² There is no scholarly consensus on the dating of the Nirukta. Kahrs briefly reviews previous views and concludes that “[m]y own inclination is that Yāska's date falls within the later period of a possible timespan between the seventh and third centuries bce”. Kahrs, E., Indian Semantic Analysis: The nirvacana Tradition (Cambridge, 1998), pp. 13–14Google Scholar. More recently, Scharf has proposed that the Nirukta is a multilayered text and that different parts of the texts were composed at different stages; specifically, he regards the first portion of the text (Nirukta 1.1-2.4), together with Chapters 13 and 14, as later additions. Scharf, P., ‘Linguistics in India’, in The Oxford Handbook of the History of Linguistics, (ed.) Allen, K. (Oxford, 2013), p. 229Google Scholar. Scharf, P., ‘The Relation between Etymology and Grammar in the Linguistic Traditions of Early India’, Bulletin d’ Études Indiennes 32 (2014), p. 259Google Scholar.

³ For a fairly comprehensive, yet by no means exhaustive, overview of modern scholarship, see Deeg, M., Die altindische Etymologie nach dem Verständnis Yāska's und seiner Vorgäger: Eine Untersuchung über ihre Praktiken, ihre literarische Verbreitung und ihr Verhältnis zur dichterischen Gestaltung und Sprachmagie (Dettelbach, 1995), pp. 67–73Google Scholar, and Visigalli, P., ‘An Early Indian Interpretive Puzzle: Vedic Etymologies as a Tool for Thinking’, Journal of Indian Philosophy 46 (2018), pp. 985–988CrossRefGoogle Scholar.

⁴ Sarup, L., The Nighaṇṭu and the Nirukta, the Oldest Indian Treatise on Etymology, Philology, and Semantics. English Translation (London and New York, 1921)Google Scholar.

⁵ Roth, R., Jâska's Nirukta sammt den Nighaṇṭavas (Göttingen, 1852)Google Scholar.

⁶ Pathak, J., Nirukta of Yāskācārya: Edited with ‘Śaśiprabhā’ Hindi Commentary and Notes (Vanarasi, 2018 [2010])Google Scholar.

We translate the text established by Sarup in his critical edition. In two occasions, we choose a different reading; see below note 57 (pravahlitam > pravalhitam) and note 101 (aharad > āharad). Our references to the Nighaṇṭu, too, follow Sarup's edition. Sarup, L., The Nighaṇṭu and the Nirukta, the Oldest Indian Treatise on Etymology, Philology, and Semantics. Sanskrit Text, with an Appendix Showing the Relation of the Nirukta with Other Sanskrit Works (Lahore, 1927)Google Scholar.

⁷ In this paper, ‘sky’ renders Sanskrit dyu. (We use ‘celestial’ as its corresponding adjective.) We chose this rendition over the usual ‘heaven’ because we think that the latter is partly misleading and does not fit well with Yāska's theology. ‘Heaven’ is a religiously loaded term, whereas ‘sky’ denotes a spatial/cosmological locus. ‘Heaven’ conjures up earth, thereby suggesting an opposition in which only heaven is predominantly associated with the supernatural. In Yāska's theology, however, all the three worlds (earth, mid-space, and sky) are cosmological loci that are equally associated with their corresponding deities (see Part I: §6.1; §8).

⁸ We use italics to refer to the word (e.g. agni) and capitalised roman to refer to the deity (e.g. Agni).

⁹ Yāska uses the following terminology: (a) ‘primary praise’ (prādhānyastuti: Nirukta 1.20; 7.1, 13; prādhānyena: Nirukta 1.20); (b) ‘incidental mention’ (naighaṇṭuka: e.g. Nirukta 1.20; and nipāta: e.g. Nirukta 7.18). For the sake of clarity, we render ‘incidental mention’ as ‘secondary praise’, to better match ‘primary praise’. Yāska uses two different terms (naighaṇṭuka/nipāta) to refer to ‘secondary praise’. He defines naighaṇṭuka in Nirukta 1.20: ‘with respect to this (i.e. the twofold classification of divine names into primary [prādhanyena] and incidental [naighaṇṭuka]), whatever [divine name, devatānāma] drops in a formula [addressed to] another deity, that is incidental (naighaṇṭuka)’ (Nirukta 1.20: tad yad anyadevate [variant reading: anyadaivate] mantre nipatati naighaṇṭukam tat).

Note that the same term naighaṇṭuka (yet in the plural naighaṇṭukāni [nāmāni], rather than in the singular) is used in Nirukta 1.20 in a different sense, to refer to the first section (=first three chapters) of the Nighaṇṭu word list. This section consists of synonyms: e.g. Nighaṇṭu 1.1 lists twenty-one words, all of which mean ‘earth’.

¹⁰ Two examples will help clarify how this twofold classification works. In Nirukta 1.20, immediately after having defined naighaṇṭuka (see note 9), Yāska cites R̥gveda I.27.1. In this verse, two divine names are mentioned, agni [Nighaṇṭu 5.1.1] and aśva ‘horse’ [Nighaṇṭu 5.3.1]. Yet, aśva is mentioned only incidentally, for R̥gveda I.27.1 is addressed to Agni. Thus, while Agni enjoys primary praise through agni, Aśva enjoys only secondary praise through aśva.

In Nirukta 12.20-21, Yāska discusses the divine name viśvānara [Nighaṇṭu 5.6.12]. He says: ‘[The name] viśvānara has been already explained [in Nirukta 7.21; below Part IV: §21]. viśvānara has an incidental mention (=Viśvānara is incidentally mentioned) in [the following] verse addressed to Indra [R̥gveda VIII.68.4]’ (viśvānaro vyākhyātaḥ | tasyaiṣa nipāto bhavaty aindryām r̥ci). This means that the mention of viśvānara in R̥gveda VIII.68.4 is incidental, for the hymn is addressed to Indra. Thus, while Viśvānara enjoys only secondary praise through viśvānara, Indra enjoys primary praise through indra, which occurs several times in the hymn.

For the R̥gveda, we use the following edition: Aufrecht, T., Die Hymnen des Ṛigveda, 2 vols (Bonn, 1877)Google Scholar.

¹¹ Take as an example the word aśva. As we just saw, it is mentioned incidentally in R̥gveda I.27.1; yet, it refers to the deity that enjoys primary praise in R̥gveda IX.112.4; see Nirukta 9.2.

¹² This can be deduced from Nirukta 7.1, in which the ‘section on the names of the deities’ (daivata [= Nighaṇṭu 5]) is defined as comprising names with which the deities enjoy primary praise; see below [I: §1]. More clearly, the point is expressly stated by Yāska, when he explains the rationale behind the arrangement of Nighaṇṭu 5; see below [I: §10].

¹³ Durga (see R̥jvarthā 7.4 [626.10-15]) gives the following example: formulas whose divine name is not specified that are used in the Agniṣṭoma should be considered as addressed to Agni, for the Agniṣṭoma is associated with Agni. However, when such formulas are recited during the first, second, and third soma pressing of the Agniṣṭoma, they should be taken as being addressed to Agni, Indra, and Āditya, respectively; for each of the three pressings is associated respectively with one of these three deities.

For the R̥jvarthā, we use the following edition: Rajavade, V. K., Durgācarya's Commentary on the Nirukta, Durgācāryakr̥tavr̥ttisametam Niruktam (Poona, 1926)Google Scholar.

¹⁴ Prajāpati is the main deity for the ritualists. As noted by Durga (R̥jvarthā 7.4 [627.1-3]), the affinity between the formulas whose deity is not specified and the deity Prajāpati consists in the fact that both are ‘undefined’ (anirukta). See Renou and Silburn's study of the terms nirukta/anirukta in the Brāhmaṇas. L. Renou and L. Silburn, ‘Nírukta and Ánirukta in Vedic’, in Sarūpa-Bhāratī: The Homage of Indology (Dr. Laksman Sarup Memorial Volume), (eds.) J. Agrawal and Bhim Dev Shastri (Hoshiarpur, 1954), pp. 68–79. See also Visigalli, P., ‘The Vedic Background of Yāska's Nirukta’, Indo-Iranian Journal 60 (2017), p. 112fCrossRefGoogle Scholar. Durga (R̥jvarthā 7.4 [627.3-12]) explains Narāśaṃsa as an epithet of Agni, the main deity in the etymologists’ pantheon. Cf. below [I: §15], where Agni is said to be all the deities.

¹⁵ See note 6.

¹⁶ The meaning of Yāska's terse formulation (‘x’ kasmāt?) has been debated in scholarship, for it is connected with an overall interpretation of Yāska's etymological project: are Yāska's etymologies best understood as derivational (kasmāt = ‘from where?’, i.e. from what root?) or causal (kasmāt = ‘why?’, i.e. why is something called ‘x’? e.g. ‘why is Agni called agni?’) explanations? In our view, Yāska's etymologies are concerned with both derivational-grammatical and causal-semantic aspects; while Yāska's primary concern is to elucidate the semantic content of a word, doing so involves providing a derivational-grammatical analysis of that word. On the meaning of kasmāt and its relation to an overall interpretation of the Nirukta, see Kahrs's and Scharf's studies. Kahrs, E., ‘Yāska's Use of kasmāt’, Indo-Iranian Journal 25 (1983), pp. 231–237CrossRefGoogle Scholar. Kahrs, Indian Semantic Analysis. P. Scharf, ‘The Natural-language Foundation of Metalinguistic Case-use in the Aṣṭādhyāyī and Nirukta’, in Papers of the 12th World Sanskrit Conference. Vol. 4, Indian grammars: philology and history, (eds.) George Cardona and Madhav Deshpande (Delhi, 2012), pp. 181–214. Scharf, ‘The Relation between Etymology and Grammar in the Linguistic Traditions of Early India’, pp. 255–266.

¹⁷ Of the three ritual fires, it is the āhavanīya or offertorial fire which Yāska has in mind. This is clear from his explanation of Agni's main actions [I: §8.1]: carrying the oblations to the gods and inviting the gods to the ritual place.

¹⁸ In Part II and III, (iii) is introduced with the same formulation: ‘One should not think that [the word] agni (III: agni [jātavedas]) [refers to] this (terrestrial fire) only. Also the well-known two upper lights (i.e. the lightning and the sun) are called agni (III: jātavedas)’: sa na manyetāyam evāgnir iti | apy ete uttare jyotiṣī agnī (III: jātavedasī) ucyete |

¹⁹ What are these two parties? In Part IV, the first view is ascribed to the ‘teachers’ (ācāryas) [IV: §23.1], and the second view to the ‘previous ritualists’ (pūrve yājñikāḥ) [IV: §23.2]. Commentators gloss ‘teachers’ with ‘etymologists’ (R̥jvarthā on Nirukta 7.22 [697.21]: nairuktāḥ ke cit ācāryāḥ; Niruktabhāṣyaṭīkā on Nirukta 7.22 [88.5]: pūrve nairuktā). It is possible, though not certain, that the two competing views mentioned in Part II and III, too, may be ascribed to the same two parties.

For the Niruktabhāṣyaṭīkā, we use the following edition. Sarup, L., Commentary of Skandasvāmin & Maheśvara on the Nirukta [Chpaters VII-XIII] Critically Edited by Dr. Lakshman Sarup with Additions and Corrections by Acharya V. P. Limaye (New Delhi, 2012)Google Scholar. [Originally published in three volumes, (Lahore, 1928–34)].

²⁰ yas tu sūktaṃ bhajate yasmai havir nirupyate ’yam eva so ’gnir [III: agnir jātavedāḥ; IV: agnir vaiśvānaraḥ] | nipātam evaite uttare jyotiṣī etena nāmadheyena bhajete ||

²¹ For the Kauśītaki-Brāhmaṇa, we use the following edition. Lindner, B., Das Kaushîtaki Brâhmaṇa, Text, I. (Jena, 1887)Google Scholar.

²² For the Aitareya-Brāhmaṇa, we use the following edition: T. Aufrecht, Das Aitareya Brāhmaṇa: Mit Auszügen aus dem Commentare von Sāyaṇācārya und anderen Beilagen herausgegeben (Bonn, 1879).

²³ Note that this is the third time R̥gveda IV.58 is cited; see above [§14.1] (R̥gveda IV.58.8) and [§14.2] (R̥gveda IV.58.1a). It is worth noticing that different views regarding the deity of this hymn (Agni, Sun, the Waters, Cows, or Ghee) are recorded in the Sarvānukramaṇī (see S. W. Jamison and J. P. Brereton, The Rigveda: The Earliest Religious Poetry of India, 3 vols [New York, 2014], p. 643) and the Br̥haddevatā (V.11). For the Br̥haddevatā, we use the following edition. Tokunaga, M., The Br̥haddevatā: Text Reconstructed from the Manuscripts of the Shorter Recension with Introduction, Explanatory Notes, and Indices (Kyoto, 1998)Google Scholar.

²⁴ For the Maitrāyaṇī-Sam̐hitā, we use the following edition: L. von Schroeder, Maitrâyaṇî Saṁhitâ, 4 bde (Leipzig, 1881–6), reprint (Wiesbaden, 1970–2).

²⁵ Durga [R̥jvarthā 603.1-7] explains that the ritualists have in mind R̥gveda X.88.12 (X.88.12b: vaiśvānaráṃ ketúm áhnām akr̥ṇvan, ‘[the gods] made Agni Vaiśvānara the beacon of the days’). He glosses ketu with kartr̥ ‘maker’. It is only the sun that is the maker of the day, in the sense that it creates the days with its own rising and setting; hence, the name vaiśvānara must refer to the sun.

²⁶ It is not clear whether the four arguments recorded in [§23.3.1-4] are by Śākapūṇi or by Yāska. We prefer the former option. Yāska's own voice seems to first occur in [§24], whereby the first view [§23.1] regarding the referent of vaiśvānara is countered. Note that Yāska would probably agree with Śākapūṇi's arguments in [§23.3.1-4], for both believe that vaiśvānara refers to the terrestrial fire. However, while Śākapūṇi maintains that vaiśvānara refers to the terrestrial fire only, Yāska says that it also refers to the other two fires [§28].

²⁷ On Yāska's identification of Indra with lightning, see Kawamura's recent article. Kawamura, Y., ‘On the Name and Role of Indra: From the Viewpoint of Yāska's Etymology and Theology’, The Hiroshima University Studies, Graduate School of Letters 79 (2019), pp. 15–28Google Scholar. This article is written in Japanese with English summary.

²⁸ For possible interpretations of the sentence-introducing tad (‘so’ in our rendition), cf. Cardona's comments on Nirukta 1.1 tad yāny (etāni) catvāri padajātāni . . . . Cardona, G., ‘Philology, Text History and History of Ideas’, Nagoya Studies in Indian Culture and Buddhism: Saṃbhāṣā 35 (2019), p. 2Google Scholar, note 5. Cf. also J. Bronkhorst, ‘Yāska and the Sentence: the Beginning of śābdabodha?’, in Subhāṣiṇī: Dr. Saroja Bhate Felicitation Volume, (ed.) G. U. Thite (Pune, 2002), pp. 57–58.

To properly parse Nirukta 7.1 (=1.20) (tad yāni nāmāni prādhānyastutīnāṃ devatānāṃ tad daivatam ity ācakṣate), one needs to recognise that the direct object pronoun tad has nāmāni as its antecedent, but it agrees in gender and number with the object predicate (daivata n.) governed by ā √cakṣ (*tāni [nāmāni] > tad). In the following sentence (‘That section is . . .’) (saiṣā devatopaparīkṣā), a similar agreement obtains between subject (*tad [daivatam] > sā) and subject predicate (devatopaparikṣā f.). Both kinds of agreement (between object and object predicate, and between subject and subject predicate) of the ta-pronoun are regular in Vedic prose. See Brereton, J. P., ‘‘Tat Tvam Asi’ in context’, Zeitschrift der Deustchen Morgenländischen Gesellschaft 136 (1986), pp. 99–101Google Scholar and Brereton, J. P., ‘Unsounded speech: Problems in the interpretation of BU(M) 1.5.10 = BU(K) 1.5.3’, Indo-Iranian Journal 31 (1988), p. 3, note 10Google Scholar.

²⁹ We take yat(kāma) to refer to ārtha(patyam). Cf. Roth's rendering ‘. . . diejenige Gottheit, welche der Rischi um den Besitz irgend einer Sache, die er zu erhalten wünscht, anruft . . .’ [emphasis added] and Macdonell's ‘the formula has that god for its deity to whom he addresses praise when desiring the possession of an object which he wants’ (emphasis added). Roth, Jâska's Nirukta sammt den Nighaṇṭavas, p. 100 (Erläuterungen) and A. A. Macdonell, The Br̥had-devatā Attributed to Śaunaka: A Summary of the Deities and Myths of the Rigveda, Critically Edited in the Original Sanskrit with an Introduction and Seven Appendices, and Translated into English with Critical and Illustrative Notes. Part I, introduction and text and appendices (Cambridge, 1904), p. 2 (ad Br̥haddevatā I.6). We take ‘[verse of] praise’ (stuti f.) as the antecedent of the pronoun sa, which agrees in gender with the subject predicate mantra (m.) (*sā > sa).

Yāska's definition of the deity of the formula is echoed in Br̥haddevatā I.6: artham icchann r̥ṣir devaṃ yaṃ yaṃ āhāyam astv iti | prādhānyena stuvañ chaktyā mantras taddeva eva sah. || As noted by Tokunaga (M. Tokunaga, The Br̥haddevatā [Kyoto, 1998], p. 159), this definition is reminiscent of R̥gveda X.121.10: prájāpate ná tvád etā́n_iy anyó víśvā jātā́ni pári tā́ babhūva | yátkāmās te juhumás tán no astu vayáṃ s_iyāma pátayo rayīṇā́m || (“O Prajāpati! No one other than you has encompassed all these things that have been born. Let what we desire as we make oblation to you be ours. We would be lords of riches.”) This translation is by Jamison and Brereton, The Rigveda, p. 1594.

³⁰ Nirukta 7.1 (132.5) tās trividhā r̥caḥ. We take tās as a pronoun having the word mantra as its antecedent; it agrees in gender and number with the predicate (r̥cas) (*saḥ [mantra] > tās).

³¹ The translation of the R̥gvedic verses follows Jamison and Brereton's work mentioned above, with minor variations, unless Yāska comments on the verses. In the latter case, our translation reflects Yāska's interpretation.

³² This verse is untraced. Durga (R̥jvarthā 1.2 [614.10-11]) cites the entire verse: indre kāmā ayaṃsata divyāsaḥ pārthivā uta ∣ tyam ū ṣu gr̥ṇatā narah ‖ (‘‘On Indra the desires were based, the celestial as well as the terrestrial. O people do praise him well’’) cf. Bloomfield, M., A Vedic Concordance (Cambridge, 1906), p. 225Google Scholar.

³³ Yāska cites only the portions of the verses that contain the seers that are addressed directly, i.e. with the second person. The remaining portions of the verses contain the objects of praise (R̥gveda VIII.1.1: Indra; R̥gveda I.37.1: the Maruts; R̥gveda III.53.11: king Sudās) that are addressed indirectly.

³⁴ On such ‘hymns of self-praise’ (ātmastuti), see G. Thompson, ‘Ahaṃkāra and Ātmastuti: Self-Assertion and Impersonation in the Ṛgveda’, History of Religions 37, 2 (1997), pp. 141–171.

³⁵ Almost the same passage is found in Kāṭhaka-Āraṇyaka III.I.214. For the Kāṭhaka-Āraṇyaka, we use the following edition. Witzel, M., Kaṭha Āraṇyaka: Critical Edition with a Translation into German and an Introduction (Cambridge, Mass., 2005)Google Scholar.

³⁶ We take the ablative kasmāc cit bhāvāt as the object of dislike (cf. vārttika 1 on Aṣṭādhyāyī 1.4.24: jugpsāvirāmapramādārthānām upasaṃkhyānam) governed by paridevanā. If it is taken as a causal ablative, then the translation would be ‘on account of a certain state’.

For the Vārttika and the Aṣṭādhyāyī, we use the following editions respectively. K. V. Abhyankar, The Vyākaraṇa-mahābhāṣya of Patañjali: Edited by F. Kielhorn, 3 vols (Bombay, 1880–5), Third edition, revised and furnished with additional readings, references and select critical notes by K. V. Abhyankar, 3 vols (Poona: 1962–72) and Appendix III (Aṣṭādhyāyīsūtrapāṭha) in Cardona, G., Pāṇini: His Work and Its Traditions, Volume One, Background and Introduction (Delhi, 1988)Google Scholar, Second edition, revised and enlarged (1997).

³⁷ Durga identifies two major kinds of formulas that are not employed in ritual: (a) formulas whose original ritual use had been lost (utsanna) (R̥jvarthā 7.4 [626.16-23]); (b) formulas that are used in non-ritual performances such as the recitation of the Veda (svādhyāya) and atonement practices (630.1-2).

³⁸ Durga (R̥jvarthā 627.3-12) explains Narāsaṃśa (Nighaṇṭu 5.2.4; cf. also nārāśaṃsaḥ Nighaṇṭu 5.3.6) as an epithet of Agni, the main deity in the etymologists’ pantheon (secondarily, it is also identified with Yajña, an epithet of Viṣṇu, and with Sūrya [630.11]). According to Findly, Narāśaṃsa is the personification of the priests’ (narā < *nárām is an old subjective genitive plural of nr̥ ‘man’, i.e. the poet-priests) praises (śaṃsa) for the deities, especially for Indra. Due to their poetic eloquence such praises were expected to satisfy the deities and thereby make ritual successful. E. B. Findly, Aspects of Agni: Functions of the Ṛgvedic Fire (unpublished PhD dissertation, Yale University, 1978), p. 174f.

³⁹ api vā sā kāmadevatā syāt. We take sā as the subject; it refers to the implied word devatā, ‘deity’.

⁴⁰ Formulas whose deities are not specified are divided into two main groups, (i) those that are employed in ritual and (ii) those that are not. Formulas (i) have as their deity the deity of (part of) the ritual in which they are employed. With respect to (ii), three views are given. Such formulas have as their deity: (a) Prajāpati, for the ritualists; (b) Narāśaṃsa (=Agni), for the etymologists; (c) kāmadevatā ‘desire deity’, i.e. one can choose the addressee deity in accordance with one's own will; (c¹) prāyodevatā ‘general/various (?) deity’.

We take (c) and (c¹) as forming the third view, which is illustrated by the concluding passage (asti hy ācāro . . .): the practice in the world is various (bahula), i.e. one and the same formula has sometimes a god as its deity, sometimes the guests, sometimes the ancestors. We tentatively take the problematic prāyo(devatā) to mean something like bahula.

⁴¹ The phrase yājñadaivato mantra iti (Nirukta 7.4) is problematic. Our interpretation is, therefore, provisional. While this phrase must refer to one or more views that have been mentioned above, it is unclear what view(s) is(are) meant exactly. We assume that yājñadaivato mantra gives a shorthand summary of all the classificatory cases for formulas that have been dealt with so far. We analyse yājñadaivata as follows. ‘[The deity] of the ritual’ (yājña) refers to view [§4.1], i.e. formulas whose deity is not specified have as their deity the deity for which (part of) the ritual is performed. ‘[Another] deity’ (daivata) refers to all the other possible cases: the name of the addressee deity is either (a) explicitly mentioned, (b) or it is not specified; (b) includes the views discussed in [§4.2] and [§4.3].

⁴² The main thrust of the objection seems this: the names listed in Nighaṇṭu 5.3.1-22, such as ‘horse’ (aśva: Nighaṇṭu 5.3.1) and ‘herb’ (oṣadhi: Nighaṇṭu 5.3.1), as well as the eight dyads listed in Nighaṇṭu 5.3.29-36, such as ‘mortar-pestle’ (ulūkhalamusale: Nighaṇṭu 5.3.29) and ‘oblation-receptacle’ (havirdhāne: Nighaṇṭu 5.3.30) cannot be divine names. For horse etc. and mortar-pestle etc. are not deities.

Durga (R̥jvarthā on Nirukta 7.4 [630.17-631.1]) explains that the deity of a formula must be able to fulfil men's desires by reciprocating their praises (see the definition of the deity of the formula in §2). Now, how can a horse or an herb understand men's praises, let alone be able to fulfil their wishes? (For a partly similar issue, cf. §7 below.)

This objection might be reminiscent of Kautsa's criticism recorded in Nirukta 1.15. Kautsa challenges Yāska's statement that the purpose of the discipline of nirvacana is to explain the meaning of the formulas (Nirukta 1.15). Kautsa claims that, if this is the discipline's purpose, then nirvacana is ‘without meaning/purpose’ (anarthaka), for Vedic formulas have no meaning at all. One of Kautsa's arguments is that formulas have impossible (anupapanna) meanings; as evidence he cites the formula ‘Save him o herb! (oṣadhi)’ (Nirukta 1.15). Kautsa's criticism is not a case of ‘early anti-vedic scepticism’ (Sarup, The Nighaṇṭu and the Nirukta, p. 71). Rather, as Otto Strauss pointed out, Kautsa's view reflects the idea that the ‘power [of the formulas] resides in their mysterious efficacy when they are pronounced, and not in their meaning’. Strauss, O., ‘Altindische Speculationen über die Sprache und ihre Probleme’, Zeitschrift der Deutschen Morgenländischen Gesellschaft 81 (1927), p. 120Google Scholar. This interpretation has been endorsed and further elucidated by Thieme and Taber. P. Thieme, ‘Grammatik und Sprache, ein Problem der altindischen Sprachwissenschaft’, in Kleine Schriften, (ed.) Georg Buddruss (Wiesbaden, 1984), pp. 27–28, and J. Taber, ‘Are Mantras Speech-acts? The Mīmāṃsā Point of View’, in Mantra, (ed.) Harvey P. Alper (Albany, 1989), pp. 144–164.

⁴³ The word ‘meanings/objects’ refer to the entities listed in Nighaṇṭu 5.3.1-22; 29-36, such as ‘horse’, ‘herb’ etc. Two interpretations are possible: (i) One should not think that objects (artha) such as horse and herb have fortuitously ended up being referred to among (taking ‘of the deities’ as a partitive genitive) the other deities mentioned in Nighaṇṭu 5; (ii) One should not think that the meanings (artha) of words such as ‘horse’, ‘herb’ etc. are fortuitous (taking ‘of the deities’ as ‘[the names of] the deities’); that is, they are no by chance, but there must be a reason why they are recorded in Nighaṇṭu 5 together with the other divine names. Yāska explains below that such names, too, are full-fledged divine names, for they ultimately refer to and derive from the one existing deity, the Self (ātman).

⁴⁴ It is unclear whether the first three views about the origin of the deities conflict with, or rather are preliminary to, the fourth view. It is clear however that the last view is endorsed by Yāska.

⁴⁵ References to a tripartite division of deities are found in Vedic literature. We cite two examples: (i) [Sūrya, Vāta, Agni] R̥gveda X.158.1: sū́r_iyo no divás pātu v_aā́to antárikṣ_aā́t | agnír naḥ pā́rthivebh_iyaḥ || (“Let the Sun protect us from heaven, the Wind from the mid-space; let Agni (protect) us from the earthly ones.”)

(ii) [Agni, Indra, Sūrya] AB II.37.17: bhūr agnir jyotir jyotir agnir indro jyotir bhuvo jyotir indraḥ | sūryo jyotir jyotir svaḥ sūryaḥ | (“bhūr, light is Agni, Agni is light; light is Indra, bhuvo, Indra is light; light is Sūrya, Sūrya is, svaḥ, light.”) In this formula, the word jyotis ‘light’ refers to Agni, Indra, and Sūrya; cf. Yāska's use of the same word to refer to the three fires, the terrestrial, the atmospheric, and the celestial (Nirukta 7.18; 20 [×2]; 23; 31).

⁴⁶ Each of the three deities has many names because of its great, manifold powers, or because it performs many different actions. The latter point is illustrated with a simile. Like one ritual agent is called Hotar, Adhvaryu, Brahman, or Udgātar, because he performs different actions, in the same way one of the three deities is called with different names, because it performs different actions. The point of the simile seems to be this: one and the same priest performs different priestly roles in the same or in distinct rituals.

⁴⁷ The fact that there are distinct praises shows that there are different deities. In the same way, the fact that there are different appellations shows that there are different deities.

⁴⁸ While the etymologists argued that one agent performs multiple actions, the ritualists counter that multiple actions are performed by distinct agents. Specifically, the ritualists seem to counter the ritual simile employed by the etymologists. While the latter say that one ritual agent performs multiple actions, the former argue that multiple ritual agents (Hotar, Adhvaryu, Brahman, and Udgātar) perform distinct ritual actions after having divided them among themselves. Note that the ritualists’ criticism is introduced with the formula yatho etad . . . iti. In its other occurrences in Nirukta 7 (I: §7.2; IV: §24.2.1-6; see also Nirukta 1.14), this formula introduces Yāska's own criticism to a previously mentioned view.

⁴⁹ [§6.4] is difficult. In our interpretation, Yāska supports the etymologists’ view by mentioning two parameters by which many deities can be reduced to one: common location and common enjoyment. One example for common location is given: while men, animals and gods are distinct entities, they occupy the same location, earth. Likewise, we are given to understand, distinct deities that occupy one and the same location—i.e. earth, atmosphere, and sky—can be regarded as one. Next, Yāska gives two examples for common enjoyment: (i) the same formula/oblation is enjoyed by Earth as well as by Parjanya, Vāyu, and Āditya; (ii) the same formula/oblation is enjoyed by both Agni, the terrestrial/ritual fire, and by one (or two) of his counterpart(s), i.e. the atmospheric fire (=lightning) and/or the celestial fire (=sun). We take (i) and (ii) to refer to ritual. The point seems to be this: distinct deities enjoy one and the same formula/oblation. Alternatively, a natural-physical interpretation may be possible. In this interpretation, the words Parjanya, Vāyu and Āditya would not refer to the names of the deities, but to natural-physical phenomena: rain-cloud, wind, and sun. Depending on whether one takes the genitive pr̥thivyāḥ ‘of the earth’ as subjective or objective, two interpretations follow: (i) (subjective genitive) One entity, the earth, enjoys distinct natural phenomena, i.e. the rain-cloud, the wind, and the sun; (ii) (objective genitive) the distinct natural phenomena enjoy one and the same object, the earth.

Additionally, note that the word earth (pr̥thivī) occurs in the example for common location and in the first example for common enjoyment (in the second example it may be implied by the reference to Agni, the terrestrial fire). The Earth is the first of the three places (sthāna) in which deities are organised according to the etymologists’ tripartite division [§6.1]. This suggests that Yāska gives examples for the first place only; examples for the remaining two places (mid-space and sky) are implied.

⁵⁰ The exact sense of this simile is unclear. Yāska seems to liken the organisation of the world of deities to that of a human kingdom.

⁵¹ Note that while the bahuvrīhi compound karmātmāna (‘whose nature is the [ritual] actions’) presupposes a masculine word, deva (‘god’), the feminine word devatā (‘deity’) was used above in ‘Now, [we] consider the forms of deities (devatānāṃ)’ [§7]. In Pāṇini (A 5.4.27; see the Kāśikāvr̥tti thereon), the word devatā is recorded as having the same meaning as deva. For the Kāśikāvr̥tti, we use the following edition. A. Sharma, K. Deshpande, and D. G. Padhye, Kāśikā: A Commentary on Pāṇini's Grammar by Vāmana & Jayāditya, 2 vols (Hyderabad, 1969–70).

⁵² The term ākhyāna ‘legend, story’ refers to a school of interpretation of the Vedic formulas. According to Gupta, the adherents of this school maintain that stories about and descriptions of the Vedic deities should not be taken literally, but understood figuratively or allegorically. Gupta, S. K., ‘Ancient Schools of Vedic Interpretation’, Journal of the Ganganatha Research Institute 16, 1/2 (1958–9), p. 149Google Scholar.

⁵³ The mention of Iḷā [Nighaṇṭu 5.5.35] as one of Agni's wives seems to conflict with its occurrence in Nighaṇṭu 5.5. Nighaṇṭu 5.4-5.5 list deities that reside in the mid-space; hence, one would expect Iḷā to be associated with Indra, not with Agni. Note also that while Pr̥thivī [Nighaṇṭu 5.3.26] and Agnāyī [Nighaṇṭu 5.3.28] are mentioned in the Nighaṇṭu section that is associated with Agni [Nighaṇṭu 5.1-3], their order is inverted in [§8.1]; cf. also note 88 below.

The association of Iḷā with Agni may be explained as follows. An oblation offered to Agni can be called iḷā ‘refreshment’. When this oblation, after going up to heaven, returns down to the earth, it becomes rainwater that is associated with Indra. Yāska may be thus referring to iḷā's previous form, iḷā as an oblation to Agni. iḷā is enjoyed by Agni as well as by Indra.

⁵⁴ Yāska seems to analyse suvidatra ‘wealth’ as follows: (i) su-vid-atra ‘something that is well (su-) found (-vid-)’, where -atra is likely to be taken as a suffix; cf. A. Debrunner, Jacob Wackernagel, Altindische Grammatik, Band II, 2, Die Nominalsuffixe (Göttingen, 1954), p. 170; (ii) ‘something that is given (dā) in a good (su-) and special/various (-vi- = viśeṣena/vividhena) way’.

⁵⁵ Some manuscripts and Roth's edition add here: agninā ca pūṣā ‘Pūṣan with Agni’.

⁵⁶ rasādānaṃ raśmibhiś ca rasadhāraṇam (some manuscripts and Roth read rasādhāraṇam). Two etymologies of āditya seem to be referred to here: (i) Āditya takes (ā √dā) the fluid and (ii) retains (ā √dhā) it by means of its rays. (i) is mentioned explicitly in Nirukta 2.13: ādityaḥ kasmād ādatte rasān ‘wherefrom āditya? He takes (i.e. absorbs) liquids’. With respect to (ii), the implied etymology would seem to require the reading rasādhāraṇam. Note also that this etymological link (āditya < ā √dhā) is already attested in the Brāhmaṇas; see the passages cited in Deeg's work. Deeg, Die altindische Etymologie nach dem Verständnis Yāska's und seiner Vorgäger, pp. 206–207. Finally, for the idea that the rays (raśmi) serve the function of retaining the liquid absorbed by the sun, cf. Nirukta 2.13: raśmir yamanāt ‘ray is from restraining’.

⁵⁷ Sarup reads pravahlita [oversight?] (Sarup, The Nighaṇṭu and the Nirukta, p. 137 [text]) and translates it as ‘all that relates to greatness’ (p. 118 [translation]); Roth reads pravalhita (Roth, Jâska's Nirukta sammt den Nighaṇṭavas, p. 119 [text]) and renders it as ‘all that is mysterious’ (‘alles Gehimnissvolle’) (p. 105 [Erläuterungen]). In Vedic pra √valh means ‘to confound, to challenge through an enigma’; see T. Gotō, Die ,,I. Präsensklasse‘‘ im Vedischen: Untersuchung der vollstufigen thematischen Wurzel-päsentia (Wien, 1987), pp. 293–294, and M. Mayrhofer Etymologisches Wörterbuch des Altindoarischen, 3 bde (Heidelberg, 1992–2001), p. 527. The connection between pravalhita and Āditya is somewhat unclear (is the sun enigmatic/mysterious because it disappears at night and reappears in the morning [?]). Indian commentators are not helpful here. Curiously, Pathak glosses pravahlitam (sic) with ‘the growth (vr̥ddhi) and nourishment of herbs and plants’ (Pathak, Nirukta of Yāskācārya, p. 504). While this interpretation would seem to fit the context—the link between growth, the sun, and the sun's actions of absorbing and retaining water is easily understandable—, it is unclear how one may assign such a meaning to pravahlitam/ pravalhitam (contamination with the roots √vr̥dh, √br̥h ?).

Werner Knobl (personal communication) offered the following perceptive remarks. We thank him and cite his words with his permission: “From a Proto-Indo-European (PIE) point of view, a root *vahl would have been impossible. Minimally, the root structure is CVC; maximally, CRVRC (R = resonant). A root *vahl, representing *CVCR, could not have occurred. A root valh (CVRC), however, could. The verbal compound pra√valh-, if it means “confuse, embarrass”, fits the action of the sun, insofar as Āditya, by absorbing the water, makes it disappear mysteriously, like a trickster or conjurer quickly—hey presto!—deceives the eye that is too slow to follow the sudden change brought about by his dexterous fingers”.

⁵⁸ We take iti (‘etc.’) as having an enumerative function.

⁵⁹ We omit translating a few lines (Nirukta 7.12 [138.4-17]), which only contain etymological explanations of a few words and interrupt the flow of Yāska's discussion.

⁶⁰ Cf. Maitrāyaṇī-Sam̐hitā II.2.11: índrāya vr̥traghná ékādaśakapālam [nír vapet] ‘[One should offer an oblation] cooked on eleven potsherds to Indra, the slayer of Vr̥tra’ and Maitrāyaṇī-Sam̐hitā II.2.10; II.6.6; IV.3.9: índrāyām̐homúcā ékādaśakapālaṃ nír vapet ‘One should offer [an oblation] cooked on eleven potsherds to Indra, the deliverer from crisis.’ Note the sandhi in am̐homucā ekādaśa- from am̐homuce ekādaśa-. This is one of the special sandhis observed in the Maitrāyaṇī Sam̐hitā: -e + V- (any accented vowel) → -ā + V-.

⁶¹ Two interpretations seem possible: (a) the actual lists that include also the epithets of the deities fail short of recording all such epithets; (b) such epithets are too many to be recorded manageably in a list.

⁶² Epithets describing the deities’ actions (e.g. ‘slayer of Vr̥tra’ etc.) merely qualify a circumstantial state/aspect of the deities, like ‘who is hungry . . . who has taken a bath . . . who is thirsty’ qualify a circumstantial state/aspect of a brahmin. For a comparable argument regarding the relation between name and the action one performs, cf. Nirukta 1.15.

⁶³ Yāska's etymologies of Agni are discussed in detail in Visigalli, P., ‘Words in and out of History: Indian Semantic Derivation (Nirvacana) and Modern Etymology in Dialog’, Philosophy East and West 67, 4 (2017), 1143–90CrossRefGoogle Scholar and Y. Kawamura, ‘How to Define the God of Fire: Fresh Perspectives on Yāska's Etymology of agní’, in Professor George Cardona's Felicitation Volume, (ed.) Peter M. Scharf (forthcoming).

⁶⁴ Śākapūṇi refers to these verbs in the form of verbal adjectives (past particles), ita-, akta-, dagdha-, and nīta-. This seems to be an archaic practice; cf. H. Scharfe, ‘A New Perspective on Pāṇini’, Indologica Taurinensia 35, p. 111, note 23. In his paraphrase, Yāska cites these verbs in the third person singular, as is common in grammatical literature.

⁶⁵ In translating khalu, we follow Emeneau, M. B., ‘Sanskrit Syntactic Particles — kila, khalu, nūnam’, Indo-Iranian Journal 11, 4 (1969), p. 258CrossRefGoogle Scholar.

⁶⁶ Schematically, Yāska's etymologies are as following:

1. [1]

   [1] agraṇī (agraṃ . . . praṇīyate) → agni.

2. [2]

   [2] aṅgam + nayati → agni.

3. [3]

   [3] aknopana → agni.

4. [4]

   [4] a (< √i; through a form having a, such as ayāni) + g (< √añj or √dah; through forms having g, such as añgdhi or dagdha) + nī → agni.

⁶⁷ On deva and devatā, see note 51.

⁶⁸ ‘Streams of ghee’ refers to water. The word ghr̥ta ‘ghee’ is glossed with udaka ‘water’ in Nirukta 7.24 and it occurs in Nighaṇṭu 1.12.10 as one of the synonyms of water. Ghee refers metaphorically to (rain)water already in the R̥gveda; see e.g. R̥gveda I.164.47, VII.62.5, VII.65.4. In Yāska's interpretation of R̥gveda IV.58.8, the streams of ghee, i.e. (rain)water, are represented as women that willingly approach the kindling stick, Agni (=lightning); the sexual imagery is clear.

⁶⁹ Yāska seems to take the word agním (R̥gveda I.164.46d) as the direct object of the verb ‘to speak’ (vadanti). The word agni is the implied direct object (x) governed by the verb āhuḥ (‘they call’ [x y], R̥gveda I.164.46ad). All the other accusatives are the object predicates (y). This means that ‘Indra’, ‘Mitra’, ‘Varuṇa’, ‘the well-winged Garutmat’ (bird, i.e. the sun), ‘Agni’, ‘Yama’ and ‘Mātariśvan’—all such names have one and the same referent, the (terrestrial) Agni, which is equated with the Great Self. Note the inclusion of ‘Agni’ (R̥gveda I.164.46a) among such names. Note also that ‘Yama’ and ‘Mātariśvan’ are not mentioned in Yāska's commentary.

⁷⁰ Yāska gives two etymologies of Garutmat: (i) garut- (= garaṇa ‘praise’) + -mat (= –vat- ‘provided with’); (ii) garu- (= guru- ‘heavy’, i.e. ‘great’) + -tmān (< ātman ‘soul’). With respect to (ii), note two things: first, the juxtaposition garu- = guru- may be facilitated by the vowel apophony in the comparative and superlative forms of guru-, i.e. garīyas/gariṣṭha. Second, the juxtaposition -tmān = ātman may find support in the Vedic forms tmánā (instrumental), tmáne (dative), and tmáni (locative), which are the old oblique cases of the word ātmán.

⁷¹ In some manuscripts, R̥gveda I.99.1 is here cited and commented upon. This passage is not commented by Durga. Sarup gives the passage within brackets (Sarup, The Nighaṇṭu and the Nirukta, p. 141 [text]).

⁷² Yāska explains aśva (jātávedasam áśvaṃ) in two ways: (i) as an adjective-participle (samaśnuvāna) derived from √aś ‘to reach’, which qualifies Jātavedas. Note that Yāska seems to take sam in (jātaveda-)sam as (also) being the preverb to be construed with √aś. That Jātavedas is all-pervasive would be consistent with etymologies [1], [2], and [3] given above in [§17]. (ii) Yāska takes aśva as a noun ‘horse’ which is employed in a simile.

⁷³ R̥gveda X.188, which consists of three gāyatrī verses, is the only gāyatrī hymn in the R̥gveda that is addressed to Jātavedas (there are other hymns that are addressed to Jātavedas, but they are not in the gāyatrī). Should the (ritual) circumstances demand it, however, it is possible to utilise verses in the gāyatrī meter that are addressed to Agni as if they were addressed to Jātavedas.

⁷⁴ We cite R̥gveda I.50.1 in its entirety: úd u tyáṃ jātávedasaṃ | deváṃ vahanti ketávaḥ | dr̥śé víśvāya sū́r_iyam || (‘‘Up do the beacons convey this god Jātavedas, the Sun, for all to see’’).

⁷⁵ The word vaiśvānara is said to be the same as the word viśvānara; the former is also said to be a derivative (tasya ‘its’) of the latter. Since both words are the same, the analysis of viśvānara as ‘fixed upon all the beings’ (pratyr̥taḥ sarvāṇi bhūtāni) must also apply to vaiśvānara. What kind of information does this analysis provide? The authors’ views differ in this regard. For Kawamura, this analysis pertains only to the semantic aspect of the word vaiśvānara; accordingly, vaiśvānara and viśvānara are the same because they share the same meaning. For Visigalli, this analysis also implies an etymological analysis: viśvānara < viśvān (=sarvāni bhūtāni) + ara (=praty-r̥ta [< √r̥]); accordingly, vaiśvānara and viśvānara are the same because they share both the same meaning and the same basic etymology. Cf. also §23.3.

⁷⁶ Possibly this means that the lack of rainfall makes impossible the performance of ritual actions or actions of any kind that needs water.

⁷⁷ Yāska explains śaṃbara, vr̥tra, and dasyu as meaning rain-cloud. śaṃvara (Nighaṇṭu 1.10.13; the different spelling ‘v’/‘b’ is not significant) and vr̥tra (Nighaṇṭu 1.10.27) are listed in the Nighaṇṭu as synonyms of cloud (megha). (śambara occurs in Nighaṇṭu 1.12.93 as a synonym of water [udaka].) Yāska's way of explaining Pūrus etymologically seems reminiscent of the Mīmāṃsakas’ exegetical strategy whereby the proper names mentioned in the Veda are taken to refer not to actual human beings but to natural elements. This strategy is part of the arguments the Mīmāṃsakas deploy to ensure the eternality of the Vedas; see Śabara's commentary on Mīmāṃsā-sūtra 1.1.31 (G. Jha, Śābara-Bhāṣya Translated into English. Vol. I, adhyāyas I-III [Baroda, 1933], pp. 49–50). Cf. also the relevant brief remarks by Kahrs and Bronkhorst regarding a possible affinity between Yāska's and Mīmāṃsā's attitude towards the Veda. Kahrs, Indian Semantic Analysis, p. 273 and J. Bronkhorst, ‘Etymology and Magic: Yāska's Nitukta, Plato's Cratylus, and the Riddle of Semantic Etymologies’, Numen 48 (2001), p. 159.

The word kāṣṭhāḥ is glossed with apaḥ ‘waters’. (Note the inverted order: *kāṣṭhā apaḥ would seem more natural.) While kāṣṭhāḥ occurs in Nighaṇṭu 1.6.5 as one of the synonyms for direction (diś), it is also said to mean water (āpo ’pi kāṣṭhā ucyante) in Nirukta 1.15.

⁷⁸ tām anukr̥tiṃ hotā . . . pratipadyate. We taka the ta-pronoun to have pratyavaroha ‘descent’ as its antecedent. It agrees in gender (tām < *tam [=pratyavaroha]) with its object predicate anukr̥ti (f.).

⁷⁹ We take the subject of ‘comes to’ to be the patron of the sacrifice, in his symbolical descent from the sky to mid-space, and then from mid-space to earth. See Heesterman, J. C., The Ancient Indian Royal Consecration: The Rājasūya Described according to the Yajus Texts and Annoted, (Mouton, 1957), pp. 12–13CrossRefGoogle Scholar. Alternatively, the subject may be the Hotar and ‘comes to’ means that the Hotar comes to sing verses addressed to mid-space deities, like Rudra and the Maruts (in the Nighaṇṭu 5, they are listed as mid-space deities), and then verses addressed to terrestrial Agni (and the Maruts). This second interpretation is followed by Pathak, Nirukta of Yāskācārya, p. 530 and Parameśvarānanda et al, Yāska's Nirukta with Sanskrit and Hind Commentaries (New Delhi, 2013), pp. 364–365.

⁸⁰ Ritualists argue that vaiśvānara refers to the sun, because a hymn dedicated to Vaiśvānara is recited by the Hotar in conjunction with the beginning stage of the sacrificer's symbolical descent from the sky to the earth. Thus, that a hymn addressed to Vaiśvānara is employed in connection with the sky would confirm that vaiśvānara refers to the sun, the celestial fire.

Ritualists also reply to a possible objection (or simply further clarify their stance). The fact that the stotriya is addressed to Agni, i.e. the terrestrial fire (and the Maruts) does not contradict their view that the hymn is addressed to Vaiśvānara that is the sun, the celestial fire. For this stotriya is recited after the sacrificer has (symbolically) descended down to earth, and has come to the terrestrial fire. In other words, the vaiśvānara-hymn and the agni-stotriya have distinct domains of application.

⁸¹ For the Śāṅkhāyana-Śrautasūtra, we use the following edition. A. Hillebrandt, The Śāṅkhāyana Śrauta Sūtra Together with the Commentary of Varadattasuta Ānartīya. 3 vols (Calcutta, 1888–97).

⁸² This passage also occurs in Āśvalāyana-Śrautasūtra VIII.10.3, in the context of the second chandoma day.

For the Vājasaneyi-Sam̐hitā and the Āśvalāyana-Śrautasūtra, we use the following editons respectively. W. L. S. Paṇśīkar, Śuklayajurveda-Samhitâ (Śrîmad-Vâjasaneyi-Mâdhyandina.) with the Mantra-Bhâshya of Mahâmahopâdhyâya Śrîmad-Uvatâchârya and the Veda-dîpa-Bhâshya of Śrîman-Mahîdhara (with Appendices & Mantra-kośa) [Bombay, 1912] and Vidyāratna, R., The Śrauta Sútra of Áśvaláyaṇa, with the Commentary of Gárgya Náráyaṇa (Calcutta, 1874)Google Scholar.

⁸³ Śākapuṇi maintains that the name vaiśvānara refers to the terrestrial fire, for the word vaiśvānara is a derivative of the word viśānara, whose dual form viśvānarau refers to the atmospheric and celestial fires. vaiśvānara denotes the terrestrial fire as being born from these two upper fires.

⁸⁴ The atmospheric (=lightning) and terrestrial fires exhibit opposite characteristics. The former burns in water and is extinguished in contact with objects; the latter does the opposite. People obtain (upa-ā-√dā) the terrestrial fire from lightning. This illustrates that the former fire (vaiśvānara) is born from the latter (viśvānara).

⁸⁵ We take this to refer to the mid-day sun, which has just (prathama-) reached (-samāvr̥tta) its highest (udīci) position in the sky.

⁸⁶ R̥gveda I.98.1 is cited above in [§22].

⁸⁷ Close connection (saṃsaṅga) appears to refer to the encounter between the flames emanating from the terrestrial fire and the rays of the sun. The former move upwards from the earth; the latter move downwards from the sky. The use of the conditional avakṣyat is a bit problematic. For the conditional usually expresses an irrealis (e.g. Had he been a millionaire, he would have bought a Porsche), which makes no sense here. It is possible that conditional avakṣyat is due to the influence of the two conditional forms (abhaviṣyan, astoṣyan) that occur immediately below.

⁸⁸ Yāska argues that if Vaiśvānara referred to the sun, then (a) the name vaiśvānara should co-occur with the names of deities residing in the sky; (b) Vaiśvānara should be praised with actions relatable to the sun. Instead, (a¹) vaiśvānara occurs in hymns addressed to Agni, the terrestrial fire; (b²) Vaiśvānara is praised with reference to actions relatable to Agni. (a) and (b) refer to two (of the three) parameters employed by Yāska in [§8] to describe the three main deities, Agni, Indra, and Āditya: namely, their shares (bhakti), i.e. associations with particular items or deities, and their characteristic actions (karman), i.e. actions with which the deities are described in the formulas.

It is interesting to compare (a) and (b) in [§23.3.4] with the two same parameters given in [§8.3], where Yāska describes Āditya. While (a) is consistent with the shares in [§8.3] (‘the group of gods transmitted [in the Nighaṇṭu 5.6 as residing] in the highest place [sky]’), (b) does not match well the corresponding actions (absorbing fluid and whatever is enigmatic/mysterious). This inconsistency may perhaps be explained by taking the actions given in [§8.3] as the main characteristic action, and those given in [§23.3.4] as secondary.

Finally, note that the order in which Bhaga, Savitar, Pūṣan, and Viṣṇu are mentioned in [§23.3.4] does not match the order in which they are listed in Nighaṇṭu 5.6: savitā [5.6.7]; bhagaḥ [5.6.8]; pūṣā [5.6.10]; viṣṇuḥ [5.6.11]. This discrepancy may suggest that the Nighaṇṭu 5 to which Yāska referred differs slightly from the one we have now; cf. note 53.

⁸⁹ Poured onto the ritual fire (=terrestrial Agni), oblations (=water) go up to the sky and return down to earth in the form of rain. Cf. Geldner's note “Wechselwirkung zwischen Himmel und Erde, zwischen Regen als Gabe des Himmels und dem Opfer als der Gabe der Menschen”. Geldner, K. F., Der Rig-Veda: Aus dem Sanskrit ins deutsche Übersetzt und mit einem laufenden Kommentar versehen, 3bde (Cambridge, Mass., 1951), p. 23Google Scholar.

⁹⁰ Both r̥ta [Nighaṇṭu 1.12.72] and ghr̥ta [Nighaṇṭu 1.12.10] are listed as one of the synonyms for ‘water’.

⁹¹ Our interpretation of Yāska's commentary on R̥gveda I.164.47 is indebted to Pathak, Nirukta of Yāskācārya, p. 538. The verse seems to refer to the two courses of the sun: the ‘northern course’ (uttarāyaṇa), from 23 December till 22 June, and the ‘southern course’ (dakṣiṇayāṇa), from 22 June till 23 December. The latter is referred to as ‘black way’ (kr̥ṣṇáṃ niyā́naṃ) in R̥gveda I.164.47. Yāska glosses niyāṇa with nirayaṇa ‘out-way’, and calls it ‘the night of the sun’, possibly because nights are longer than days during the ‘southern course’. During the southern course, the sun-rays absorb water and carry it to heaven, where water is thought to be stored in the sun. During the northern course, the water stored-up therein returns to earth in the form of rain.

⁹² Cf. Kāṭhaka-Sam̐hitā XI.10, Taittirīya-Sam̐hitā II.4.10.2, and Maitrāyaṇī-Sam̐hitā II.4.8. For the Kāṭhaka-Sam̐hitā and the Taittirīya-Sam̐hitā, we use the following editions. L. von Schroeder, Kâṭhakam: Die Saṃhitâ der Kaṭha-Çâkhâ, 3 bde (Leipzig, 1900–10) and A. Weber, Die Taittirîya-Saṁhitâ, 2 Bde (Leipzig, 1871–2).

⁹³ This picks up on āmnātaḥ (‘. . . has been traditionally considered . . .’) in [§23.2.1], i.e. that statement is based on human tradition, it does not have the same degree of authority as evidence found in śruti.

⁹⁴ Cf. Taittirīya-Sam̐hitā II.4.10.2: sauryám ékakapālam. The ritualists identify Vaiśvānara with the sun, since an oblation cooked on twelve-potsherds is offered to Vaiśvānara, and the number twelve (=twelve months) is associated with the sun. Yāska counters that there is not a necessary association between the number twelve and the sun, because oblations cooked on one-potsherds and on five-potsherds are also offered to the sun. Hence, it cannot be concluded that Vaiśvānara is the sun from the fact that it is the recipient of an oblation cooked on twelve-potsherds. Note the word ‘non-explanation’ anirvacana (< nir√vac). This word is reminiscent of a specific Vedic usage of derivatives of nir√vac: a formula is nirukta if it contains the name of a deity or one of its recognisable epithets or symbols; if these are missing, the foruma is anirukta; see Renou and Silburn, ‘Nírukta and Ánirukta in Vedic’ and Visigalli, ‘The Vedic Background of Yāska's Nirkta’, pp. 112–114.

⁹⁵ See, for example, Śatapatha-Brāhmaṇa XIII.3.8.3: iyáṃ vái vaiśvānaráḥ; Śatapatha-Brāhmaṇa VI.6.1.5: saṁvatsaró vaiśvānaráḥ; Taittirīya-Brāhmaṇa III.7.3.2: agnír vaiśvānaráḥ | yád brāhmaṇáḥ |

For the Śatapatha-Brāhmaṇa and the Taittirīya-Brāhmaṇa, we use the following editions. A. Weber, The Çatapatha-Brâhmaṇa in the Mâdhyandina-Çâkhâ with Extracts from the Commentaries of Sâyaṇa, Harisvâmin and Dvivedagaṅga, The White Yajurveda, part II (Berlin and London, 1855), reprint (Leipzig, 1924) and Kr̥ṣṇayajurvedīyaṃ taittirīyabrāhmaṇam srīmatsāyaṇācāryaviracitabhāṣyasametam, 3 vols (Poona, 1898).

⁹⁶ One part of the nivid formula (“[Agni Vaiśvānara] shines in the direction of heaven, in the direction of the earth”) was cited as an argument for the view that Vaiśvānara is the sun. Now, Yāska cites another part of this same formula, which shows that Vaiśvānara is the terrestrial fire. We give the nivid formula in full, followed by Caland's translation. Śāṅkhāyana-Śrautasūtra VIII.22.1 (I.94.11-16): agnir vaiśvānaraḥ somasya matsat | viśveṣāṃ devānām samit | ajasraṃ daivyaṃ jyotiḥ | yo viḍbhyo mānuṣībhyo dīdet | dyuṣu pūrvāsu didyutānaḥ | ajara uṣasām anīke | ā yo dyāṃ bhāty ā pr̥thivīm | orvantarikṣam | jyotiṣā yajñiyāya śarma yaṃsat | agnir vaiśvānara iha śravad iha somasya matsat | premāṃ deva iti samānam | (Caland [1953, p. 224]: ‘‘“May Agni Vaiśvānara enjoy the Soma—the fuel stick of all the Gods—the imperishable divine light—who hath shone for the clans of men—who hath shone on the days of old—imperishable on the foremost rank of the dawns—who illumines the sky and the earth—and the wide intermediate region—by his light may he bestow protection on him who is worthy of sacrifice—May Agni Vaiśvānara hearken here, may he enjoy the Soma. May he favour this God-invocation” etc. as above’’.) W. Caland, Śāṅkhyāyana-Śrautasūtra: Being a Major yājñika Text of the Ṛgveda, Edited with an Introduction by Lokesh Chandra (Nagpur, 1953), reprint (Delhi, 1980).

⁹⁷ The context in which this passage occurs is the first chandoma day.

⁹⁸ While kám seems to correspond to ‘this terrestrial fire’ in Yāska's commentary, it is unclear how Yāska understood it. Yāska may have taken kám as: (i) an interrogative pronoun (our transl. of R̥gvedic verse); (ii) a verse-filling particle (pādapūraṇa; so the Niruktabhāṣyaṭīkā); (iii) as a qualifier of Agni. ka is listed in Nighaṇṭu 5.4.14, the section of the names of mid-space deities.

⁹⁹ Yāska cites the verse as further evidence that Vaiśvānara does not denote the sun but refers to the terrestrial fire. Yāska seems to interpret the verse as describing the origin of the terrestrial Agni Vaiśvānara. Its source is the celestial abode of the waters that is in the sky. This abode is identified with the sun, the Shining one (vivasvān). The mid-space deities grasp the atmospheric fire from its source, and the wind, the atmospheric carrier par excellence, brings it down to earth. Considering §9.3.4.3, where ‘vaiśvānara’ is said to derive from the ‘viśvānara’ which denotes the atmospheric and the celestial fires, it is here conceivable that Yāska intimates that ‘vaiśvānara’ derives from ‘vivasvān’.

¹⁰⁰ Yāska analyses mahiṣā in two ways, as formed by locative mahati ‘the great’ (=mid-space) and āsīnāḥ ‘residing/dwelling’; or as a participle plural ‘those who are powerful’ based on √mah.

¹⁰¹ Sarup's edition reads aharad (Sarup, The Nighaṇṭu and the Nirukta, p. 146), but Yāska's gloss below—‘bringer’ (āhartāram)—suggests solving the sandhi vāharad as vā-āharad.

¹⁰² The particles ū tu are glossed with tu in Yāska's commentary. If they retain an adversative sense, this sense seems weak.

¹⁰³ By night, Agni resides on earth in the form of actual fire. At dawn, the same Agni is born as the rising sun. The knowledge that one and the same Agni is the terrestrial fire at night and the sun by day is seen as the esoteric knowledge of the gods. Agni performs the action of rising from the earth and reaching the sky as the sun, moving across the atmosphere, thereby traversing the three worlds. The word tataḥ can have a temporal (after, i.e. once the night is over) or spatial (from there, i.e. from the earth) sense. The genitive yajñíyānām ‘worthy of sacrifice’ can be either subjective (the wisdom (prajñā [=māyā]) possessed by the gods) or objective (the wisdom about the gods).

¹⁰⁴ Durga (R̥jvarthā [717.18-22]) says that R̥gveda X.88.6 describes only two worlds (sthāna); whereas R̥gveda X.88.10 mentions all the three worlds.

¹⁰⁵ The last pāda (“only . . . [them]”) is not commented by Yāska; cf. R̥gveda X.88.10d in Nirukta 7.28. The translation of this pāda is from Jamison and Brereton, The Rigveda, p. 1534.

¹⁰⁶ Yāska seems to give two analyses of the word mithuna: [i] when mithuna refers to a godly pair; [ii] when it refers to a human pair. [i] The word is analysed as comprising three parts: (1) mi- in the sense of leaning/relying on (śrī-); (2) a nominal affix -thu- or -tha-; (3) the final part of the word ‘-na’ is linked with nayati (nī-) (‘lead’) or with van- (‘seek for’ [?]). Note two points. With respect to (2), Yāska may give -thu- first because it phonetically matches (mi-)thu(-nau); then he gives -tha- because it is a more common affix; is Yāska perhaps suggesting that the odd -thu- derives from the standard -tha-? With respect to (3), Yāska's analysis of van can be explained in two ways. First, van becomes (mi-th-)un(<van)-a, because of samprāsaraṇa. Second, Yāska's analysis presupposes the dual form mithunau: (mithu)-nau < na(-yati); (mithu)-nau < van-. In this case, the two items (nau < van-) consist of the same sounds, yet the order is reversed. [ii] mithuna < meth- + van. It is noteworthy that Yāska differentiates between mithuna as referring to gods and to humans. On etasmād eva in the Nirukta, cf. Kahrs, Indian Semantic Analysis, p. 131.

¹⁰⁷ Tentatively, we think that Yāska takes tat as a gloss of it (R̥gveda X.88.17c), as a pronoun that correlates with yatrā.

¹⁰⁸ Yāska's analysis of sákhāyaḥ ‘companions’ as samāna ‘same, identical’ plus ākhyāna ‘thought/knowledge’ is reminiscent of analogous analyses, samāna plus khyāti/khyāna ‘knowledge, insight’. Compare the following analyses of sakhāyaḥ in R̥gveda X.71.2. This verse is cited in Nirukta 4.10 and in Mahābhāṣya (I.4.10-11). While neither Yāska nor Patañjali say much about sakhāyaḥ, the analyses given in their respective commentaries are reminiscent of the analysis in Nirukta 7.30. For the Mahābhāṣya, we use the following edition. K. V. Abhyankar, The Vyākaraṇa-mahābhāṣya of Patañjali: Edited by F. Kielhorn, 3 vols (Bombay, 1880–5), Third edition, revised and furnished with additional readings, references and select critical notes by K. V. Abhyankar, 3 vols (Poona, 1962–72).

R̥jvarthā (I.372.11-12): sakhāyaḥ—samānakhyānāḥ / samānakhyānānām eva samāneṣu śāstreṣu kr̥taśramāṇām tad yathā vaiyākaraṇānām vaiyākaraṇā eva nairuktānāṃ nairuktā eva | sakhyāni sakhibhavān saṃjānate [. . .] | (‘‘they know the friendship [i.e.] the state of being a friend only of those who possess the same knowledge [i.e.] those who have accomplished an effort with respect to the same śāstras—as for example only the grammarians [know the friendship] of [other] grammarians, only the nairuktas [know the friendship] of other nairuktas’’.)

¹⁰⁹ In Yāska's interpretation, this difficult verse seems to express a parallelism. The debate between the two divine Hotars, the terrestrial and celestial fires, has its counterpart in a debate among human priests.

¹¹⁰ The particle iva is commonly used to express comparison. In citing a sentence example, Yāska says that iva also means ‘just now’. He ascribes this same sense (‘just now’) also to na, which is often used in Vedic as a particle of comparison.