary: … And that [cognition through authoritative verbal testimony] is intrinsically valid (svataḥ-pramāṇa). It is true (yukta), because it is entirely free from suspicion of [any] defectiveness inasmuch as it is born by sentences of the authorless (apauruṣeya) Vedas. Thus the cognition born by sentences of the smṛtis, 22 Itihāsas, and Purāṅas, which are rooted in the Vedas (veda-mūla), is also true (yukta). And the primeval sage Kapila at the beginning of the kalpa remembers śruti studied [by him] during the [previous] kalpa(s)… 23 Vācaspati Miśra states directly that the Vedas have no author-by calling them apauruṣeya, which means literally 'not coming from puruṣa". The authorlessness of the Vedas is used here, like in Mīmāṁ sā, as the argument for their unquestionable authoritativeness. No doubt is possible about knowledge generated by sentences of the Vedas-for they have no author whose reliability can be doubted. A doubt about the validity of Vedic sentences cannot arise-for there is no reason for it. It is obvious that Vācaspati's commentary is influenced by Mīmāṁ sā. In the small fragment cited above we find the following ideas which were systematically developed in Mīmāṁ sā: the idea of intrinsic validity of cognition through authoritative verbal testimony; the idea of the authorlessness of the Vedas regarded as the argument for their unquestionable authoritativeness; the idea that authoritativeness of those texts which have an author lies in their being based on the Vedas.
In the TK there is one more instance of direct evidence in the question of the authorship of the Vedas. In TK 2 Vācaspati says that the Vedas are only transmitted from teacher to pupil, "but they are not created by anybody" (na tu kenacit kriyate). Vācaspati states here directly that the Vedas have no author. In that fragment also we see the influence of Mīmāṁ sā which claims that the Vedas were not created by some author, human or divine, but came to us through unbroken tradition of the Vedic recitation.
But how is it possible to fit this Mīmāṁ sā view on the Vedas in the doctrine of Sāṁ khya which, unlike Mīmāṁ sā, accepts the conception of periodic dissolutions (pralaya) of the world? How is it possible to introduce the idea of the unbroken tradition of the Vedic recitation into the Sāṁ khya cosmological framework? According to Mīmāṁ sā, both the Vedas and the world have no beginning in time. They have always been. They never arose. The present-day Vedic teachers have heard the Vedas from their teachers, and their teachers also had teachers from whom they learned the Vedas, and so on-without beginning. According to Sāṁ khya, all the world, except puruṣa and prakṛti which are eternal, undergo destruction (i.e. dissolution in prakṛti) during pralaya, emerging from prakṛti again at the beginning of a new kalpa. The Vedas also undergo dissolution during pralaya, and the tradition of their recitation breaks. 22 By smṛtis Vācaspati means dharmaśāstras. 23 … tac ca svataḥ-pramāṇam apauruṣeya-veda-vākya-janitatvena sakala-doṣâśaṅkā-vinirmuktatvena yuktaṁ bhavati evaṁ veda-mūla-smṛtîtihāsa-purāṇa-vākya-janitam api jñānaṁ yuktam ādi-viduṣaś ca kapilasya kalpâdau kalpântarâdhīta-śruti-smaraṇa-sambhavaḥ ….
Vācaspati Miśra seems to be aware of the difficulties of fitting the Mīmāṁ sā view on the Vedas into the Sāṁ khya doctrine. How does he solve these difficulties? We can reconstruct his solution on the basis of the following passage: "And the primeval sage Kapila at the beginning of the kalpa remembers śruti studied [by him] during the [previous] kalpa(s)…" (see TK 5 cited above). By saying that Kapila at the beginning of the kalpa remembers the Vedas studied by him before the time of dissolution (pralaya), Vācaspati tries to preserve the continuity of the Vedic recitation before and after pralaya. From this passage it is clear that the Vedas of the new cycle of the existence of the world are similar to the Vedas before pralaya, and that there are persons (or a person, namely, Kapila, who is identified by Sāṁ khya as its founder) with extraordinary capacities who are able to remember the Vedas studied before pralaya and, in that way, to recommence the tradition of the Vedic recitation interrupted by pralaya.

The Jayamaṅgalā
The JM stands out against the background of most Sāṁ khya texts (i.e. the texts containing suggestions or direct statements that the Vedas are without an author), for it follows from its commentary to the definition of authoritative verbal testimony that the Vedas have an author or authors. Below I shall cite the relevant fragment of this commentary. JM 5: "And (ca) authoritative verbal testimony is the word of authority"-thus [it is said in the SK]…. That word ( Unlike most Sāṁ khya commentators, the present commentator speaks only about one type of authoritative verbal testimony, namely, the word of authoritative persons. The author of the JM does not draw a distinction between Vedic utterances and other authoritative utterances. He holds that all the authoritative utterances are based on perception or inference of certain objects by trustworthy persons (according to Sāṁ khya theory of the scope of the third pramāṇa these objects are the objects reachable neither by perception of ordinary people nor by inference which rests upon such a perception). It follows from this that the Vedic utterances also have an author or authors. It is not clear if the Vedas have one author (for example, Kapila) or many authors (for example, different ṛṣis). It is not clear, also, 24 āpta-śrutir āpta-vacanaṁ cêti | (…) ‖ āptebhyo yā śruti-paramparayā śrutir āgatā tad āpta-vacanam tair dṛṣṭo'numito vârthaḥ paratra svabodha-sadṛśa-bodhântarôtpattaye śabdenôpadiśyate | yad āpta-vacanaṁ tan na plavate | Authorship of the Vedas 463 whether the capacity of the direct cognition of unseen reality of this author/these authors is limited. It is said that authoritative persons base their reliable statements both on perception and inference, from which it follows that all of these authorities or some of them are not capable of perceiving all the objects. If perception of each authoritative person embraced all the objects, there would be no mention of acquiring knowledge by inference. It is not specified in the JM if all the authorities obtain knowledge by inference also or it refers only to those authorities which are not the authors of the Vedas; so it is not clear whether the authority/authorities which composed the Vedas based their trustworthy utterances not only on perception but also on inference. It is difficult not to notice a close resemblance between the above-cited definition of authoritative verbal testimony of the JM and the definition of the Yogasūtrabhāṣya (ca. 500-700 CE (?)), given in its commentary on sūtra I, 1, 7 of the Yogasūtra (though the author of the JM uses for authoritative verbal testimony the Sāṁ khya term āpta-vacana and not the Yogic term āgama we find in the Yogasūtrabhāṣya). 25 I would like to emphasize that the above-cited fragment of the JM was interpreted in the context of the JM and Sāṁ khya (and not in the context of Yoga). Unlike Sāṁ khya (see kārikā 6 of the SK and the commentaries to this kārikā), Yoga (see such texts as Yogasūtras, Yogasūtrabhāṣya, Tattvavaiśāradī (ca. 841 CE or ca. 976 CE) of Vācaspati Miśra, and Rājamārtaṇḍa (ca. 1150 CE) by Bhoja Rāja) does not state that authoritative verbal testimony has independent scope including only those things which cannot be validly cognized by perception or inference of ordinary people (i.e. people not endowed with an extraordinary capacity of perception of imperceptible reality). Hence, in Yoga authoritative verbal testimony is a source of valid knowledge of those things also which are reachable by perception and inference of ordinary people, and inference mentioned in the definition of āgama given in the Yogasūtrabhāṣya is probably a means of cognition of those authorities which are ordinary people (and not of those authorities which are capable of direct cognition of unseen reality). It follows from this that in the context of Yoga (unlike in the context of Sāṁ khya) this mention about an inferential cognition of authoritative persons does not necessarily lead to the conclusion that all or some of the authorities which communicate about imperceptible reality cognize this reality not only by means of perception but also by means of inference. 25 Cf. Yogasūtrabhāṣya I, 1, 7: āptena dṛṣṭo'numito vârthaḥ paratra sva-bodha-saṁkrāntaye śabdenôpadiśyate śabdāt tad-arthaviṣayā vṛttiḥ śrotur āgamaḥ | yasyâśraddheyârtho vaktā na dṛṣṭânumitârthaḥ sa āgamaḥ plavate | mūla-vaktari tu dṛṣṭânumitârthe nirviplavaḥ syāt ‖ An object which has been perceived or inferred by authoritative person (

Conclusion and Perspectives for Future Research
The aim of this paper was to reconstruct the classical Sāṁ khya view on the authorship of the Vedas. The research was based on the commentaries to the SK definition of authoritative verbal testimony (given in kārikā 5) which contain the main evidence (both direct and indirect) for reconstruction of this view. Besides the fragments where the classical Sāṁ khya writers explain the SK definition of authoritative verbal testimony, the fragment of TK 2 containing the direct evidence in the question of the authorship of the Vedas was considered. In that way all the direct evidence I had found in the texts of classical Sāṁ khya (it seems not to be numerous, formulated three times only-in YD 5, TK 5, and TK 2) was taken into account.
The analysis presented in this article leads to the following conclusion. According to most classical Sāṁ khya commentaries, the Vedas have no author. Such conclusion is supported by all the direct evidence we find in the classical Sāṁ khya texts and by indirect evidence given in five classical Sāṁ khya commentaries to the definition of authoritative verbal testimony. In YD 5, TK 5, and TK 2 the Sāṁ khya writers state directly that the Vedas have no author. The fragments containing the direct evidence are influenced by Mīmāṁ sā. Vācaspati Miśra in the TK, trying to fit the Mīmāṁ sā idea of the uninterrupted tradition of recitation of the authorless Vedas in the doctrine of Sāṁ khya which accepts the conception of periodic dissolutions (pralaya) of the world, suggests, first, that the Vedas after pralaya are similar to the Vedas before pralaya, second, that there are persons (or a person, i.e. Kapila) capable of remembering the Vedas existing before pralaya and recommencing in that way the tradition of their recitation interrupted by pralaya. As regards the indirect evidence, it is as follows: first, in SVṙ 5 and SSVṙ 5 the commentators, distinguishing between authorities who uttered the Vedas and authorities to whom we owe dharmaśāstras, call the authors (kāra, kārtṛ) only the latter authorities; second, in GB 5, MV 5, and YD 5 utterances of authoritative persons and utterances of the Vedas, which are not associated with any authorities, represent two separate types of authoritative utterances.
The only commentary which seems to suggest that the Vedas have an author (or authors) is the JM. According to the JM definition of authoritative verbal testimony, which is influenced by Yoga, all the authoritative utterances are based on perception or inference of certain objects by authoritative persons (see JM 5). It follows from this that the Vedas too have an author or authors.
Future research perspectives include, first of all, clearing up whether the classical Sāṁ khya texts contain other indirect evidence in the question of the authorship of the Vedas. In order to clear it up we should examine carefully all the fragments where there are mentions of the Vedas, authoritative utterances, Kapila, ṛṣis, and the like persons. After that we should study pre-Īśvarakṙṡṅa Sāṁ khya material and postclassical Sāṁ khya texts, as well as external evidence, i.e. the evidence presented in the texts which do not belong to Sāṁ khya tradition.

Authorship of the Vedas 465
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.