Notes

EXEGETICAL AND CKITICAL NOTES CHAPTER I

1. 1.] Muir translates the passage as follows : A sacred record (samamnaya) has been compiled, which is to be expounded. This is called the Nighantus.’ 1 His rendering of samamnaya by’ a sacred record * is not quite correct. ‘It may be that he has been led to this conclusion on account of misunderstanding the following explanatory note of Durgacarya :’ The meaning is, that this compila tion has been put together by seers, as being an illustration for understanding the meaning of the Vedic stanzas, in a single list, divided into five chapters/ He explains the word samamnaya as follows :’ The list of words beginning with gaus and ending with devapatnl is called samamnaya. . . . It’is called samamnaya because it is handed down by tradition.’ That the word samamnaya means ’ a listf or* a traditional list’ is further shown by the following passages : AtMto varnasamamnayam vyakhyasyamah.2 1 Now we shall explain the list of letters.’ Atha varnasamamnayah* ‘Now the list of letters.’ The Tn-lMsya-ratna has the following comment : ’ saw denotes aggregation ; a tradition ; and mnaya signifies the instruction handed down from generation to generation in succession.’ Of. also : padaksara samamnayam chandasy eva pratitthitam.* Muir also leaves out the words imam samamnayam in his translation of the third sentence, which may be translated as’ this same list ‘.
2. 3.] Yaska gives three derivations of the word Nighantu: (1) from^gam with the preposition ni, (2) fromVhan with ni, (8) fromVhr with ni. None of them is satisfactory as they do not account for the cerebral t. Durga, fully conscious of the unsatisfactory character of Yaska’s etymologies, tries to get over the difficulty by the following ingenious theory of his own. He says : ’ The arrangement of words is indeed threefold, i. e. those whose grammatical form is (1) direct, (2) indirect, and (8) obscure. With reference to them, the root is explicitly stated in words of direct grammatical forms ; it is inherent in those of indirect forms. As regards words of obscure forms, the process of

• The bracketed figure does not representJ Muir, op. cit., vol. ii, p. 165. the chapter and the sections, but the sections 2 Vpr. viii. 1. 8 Tpr. i. 1. and the line. 4 Anuvdkanukramanl, 1. 6.

1. ;] EXEGETICAL AND CRITICAL NOTES 201
   explanation is the following : having been reduced from the state of obscurity to that of the words of indirect forms, they should be explained by those of direct forms, e. g. ni-ghantavah is a word whose grammatical form is obscure. The same (being reduced to) ni-gantavah attains to the state of a word of indirect form, and as ni-gamay’darah to that of a word of direct form.’ This theory acquires some plausibility from the explanation of Aupamanyava, but there is no evidence to show that Yaska agreed with Durga’s threefold classification. On the contrary, Yaska does not follow the process laid downby Durga, that words of obscure forms should be explained by those of direct forms through the intermediation of those of indirect forms. This fart alone is sufficient to indicate that Yaska did not subscribe to the view of the com mentator. Another more or less fanciful derivation of the word nighantu is suggested from the root granth or grath by transposition, which, however, has the merit of accounting for the cerebralization through r, i. e. ‘a list of words which have been strung together ‘, or from*/ghat orVghant with fit,’ to join together ‘, i. e. ’ a list of words which have been joined together ‘.
2. 6.] Gune translates the passage as follows :’ Nouns are where being pre- dominates and a verb is where becoming predominates respectively/1 The translator has changed the order of the original without any justification.
3. 7.] Owing to the want of precise punctuation the commentator has not clearly understood the passage. He ends the sentence with Wiavatah and gets over the difficulty of grammar by construing Wiavatah twice as follows : Roth 2 seems to agree with the commentator in the punctuation and interpreta- tion of the passage. Both interpret cT^Hf as referring to a sentence, but there is hardly any justification for attributing this sense to fnr. It immediately follows Yaska’s definition of nouns, which definition is not comprehensive, for it excludes all verbal nouns whose fundamental notion is more a becoming than a being. In order, therefore, to reconcile his definition with this class of nouns and to draw a clear line of demarcation between verbs and verbal nouns, Yaska expresses his meaning more definitely at greater length in this sentence. There should be no full stop after Wiavatah. The sentence should be read thus: l Mfaif<f?f I ’ Where both are dominated by a becoming, a becoming arising from a former to a latter state is denoted by a verb, as “he goes “, " he cooks “, &c. ; the embodiment (of the whole process) from the beginning to the end, which has assumed the character of being, by a noun, as “going “, cooking”, &c.’ The difference is this: a becoming in the course of a process or state of flux is denoted by a verb, but the embodi ment of the complete process is denoted by a noun.3 1 IA., vol. xlv, p. 158. 2 Cf. op. cit., p. 4. 8 Cf. Gune, foe. et*. 202 EXEGETICAL AND CRITICAL NOTES [1. 7 The commentator cites the following two stanzas to show clearly the difference between nouns and verbs : ‘They call that notion by the term verb, which is connected with many actions, which proceeding from a former to a later state is yet one, and which is accomplished through the termination of the actions. A becoming, produced by the completion of action, capable of being expressed by a word ending in a primary affix, and joined with number, case, inflexion, and gender, should then be regarded as a noun.’
4. 8.] With a view to further distinguish nouns from verbs, Yaska says in this sentence, that there is a specific difference in the use of terminology applicable to nouns and verbs, e. g. beings can only be pointed out by a demon strative pronoun, as* this cow ‘,’ that elephant ‘, &c. Verbs, on the other hand, cannot be so pointed out. In order to indicate them, one is obliged to use the verb ’ to be ‘. The use of different technical terms, which are non-interchange able, shows that there is some fundamental difference of notions between them, the characteristics of which have already been mentioned.
5. 10.] In the original text, the sentence forms a part of the first section, and is immediately followed by the second. It introduces a controversy, i. e. whether words are permanent or impermanent, a controversy which in its character differs altogether from the subject-matter of the first section. To begin the second section with this sentence would have been therefore a more logical division of the sections, and more in harmony with the modern concep tion of what constitutes a paragraph. That a section of the Nirukta more or less corresponds to a paragraph is indicated by the evidence of older MSS. which place the full stop, i. e. danda, at the end of, and very seldom within the section itself, excepting the commencement and the conclusion of a quotation. This is further supported by the fact that, in most cases, one section is devoted to the explanation of one Vedic stanza only. Hence it is argued that the division of the text of the Nirukta into sections, as constituted at present, is illogical and arbitrary. It is therefore proposed* to discard in this respect the authority of the MSS., which has been hitherto strictly followed, and ’to make sections according to the most natural division ‘. 2 ’ Faithfulness ‘, says Gune,’ is indeed a merit, but it should not be overdone, at least not where reason says other wise.’ 3 The suggestion is rather a bold one and, I think, contrary to the canons of modern editorship. The suggested improvements can very well be shown in foot-notes, but the wisdom of rearranging the text itself in opposition to the evidence of the MSS. is doubtful. However, there are practical difficulties in accepting this suggestion. Redivision of sections would involve the transference of a considerable number of passages to new sections, and would thereby reduce the utility of various books of reference, as far as these passages are concerned. Further, if the sections of the Nirukta do not harmonize svith the modern con 1 Cf. Gurio, op. cit., p. 157. 2toe. tit. 3 loc. cit.
6. IQ] EXEGETICAL AND CRITICAL NOTES 203 ception of what constitutes a paragraph, does it necessarily follow that they are illogical ? Is this, by itself, a conclusive proof of their arbitrary character? Tomy mind, the answer is in the negative, for the ancients may have had a different conception of the constitution of a paragraph. As far as Yaska is con cerned, a careful examination of all the sections of the Nirukta indicates that Yaska proceeds methodically in his division of the text into sections, whichdivision is based on a general principle. By the time of Yaska very great weight was attached to the Vedas, especially by that scholar himself, as is evident from Chapter I, particularly from his rejoinder to the adverse criticism of Kautsa. To him, a Vedic stanza was of the utmost importance, and accord- ingly formed a very suitable beginning for a new section. There are 400 sections altogether in the first twelve chapters of the Nirukta, distributed among those chapters as follows : Chapter. I 204 EXEGETICAL AND CRITICAL NOTES [1. 10 at the end of the previous section, e. g. Athdpi prathama bahuvacane. 1 It appears that this method of putting a short sentence of a section at the end of a previous section which was a necessity in the case of sections beginning with Vedic stanzas has been rr echanically extended to the eight sections mentioned above. From what has gone before, it will be clear that the sections in the Nirulcta are not illogically nor arbitrarily divided, but are based on a general principle adopted by Yaska. Gune’s suggestion to rearrange the sections and to discard the authority of the MSS. is therefore unacceptable. ’ Speech is permanent in the organs only’ means that the character of speech is evanescent. Sounds disappear as soon as they are uttered. They have no existence beyond articulation by the vocal organs and their corresponding cognition by the sense of hearing. With this may be compared the remarks of Patanjali : ’ BECAUSE SPEECH is SET IN MOTION BY INDIVIDUAL LETTERS AND BECAUSE THE LETTERS ARE ANNIHILATED AS SOON AS THEY ARE UTTERED.’

• Speech is set in motion by individual sounds. One cannot produce two sounds simultaneously. Take the word gauh. for instance : when the sound g is being produced, neither au nor the visarjariiya can be uttered ; when au is being produced, neither g nor the visarjanlya can be given utterance ; and when the visarjanlya is being produced, one can neither pronounce g nor au, because they are annihilated as soon as they are uttered. The sounds indeed are perishable after being uttered, i. e. as soon as they are uttered they have perished.‘2 Durga has the following comment :’ As soon as the speech of the speaker is an act of utterance in the organ, so long only it is possible to say that it exists, but fallen from lips it no longer exists. . . . And the sounds that have perished and that have not perished cannot be conjoined.’ The question whether words are eternal or non-eternal was a very important subject of discussion among the ancient Indians. For instance, cf. Patanjali : 3 ’ But again, is the word eternal or non-eternal (lit. created) ? Whether it is eternal or non-eternal is thoroughly discussed in the Samgraha.* There the 1 N. 1. 15. The reference in IA., loc. cit., Katyayana’s Varttika 45 on Panini’s Sutra, of this passage to p. 43 in Roth’s edition is i. 2. 64. As a lexicographer and medical wrong, 43 being a misprint for 34. author, he is quoted by numerous writers on 2 ifah<ibha$ya, i. 4. 4 ; Kielhorn’s edition, those subjects, see Catdhgus catalogonim, vol. i, vol. i, p. 356. p. 618. Aufrecht attributes the Samgraha to 3 op. cit., i. 1. 1, vol. i, p. 6. Patanjali himself although it is quoted by 4 According to Nagoji, Savigraha is a gram- him. Tradition mentions Patanjali as the matical treatise by Vyadi. Cf. Catalogus cata- curer of the three evils, (1) the evils of body, logorum, vol. i, p. 686. The name of Vyadi is (2) evils of speech, and (3) evils of mind, well known in Sanskrit Literature. He and represents him as having written three seems to have been a man of versatile genius, masterpieces, i. e. (1) the Caraka Sanihita, being famous as a grammarian, lexicographer, (2) the Mahdbha$ya, and (3) the Yogasutras, writer of authoritative books on medicine and in order to remove the three evils respec poetics. As a grammarian, he is mentioned tively. But nowhere is he credited with the in Rpr. iii. 14. 17 ; vi. 12. 13 ; xii. 15 ; authorship of another grammatical treatise.

1. IG] EXEGETICAL AND CRITICAL NOTES 205 arguments against (the proposition) are stated, as well as the arguments in favour thereof. The conclusion (arrived at) in that work is this, that although the word is eternal, it is also non-eternal ; the definition being applicable in both ways.’ Patanjali, however, does not seem to agree with the conclusion arrived at in the Samgraha. He remarks : 1

• The twofold cJiaracter of words is not proved. ‘Twofoldness of (the character of) words is not proved. If it be said that twofoldness has been proved, this cannot be maintained.’ He says further : 2 ’ Words are eternal. In the eternal words there must be sounds which are perpetual, unchangeable, and free from elision, addition, and modification. ’ This is proved by the eternity of sounds . . . words are eternal.’ The character of words is discussed by him at greater length in his comment on the varttika : fa% TJ^T^^R^t3 The argument may be summarized as follows : The relation of words to the objects they denote is eternal. This power of denotation of objects is natural and uncreated,4 and so are the objects. It may be objected that although matter is uncreated, yet the various forms into which it is moulded are non-eternal ; e.g. the earth is not created by man, but the various pots and jars, into which it is shaped, are the creations of a potter, hence the words denoting these forms cannot be eternal. To this the answer is given that the form is eternal also, for it is not something radically different from the matter, but only a particular state which the matter assumes at a particular time. ’ But how is it known that word, object, and their mutual relation are eternal?’ From the ordinary usage of the world. In daily life, people perceive objects and use words to denote th jm. They do not make any effort in creating them, because effort is necessary in producing what is non eternal only ; e. g. a man desirous of using a pot goes to the house 01 a potter and says,’ Please make me a pot, I want to use it ‘, but a man desirous of employing a word does not go to the house of a grammarian and say,* Please, sir, make me a word, I want to use it ‘. People perceive objects and use words to denote them without any effort. 5 The way in which Patanjali refers to the except a solitary MS. of Vyadi paribhuta Krttf, Samgraha shows that it was a work of some in the temple library of the Maharaja of eminent scholar and was very well known in Jammu and Cashmere. The MS. is described his time. Further Patanjali does not agree in Stein’s Catalogue of Skt. MSS. on p. 47. The with the conclusion arrived at in the Sa/- description of the MS. is very meagre, so it graha; he is therefore not likely to be its is impossible to judge whether or not it is author. Nagoji seems to be right in ascribing a genuine work by Vyadi. its authorship to Vyadi, who, besides being l op. cit. i. 1. 1. 6, vol. i, p. 104. quoted several times, is mentioned as an 2 op. cit. i. 1. 1. 8, vol. i, p. 136. ftcrtJT/a along with Sdkalya and Gurgya in Rpr. 3 op. cit. i. 1. 1. 1, vol. i, p. 6. xiii. 12. * Cf. Jaimini, PM. i. 1. 5. No MS. of the Samgraha, nor of any other 5 Cf.‘Bh&nda.rli&r, Wilson, Philological Lectures, work by Vyadi, has been so far discovered, p. 201. 206 EXEGETICAL AND CRITICAL NOTES [1. 10 With this may be compared the remarks of Plato in the Cratylus : l Soc.l And speech is a kind of action ?’ Her. ‘True.’ Soc.’ And will a mau speak correctly who speaks as he pleases ? Will not the successful speaker rather be he who speaks in the natural way of speaking . . . ?’ Soc. ‘And we saw that actions were not relative to ourselves, but had a special nature of their own.’ The question of the eternity of words was important not only to the grammarians and philologists, but also to the followers of the orthodox systems of philosophy. To their mind authoritativeness of the divine relation, i. e. the Veda, seemed to be involved and to depend on the solution of this principle. Hence the question forms a topic of discussion in almost every school of thought. It has been very fully treated by Jaimini in the first of his Purvamlmamsu. The arguments against the eternity of words are set forth in i. 1. 6-11 ; their refutation is contained in i. 1. 12-17; further arguments in support of the eternity of words are given in i. 1. 18-23. Some of the arguments are as follows : Words are Eternal. (1) Words are eternal because they are not uttered for the sake of utterance alone, but to express some meaning, and no meaning could ever be comprehended, had the words been transient. (2) Because everywhere there is a universal cognition of words. (3) Because the number is not used. When one word has been used several times, it is usually spoken of as having been used so many times and not that so many words have been used. This shows that the sameness or the unchangeable character of the word is recognized even when it is uttered or used after intervals ; hence words are eternal. Audumbarayana’s view as to the nature of words may be regarded as based on a doctrine akin to that of flux. With this may be compared the following remarks of Plato in the Cratylus : 2 Soc.’ I myself do not deny that the givers of names did really give them the idea that all things were in motion and flux ; which was their sincere but, I think, mistaken opinion. . . . Tell me, whether there is or is not any absolute beauty or good, or any other absolute existence ?’ Crat.’ Certainly, Socrates, I think so.’ Soc. ‘.. . . But let me ask whether the true beauty is not always beautiful.’ Crat. ‘Certainly.’ Soc.’ And can we rightly speak of a beauty which is always passing away ? . . . Must not the same thing be born and retire and vanish while the word is in our mouth ?’ 1 Jowett, Dialogues of Plate, vol. i, p. 327 2 Jowett, Dialogues of Plato, vol. i, pp. 387-8. (3rd ed.).2. i] EXEGETICAL AND CRITICAL NOTES 207
  Soc.’ Nor can we reasonably say, Cratylus, that there is knowledge at all, if everything is in a state of transition and there is nothing abiding.’ 2. 1.] As most of the MSS. do not mark the avagraha, the sign of a, it is not clear whether the reading represents yugapat or ayugapat. The few MSS. that do mark the avagralia are inconsistent, as they use it only occasionally. Their evidence is therefore not cogent. However, none of the MSS. mark the avagralia in this particular instance. Nevertheless I think the reading represents ayugapat. My reasons are as follows: (1) From the physical point of view, to produce simultaneously more than one sound is an impossibility ; yugapat would there fore convey no sense. (2) The context points to ayugapat, for the sentence is intended to show the absurdity of Audumbarayana’s view that speech is permanent only in the organ. But if we read yugapat the whole sentence becomes meaningless, for then the grammatical relation of the sounds which are produced simultaneously is possible, even if it is held that speech is permanent in the organ only. (3) This is in agreement with the view taken by Durga, who reads ayugapat and gives yugapat as a variant. (4) Another objection to the reading yugapat is that it would make grammar superfluous. When words are produced simultaneously their mutual connexion is simultaneous also. Durga remarks,’ The unconnected alone is connected (with something). In this case (i. e. reading yugapat) the roots are for ever connected with prepositions and affixes ; and affixes with elision, addition, and modification of letters/ Durga attributes the statement rf^ ^T<J5 ftTSf to the Purvapdksa and thinks the rejoinder to begin with the words <nffl*ne|i^Tj *i^t$l, a view which cannot be maintained. The sentence <T^ :^<p, &c., is a negative sentence ; it refutes the view of Audumbarayana. According to Durga’s interpretation, the controversy would be divided into three parts as follows : (1) Audumbarayana. Words are permanent in the organs only. (2) Parvapaksa. In that case the fourfold classification, &c., of words cannot be maintained. (3) Yaska’s rejoinder. On account of the pervasiveness and minuteness of words, the fourfold classification can be maintained. Durga thus introduces a third party into the discussion and tends to make out Yaska as supporting the view of Audumbarayana, by refuting his critic. I think there are only two parties, i. e. Audumbarayana and Yaska ; No. 1 gives the view of the former, No. 2 the rejoinder of the latter ; while No. 3 does not relate to this controversy at all, but deals with an altogether new topic, i.e. the superiority of speech over gestures, &c. Yaska’s rejoinder begins, therefore, with THf ^J?, &c.

2. 2.] It may be objected that there is no necessity for using words, for the meaning can very well be conveyed by means of gestures. To this Yaska replies that gestures, facial expressions, movements of hands and eyes, require 208 EXEGETICAL AND CRITICAL NOTES [2.2
   comparatively greater effort in their production, and are always indefinite. Their meaning can never be absolutely clear, it always involves some doubt. The word ‘minute* implies comparatively less effort and greater accuracy. Pataiijali refers to the same subject as follows :’ The meaning is clearly under stood when the word is uttered. “Bring the cow”, “Eat the curd”: these words having been uttered, the cow is brought, and the curd is eaten/ With this may also be compared the remarks of Plato in the Cratylus : 1 Soc.’ And here I will ask you a question : suppose that we had no voice or tongue, and wanted to communicate with one another, should we not, like the deaf and dumb, make signs with the hands and head and the rest of the body ?’ Her. ’ There would be no choice, Socrates.’ Soc.l We should imitate the nature of the thing ; the elevation of our hands to heaven would mean lightness and upwardriess ; heaviness and downwardness would be expressed by letting them drop to the ground ; if we were describing the running of a horse, or any other animal, we should make our bodies and their gestures as like as we could to them.’
3. 3.] Words are used to designate objects not only by men, but by gods also. Like Yaska, Plato also thinks that gods use words in giving names to things, and it follows therefore that the names given by gods would be the right names : 2 Soc.l He often speaks of them ; notably and nobly in the places where he distinguishes the different names which Gods and men give to different things. . . . For the Gods must clearly be supposed to call things by their right and natural names ; do you not think so ?’ Her. l Why, of course they call them rightly, if they call them at all. But to what are you referring ?’ Soc.’ Do you not know what he says about the river in Troy …” Whom “, as he says, " the Gods call Xanthus, and men call Scamander ‘V Her. ’ 1 remember.’ Soc. V- Or al>out the bil’d which, as he says, “The Gods call Chalcis, and men Cymindis”: . . . Or about Batieia and Myrina : " The hill which men call Batieia and the Immortals the tomb of the Sportive Myrina”/ If words are used with reference to gods human knowledge being neither perfect nor permanent mistakes are likely to be made in addressing and invoking gods, which will make them angry and render various acts of worship, like sacrifice, c., fruitless. In order to do away with such mistakes, injunctions, invocations, and hymns, &c., which are meant to complete and make worship fruitful, are laid down in the Vedas. A part of this paragraph is written in the sulra style, which style may therefore be regarded to have commenced about the time of Yaska. This is probably the earliest specimen of the sulra style. 1 Jowett, op. cit. vol. i, p. 368. 2 Jowett, op. cif., vol. i, p. 333. 4.i] EXEGETICAL AND CRITICAL NOTES 209
   Like Yaska, Jaimini also lays emphasis on the Veda as being the source of dharma : ‘Dharma is the object, the source of which is the Vedic injunction.‘1 And also :’ On account of the prescription of action being the object of the Veda/ 2
4. 3.] Of. Vajasaneyipratisakhya:3 ‘A verb denotes an action, and a preposi- tion makes that action specific.’ The Rgvedapratisdkhya : * 1 The prepositions are twenty ; with the other two (i. e. noun and verb) they express a meaning. They are : pra, abhi, a, para, nih, duh, anu, vi, upa, apa, sam, pari, prati, ni, ati, adhi, su, ud, ava, and opt.’ This list is identical with that of the Nirukta, except that they are enumerated in a different order. Cf. the Taittirlyapratitakhya5:’ a, pra, ava, upa, abhi, adhi, prati, vi, ni, are prepositions.’ It contains only half the number of prepositions. Cf. Panini : G 1 Pra, &c., are called prepositions when joined with verbs.’ The list of the pre- positions is given in the gana called pradayah, which is identical with that of the Nirukta, if the double forms of nih and duh, in the former, are not taken into consideration. The technical term used by Panini for preposition is karmapra vacanlya. He uses the word upasarga in a wider sense, i. e. covering both pre positions and adverbs. The various meanings of the prepositions are explained by Panini, i 4. S3-97.7 Cf. also the Atharvavedapratttakhya : 8 * Disjoined from the verb, however, are such as are used without significance . . .’ 9 4. 1.] Particles are classified by Yaska under three groups : (1) Particles of comparison, which are only four in number, and whose meanings and uses are illustrated by suitable examples. Yaska does not define the term upama (comparison) in this connexion, but discusses its meaning later on. 10 (2) Particles which denote karmopasamgraha, Yaska’s explanation of which term is not clearly expressed, nor does Both 11 make it more lucid. The obscurity has led Durga to misinterpret the sentence. He takes karmopasamgraha as equivalent to samuccaya, i. e.’ aggregation’; this is only one of the meanings expressed by the said term, which is very comprehensive. It is quite obvious that the term is designed to cover all the meanings expressed by the particles enumerated in the second group, i. e. those beginning with ca and ending with 1 PM. i. 1. 2. of the Nirukta. Op.cii., L 2. 1. Questions like the eternity xii. G : 702. of fabda, the impermanence of human know- 5i. 15. ledge, the infallibility of the Veda, form topics 6i. 4. 58. of discussion in almost every system of philo- 7 Cf. Patanjali, i. 4. 4 : vol. i, pp. 341, sophy ; cf. VSu. i. 8. 28 ; i. 4. 28 ; ii. 1. 4 ; ii. 345-9 ; see foot-note (in the text). 4. 20 ; NSu. i. 1. 7 j ii. 1. 54-7 ; ii. 2. 18-17 ; 8 iv. 8. Hi. 2. 49 ; VaiSu. ii. 2. 21-87 ; vi. 1. 1 ; z. 2. 9 ; 9 Translated by Whitney, J.A.O.S., vol. vii, SpSu. v. 45-51 ; v. 37-40 ; v. 58-61. p. 515.

• viii. 54. The list of the prepositions is " N. 8. 13. given in vi. 24, which IB identical with that " Erl&utsrungcn, p. ft,

210. EXEGETICAL AND CRITICAL NOTES [4. i
     tea. Gune translates the sentence yasyagamat, &c., as follows : 1 ’ Owing to whose advent (i. e. use) separateness of the ^T^ (senses or ideas) is indeed known, but not as in simple enumeration owing to separate position or inde- pendent mention, that is quJTmm^, i. e. adding or putting together of the senses or ideas.’ One must admit that the meaning of the term is not quite clear. One does not know what is the precise difference between ‘simple enumeration* and ’ putting together ‘. My translation of the sentence is the following :’ That by whose addition separateness of notions is indeed recognized, but not as an enumerative one, i. e. on account of a separateness by isolation, is a conjunc tion.’ I think l conjunction’ is the nearest corresponding word in English which will cover all the meanings expressed by the particles of the second group, and which is therefore equivalent to karmopasarrigraha. For example, in the sentence, ^fTf gftiRflHll Pt<V||ftf ^f ^T I ’ Ha ! * wil1 Put this earth here or there’,8 separateness of place is recognized by the use of the particle va ; the repetition of ilia by itself would have failed to convey the idea of different places. This notion of separateness is expressed, not by a categorical enumeration as horse, cow, man, elephant, &c., but by the use of the particle. Again, in the sentence, ^qW^^ <0tf^fK1,’ Let one mando this, the other that , 3 two distinct actions are mentioned, the notion of whose distinctness is conveyed by the particle alia. The particles of the second group are explained with appropriate examples in the Nirukta, i. 4-9. With this definition of a conjunction may be compared the following remarks of Aristotle: 4 ‘A conjunction is (a) a non-significant sound which, when one significant sound is formable out of several, neither hinders nor aids the union, and which, if the speech thus formed stands by itself (apart from other speeches), must not be inserted at the beginning of it, e. g. /,/, S//, rot, 8e ; or (&) a non-significant sound capable of combining two or more significant sounds into one ; e. g. d/u^c, TTfpl, &C. (3) Particles which do not express any meaning, but are merely used to fill up a sentence in prose or verse, are expletives. They are enumerated towards the end of the ninth section ; they are four in number : kam, lm, id, and u ; but later (in the tenth) iva is added to this list. Quotations showing their uses are cited and explained in the tenth section. The meaning of id, when com bined with na and na ca, is discussed and illustrated by suitable examples in the eleventh section. Thus id combined with na is used to denote apprehension : * lest we should ‘, &c. ; id combined with na ca is used in asking questions : don’t they drink ?’ The above-mentioned list of the particles of the third class is by no means complete. Some particles of the second group, i. e. u, 1 1A., vol. xlv, p. 159. 3 N. 1. 5. 2 N. 1. 4. 4 Poetics, 20. 1467* (ed. Bywater), p. 59.
211. a] EXEGETICAL AND CRITICAL NOTES 211 nunam, are occasionally used as expletives, and slm in Vedic Sanskrit. Sim was originally the accusative singular of a pronoun, related to sa, as Jam to Ita. It appeara in the Rgveda as an enclitic particle.1
212. 8.] Roth thinks that Yaska explains nunam by adyatanam, ’to-day’, and remarks,’ J. hat darin Unrecht *. 2 This view is erroneous. According to Yaska, the particle nunam is used (1) to signify ‘uncertainty ‘, and (2) as an expletive. These two uses are exemplified by two respective quotations, the former illustrating its meaning of ‘uncertainty’, the latter showing its use as an expletive. Hence it is that, in the latter case, Yaska neither repeats nor para- phrases nunam by any other word in his commentary, indicating thereby that it is an expletive, while in the former quotation nunam is repeated by Yaska in his explanation, showing thereby that it has the sense of * uncertainty ‘. The word adyatanam is not intended to paraphrase nunam, but is added for the sake of amplifying Indra’s brief statement. 12.2.] Cfc the following passages: AA. ii. 1. 3: ’ Now, therefore, the origin of seed. . . . Heart is the seed of people, mind is the seed of heart, speech is the seed of mind, and action is the seed of speech.’ AA. ii. 1. 6: $ 3Nll^d I ’ Speech is his thread, names the chords. All this is woven by this speech of his, in the form of thread and names as chords. For all this is names, and by his speech he names everything.’ Of. also Ch. U. vi. 5. 1-4 ; 6. 1-5 ; vii 2. 1 ; vii. 12. 2. See Keith, A. A. Translation.
     Sakatayana : a famous ancient grammarian, quoted in RV. Pra. i. 13, 17 ; xiii. 16, 747 (Max Mlillers ed., pp. 13 and 271) ; in Vaj. Pra. iii. 8, 11, 86; iv. 4, 126, 188 ; in A. Pra. ii. 24 ; in Astadhyayl iii. 4, 111 ;. viii. 3, 18 ; 4, 50 ; and in N. 1. 3, 12. No work of his has been preserved The Sabdanusasana quoted by Vopadeva and various other writers is the work of a modern Jain grammarian called Sakatayana. 3 Aufrecht regards him as the author of the Unadisutras. They, however, cannot be the work of the ancient Sakatayana, for they clearly bear a modern stamp. Belvalkar 4 attributes the Unadisutras to Panini. His argument is that ’ they use samjiids such as hrasva, dtrgha, pluta, uddtta, lopa, wmprasarana, and abhyasa in the same sense in which Panini uses them ‘. This airgument is inconclusive because, the works of Panini’s predecessors being lost, we have no means of judging whether or not he is indebted to them for those terms. Some at least of these saihjnas or technical terms were derived from a common stock ; Yaska himself, for instance, makes use of a few of them, as 1 See Professor Macdonell, A Vedic Grammar modern Sakatayana, see IA., vol. zliii, pp. for Students, pp. 249, 452. 205-12. 2 op. ctX, p. 6. * Systems of Sanskrit Grammar, p. 25. 3 On the authenticity and date of theo2 212 EXEGETICAL AND CRITICAL NOTES [12. a lopa, abhyasa, &c. Again, in some cases Panini’s teaching, as pointed out by Belvalkar himself, runs counter to the Unadisutras.1 Thus to ascribe the authorship of the U. Sutras to Panini does not rest on firm ground. Moreover, Patanjali’s defence of Panini against the adverse criticism of Katyayana2 is that ’the Unadi words are crude forms of nouns which are not derived’. 3 This statement implies that Patanjali did not regard all nouns as derivable fromverbs. He also attributes the same doctrine to Panini, and Patanjali, I think, cannot be accused of not knowing well, or misunderstanding Panini. So, according to this testimony, both Panini and Patanjali were the followers of the school of Gargya. The Unadisutras, on the contrary, are the product of the school of dakat&yana, whose fundamental doctrine was that all nouns are derived from verbs ; they cannot therefore be attributed to Panini. It is probable that, in their original form, they were written by Sakatayana, but were extended and modified by subsequent writers, and in spite of their modernness still bear traces of their ancient origin. Gargya: an ancient grammarian, quoted by Yaska i. 3, 12; iii. IB, andPanini, Jv. 1. 106 ; vii. 3. 99; viii. 8. 20 j 4. 67, and mentioned by Durga as the author of the padapatha of the Samaveda. No work of his has survived. Both Panini and Patanjali seem to be his followers, for they regard the Unddis as under!vable. It is therefore no wonder that his work, after the appearance of the AstOdhyayi, has not survived.
213. 3.] The sentence TRPI ^<4l4ft 1 ^Mt $^ftf?f which is somewhatdifficult, is differently interpreted by various writers. The crux lies in the word sam-vijnatani. Durga paraphrases this word as follows : samam vijnatani aikamatyena vijnatani,’ discriminated unanimously ; i. e. discriminated withabsolute agreement ‘. Max Mtiller * translates it by’ intelligible ‘, Both by’arbitrarily named ‘. 5 Roth’s translation seems to be based on Burga’s second explanation of the same term, which is as follows: sam-wjnana-padam itlha fastre rudhi-$abdasyeyam samjna : t in this (branch of) knowledge, the term sam vijnana is a technical expression used for a conventional word.’ Durga resorts to the Comparative Method and quotes : tany apy eke samamananti . . . sam-vijnana bhUtam syatf in support of his explanation. He is further corroborated by a comparison of all the passages of the Nirukta in which the word sam-vijnana or (with the omission of the prep, vi) sam-jnana occurs.7 We may therefore take the word to signify’ a conventional term ‘. The next problem in the sentence is the punctuation. Max Mttller 8 takes sam-vijnatani, &c., as the principal clause to complete the relative clause tad yatra 1 Panini’s sutra vi. 2. 139 is opposed to 4 History of Ancient Skt. Lit., p. 165. -USu. iv. 226. 5 op. cit, p. 9, ‘willkurlich benannt’. 1 On the relation of the three grammarians, 6 N. 7. 13. Me Kielhorn, Panini, Kdtydyana, and Patanjali.T Cf. Gune’s note, IA., vol. xlv, p. 173. * The Hahdbhdtya, viii. 1. 1, vol. iii, p. 241. ’ op. cit., p. 166.
214. 3] EXEGETICAL AND CRITICAL NOTES 218 . . . syatam, and translates as follows :* For first, if the accent and formation were regular in all nouns and agreed entirely with the appellative power (of the root), nouns such as go (cow), ava (horse), purusa (man) would be in themselves intelligible/ He succeeds in thus construing the sentence by translating yatra by’ if’ ; leaving out torn ; and by attributing to sam-vijjidtani a meaning not borne out by the comparison of passages. Roth divides the sentence by placing a semicolon after syatum and takes sam-injnatani, &c., as a co-ordinate clause ; but in order to connect the two clauses, he supplies the word dagegen. Durga offers two interpretations. Firstly, he places a full stop after tani and takes the words yatlia gaur ava. &c., as a co-ordinate clause, supplying, however, the words na punah ; the translation of the sentence according to this interpretation would be the following: ‘The words whose accent and grammatical formation are regular and which are accompanied with an explanatory radical element are unanimously recognized to have been derived from roots; but not words like “cow”, “horse”, “man”, “elephant”, &c.’ Secondly, he places a full stop after syatdm and takes sam-vijndtdni, &c., as an independent sentence. According to this division, the first sentence would consist of one single relative clause, without any principal clause. To meet this difficulty he remarks :’ Tad dJchydtajam gundkrtam iti pratlma iti vdkya ScsahS ‘We think that the words, “that is derived from a verb”, must be supplied as a supplementary clause.’ The translation according to this interpretation is the following :’ Those words whose accent and grammatical formation are regular, and which are accompanied with an explanatory radical element, are derived from roots. Words like " cow “, " horse “, " man “, " elephant “, are conventional terms.’ Gune does not seem to be aware of this second interpretation of Durga and independently arrives at a conclusion l identical with that of Durga, and suggests the adoption of the supplementary words : satvam tat pradeSikam.* These words occur in Yaska’s rejoinder in section 14. His argument is that, in his rejoinder, Yaska always first repeats the words of his opponent and then answers ’the objection. According to Gune, the sentence placed within the words yatho etad and iti exactly represents the original statement of the critic. The sentence placed within these words in Yaska’s rejoinder in section 14 contains the supplementary clause sarvam tat prdde&kam, which, being thus assigned to the critic by Yaska himself, must have formed a part of the sentence under discussion. He remarks, ‘And we are also sure, comparing the initial passage [i. e. in the purvapaksd] with its counterpart in Yaska’s reply at R. 36. 10, that *pf TTft Hl^filHiH must have been there. Its omission is strange and unaccountable. Perhaps it is the scribe’s mistake. . . .’ 3 In other words, Gune thinks that the passage in Yaska’s rejoinder could be used as a MS. (archetype). furnishing evidence .which cannot be challenged, for the critical edition of the 1 1A., loc. tit. 2 N. 1. 14. 3 IA., foe. cit. and p. 174. 214 EXEGETICAL AND CRITICAL NOTES [12.3 original passage of the critic in section 12. A closer examination, however, does not support this theory, for a comparison of the statements of the critic with those assigned to him by Yaska, in his rejoinder, shows that Yfiska repeats, and puts between yatho etad and iti, only so many words of his opponent as are necessary for the controversy. He does not repeat them in toto. It is evident, if one compares section 13 and section 14 ; Purvapaksa. Yaska’s rejoinder^.-…
215. 13: ^rerft ^ T^IT tU4|qi*t|hI- 1. 14: Zftft In both these cases, Yaska repeats only a part of his opponent’s statements. Gune’s assertion is therefore unfounded ; hence his suggestion as regards the adoption of a supplementary clause cannot be accepted. The sentence can, however, be explained without having recourse to an assumed interpolation. The difficulty will disappear if a full stop be placed after syatam and the passage na sarvanlti . . . syatam be construed as one sentence. I would then translate :’ Not all the words,’ say Gargya and some other grammarians,’ but only those, the accent and grammatical form of which are regular and which are accompanied by an explanatory radical element. Those such as “cow”,“horse”, “man”, “elephant”, &c., are conventional terms.’ Durga has the following theory about nouns :’ There is a threefold order of nouns; i. e. (1) those whose roots Are apparent; (2) those whose roots can be inferred ; (3) and those whose roots are non-existent. With reference to this, the nouns whose roots are apparent are such as “doer”, “bringer”, &c. Nouns whose roots can be inferred are such as “cow”, “horse”, &c. Nouns whose roots are non-existent are such as dittha, davittha aravinda, and varvinda, &c.’ It is quite evident that Yaska, a follower of the school of etymologists whose fundamental doctrine is that all nouns are derived from roots could not have recognized the third category of his commentator, who thus appears to be a follower of the school of Gargya. 14.] With the Nirukta controversy about the origin and nature of names may be compared a somewhat similar discussion in the Cratylus, where Plato propounds three theories, represented by the three characters of the dialogue. He puts in the mouth of Hermogenes the doctrine that names are conventional. The opinion of Cratylus that names are natural is diametrically opposed to this. Socrates takes an intermediate view. He refutes the view of Hermogenes that names are given arbitrarily and altered at will : 14] EXEGETICAL AND CRITICAL NOTES 215 Soc.’ Well, now, let me take an instance. Suppose that I call a man a horse or a horse a man, you mean to say that a man will be rightly called a horse by me individually, and rightly called a man by the rest of the world ; and a horse again would be rightly called a man by me and a horse by the world : That is your meaning ?’ l He then expounds the principle of the natural correctness of names. Soc. ‘Then the actions also are done according to their proper nature, and not according to our opinion of them ? In cutting, for example, we do not cut as we please . . . but we cut . . . according to the natural process of cutting ;’. . . Her. ’ I should say that the natural way is the right way.’ . . . Soc.’ And this holds good of all actions ?’ Her. ‘Yes/ Soc.’ And speech is a kind of action ?’ Her. ‘True.’… Soc.’ And is not naming a part of speaking ? for in giving names men speak. Her. ‘That is true.’ Soc. ‘. . . Is not naming also a sort of action ?’ Her. ‘True.’ Soc.’ And we saw that actions were not relative to ourselves, but had a special nature of their own ?’ Her. ’ Precisely.’ Soc.’ Then the argument would lead us to infer that names ought to be given according to a natural process, and with a proper instrument, and not at our pleasure : in this and no other way shall we name with success.’ 2 Socrates also admits the element of convention in names : Soc.’ Yes, my dear friend ; but then you know that the original names have been long ago buried and disguised by people sticking on and stripping off letters for the sake of euphony, and twisting and bedizening them in all sorts of ways : and time too may have had a share in the change.’ . . . 3 Also: Soc.’ To say that names which we do not understand are of foreign origin ; and this is very likely the right answer, and something of this kind may be true of them ; but also the original forms of words may have been lost in the lapse of ages ; names have been so twisted in all manner of ways, that I should not be surprised if the old language when compared with that now in use would appear to us to be a barbarous tongue.‘4 Cratylus maintains that names are either true or not names at all : Orat.’ Very true, Socrates ; but the case of language, you see, is different ; for when by the help of grammar we assign the letters a or Z>, or any other letters, to a certain name, then, if we add, or subtract, or misplace a letter, the 1 Jowett, Dialogues of Plato, Oratylw, p. 886. * Ibid., p. 414.

• Ibid., p. 887. * Ibid., p. 421. 216 EXEGETICAL AND CRITICAL NOTES [14 name which is written is not only written wrongly, but not written at all ; and in any of these cases becomes other than a name.’ . . . Soc.’ I believe that what you say may be true about numbers, which must be just what they are, or not be at all ; for example, the number ten at once becomes other than ten if a -unit be added or subtracted, and so of any other number: but this does not apply to that which is qualitative or to anything which is represented under an image/ l As regards the derivation of names : Soc.* All the names that we have been explaining were, intended to indicate the nature of things/ Her. ’ Of course.’ Soc.’ And that this is true of the primary quite as much as of the secondary names.’ . . . Soc. ‘But the secondary, as I conceive, derive their significance from the primary/ Her. * That is evident/ Soc. ’ Very good, but then how do the primary names, which precede analysis show the nature of things ?’. . . Soc. ‘But how shall we further analyse them? . . . Ought, we not, therefore, first to separate the letters, just as those who are beginning rhythm first distinguish the powers of elementary, and then of compound sounds P’. . .* Plato’s doctrine,* that names rightly given are the likenesses and images of the things which they name ‘, 3 does not hold good, for in a large number of cases names are not the images of things themselves, but of our concept of them. It was Locke who first pointed this out, and Max Mtiller amplified his argument as follows: ‘Each time that we use a general name, if we say dog, tree, chair, we have not these objects before our eyes, only our concepts of them ; there can be nothing in the world of sense corresponding even to such simple words as dog, tree, chair. We can never expect to see a dog, a tree, a chair. Dog means every kind of dog from the greyhound to the spaniel ; tree, every kind of tree from the oak to the cherry; chair, every kind of chair from the royal throne to the artisan’s stool. . . . People often imagine that they can form a general image of a dog by leaving out what is peculiar to every individual dog.’* In an elaborate discussion, Madhava brings together the views of various grammarians as to the meaning of words and its cognition: Vajapyayana* maintains that all words express a generic meaning and a particular substance is apprehended after the apprehension of the genus which has an intimate relation with them. Vyadi maintains that words mean individual things, that they explicitly express the individual substance, while the genus is implied. 1 Jowett, Dialogues of Plato, Cratylus, p. 378. * Science of Thought, pp. 77, 78, cited by * Ibid., pp. 867-70. Moncalm. 3 See ibid., p. 387.16. i] EXEGETICAL AND CRITICAL NOTES 217
  Pfinini, he says, accepts both views. He attributes to words a generic meaning,for he says the singular is used to denote the class, while the plural may be optionally used (Pa. i. 2. 58) ; on the other hand, by i. 2. 64, his acceptance of the individualistic theory is quite apparent (Sarvadarana-sarngraha, p. 145). But the next question is, how does the cognition of the meaning,whether generic or individualistic, take place ? For instance, when the word* cow’ is pronounced, there is a simultaneous cognition of dewlap, tail, hump,hoofs, and horns. Is this cognition produced by the single letters composingthe word cow ‘, or by their aggregation ? The first alternative is not tenable, for the cognition of the object cannot be the result of a string of separate andindividual letters without some unifying cause, as a garland cannot be madefrom a collection of separate flowers without the unifying string. The second alternative is impossible, for there can be no aggregation of sounds, each of which ceases to exist as soon as it is pronounced. If you attribute a’ manifesting* power to the letters, this power can be exercised only in succession and notsimultaneously. Moreover, if each letter has a separate manifesting power, then the pairs of words rasa, sara ; vana, nava ; nadi, dlna ; rama, mara ; raja, jara, &c., since the pairs have the same number of the same letters, should each have had the same meaning. Moreover, the baseless assumption of aggregation would involve mutual dependence. Since it is impossible that letters shouldconvey the meaning by themselves, we have to accept the hypothesis of aunifying factor which is all-pervading and whose existence is independent of letters, and which is technically called sphota (Sarvadarfana-sam., Bib. Ind., pp. 140-4).

16. 1.] This is a very important statement made by Y&ska. He is fully aware of the close relationship of the Yedic language with the lhasa of his owntimes, which could be no other than the classical Sanskrit Further, he seemsto be conscious of the historical growth of the latter from the former, when hesays that the nouns of the lhasa are derived from Vedic roots. It is needless to point out that it completely answers Kautsa’s criticism that the Vedas are meaningless ; for many words which they daily use in the Wiasa are identical with those used in the Vedas, and if in the Uulxa they do express any meaning,they must do so in the Veda. It would be absurd for Eautsa to deny that the words in the spoken language express a meaning. Thus he is constrained to admit that the Vedas have a meaning. A similar objection appears in Sayana’s preface to the Rgveda : From all these reasons, (it is clear) that there is no intelligible significance in the Vedas.’ One of the examples here given is as follows :’ Suppose a womannamed Parhika is plying the pestle. A Brahmana boy called Manavaka is com mitting to memory a pestle-stanza, just close to her. The recitation of the stanza does not keep time with the fall of the pestle at every stroke. No mean ing is conveyed to the woman, nor is there any intention to convey any meaningto her. Similarly, at the performance of sacrifices, the recitation of Vedic stanzas 218 EXEGETICAL AND CRITICAL NOTES [16. i
    does not convey any meaning/ In answering this, Sayana quotes (PQ. Mr. i. 1. 31) :’ But the meaning of (Vedic) expressions is not different,’ his comment being qir4iu ^qiq^^Ti \fafais: i The meaning of expressions of the Vedic Sanskrit and of the popular speech is not different.’ As to the example, the Brahmana boy does not intend to convey any meaning at the time of learning his lesson, and the woman therefore does not understand, a woman, moreover, being incapable of understanding the Veda: at sacrifices the priest does intend to convey a meaning, and others do understand it. And if at any particular time there is no intention of conveying a meaning, it would be altogether unjustifiable to conclude that there is no meaning at all. 16. 4.] The objection is that because the Brahmana text prescribes the com plete form of the sacrificial acts, the Vedic stanzas are superfluous, the only use being their mere recitation. Yaska’s reply is that the Brahmana text merely reiterates what is enjoined by the Veda. Sayana here quotes: ’ because it is mentioned in the Vedic stanzas ‘, and his comment is : T jf^fa $ft r ^rR^rata wpufli T w^fa ^ * iro^rafa i ‘just as in popular speech, he who says” do this “is the causer of the performance of the action, so in this case too, he who says” spread"is the prescriber of the act of spreading ‘. And to suppose that the use of the Vedic stanzas consists solely in recitation would be attributing a transcendent importance to them. We need not go so far ; for ‘, says Sayana,’ their non-transcendent use is the expression of meaning’: -isnfrern!I^[ <J ^fS ^&ft I
17. 5.] Cf. Sayana’s preface (p. 3): H ffifiRfd U^fa^n I I cflft foMOdl^^^^H^^T^+C I’ The stanza, “O herb, save him”, is about grass. The stanza, “O axe, do not injure him “, refers to an axe. … In these stanzas, inanimate objects, grass, axe, and stones, are addressed like sentient beings. Therefore the Veda, on account of such absurd invocations which are contrary to common sense, is of no authority/ He further quotes the Satra (Pu. Ml. i. 2. 35), ’ because a meaning is attributed to lifeless things ‘, and his comment is that lifeless things^euch as grass, stones, and axe are described as if they were living beings and had the powers of saving, injuring, hearing, &c. The answer is as follows: The inanimate things themselves are not addressed in these Vedic stanzas, but their immanent deities, which have been treated at length by Badarayana. He further quotes, ^rfTOT%J^l^ ‘There is a figurative description in such expressions.’ His comment is: ‘This is very frequently employed in poetical compositions. For instance, a river is described as having a pair of cakravaka birds for her breasts, a row of swans for her teeth, a tea plant for her garment, and moss for her hair. Similarly, the Vedic texts invok ing inanimate objects should be construed as implying praise. If by cultivation
18. i] EXEGETICAL AND CRITICAL NOTES 219
    the plant will protect, muck more so will the cultivator ; if even the stones listen to the morning recitation of the Vedic texts, how much more will the learned Brahmanas.
19. 6.] Of. Sayana’s preface (p. 8) : ‘The Veda is not authoritative, because statements like “One Rudra alone, &c.” are contradictory, as if one were to say that he is observing the vow of silence for life. . . .’ ‘There is no such contradiction, because even one Rudra by his greatness can take on a thousand forms.” He further quotes the fifth Sutra: ’ On account of the figurative description, there will be I no contradiction ’ (ibid. i. 2. 47).
20. 8.] Of. Sayana’s preface : i ….

• Now some of the stanzas convey no meaning . . . amydksa . . . yadrmin, . . . &c. These stanzas express no meaning at all.’ To this Sayana replies by repeating Yaska’s sentence that it is not the fault of the post if the blind mandoes not see it. He further quotes: WC MMJfafJMH, ‘The meaning exists, but it is obscure* (ibid. i. 2. 49). His comment is: : |’ The meaning does exist, but it is not recognized by people on account of their neglect, laziness, &c. Their meaning should be inferred from the root in accordance with etymology and grammar and parallel passages.’ Then he explains jarbhafi and turpharitu as names of the Asvins.

17. 1.] After pointing out the great importance of etymology for an accurate analysis of words into their constituent elements, YasT :a incidentally defines samhita as’ the closest conjunction (of original words) by means of euphonic combination ’ or as’ based on original words ‘. By means of using certain words, Yaska always gives a general indication of the source of his quotations. If the quotation is from the Vedas or Samhitas like the MS. KS., &c., he regularly uses the phrase athapi nigamo bhavati ; if it is from the Brahmanas, he uses the words iti viynayate, iti ca, &c. Quotations from other sources are similarly indicated; thus two quotations one from some metrical law-treatise are intro- duced with the words : tad etad rfahalokabhyam dbhyuktam (see 3. 4), but he nowhere cites without acknowledgement ; as his definitions of Samhita are not qualified with any such phrase, I take them to be Yaska’s own. The first is 220 EXEGETICAL AND CRITICAL NOTES [17. i
    adopted by Panini (L 4. 109), the second by the RV. Pra. (u. 1): Yaska is con- sequently earlier than Panini and the extant R. Pratttakhya. It is, however, very likely that the Pratisakhyas were known to Yaska in an earlier form. Probably the sentence,* that the grammatical treatises (parsadani) of all the different schools are based on the original forms of words ‘, alludes to the Pratisakhyas, these being the oldest grammatical treatises. Sometimes the words parsada and pratttakhya are interchanged, as is shown by the evidence of a MS. in the Bodleian, which uses the word parsada in the place of pratUakhya. This leads to the conclusion that Yaska knew some pratiSakhyas, although he is earlier than the modern R. Prati$akhya.
18. 6.] Yaska here intends to point out the practical utility of etymology for the performance of sacrificial rites. For the success of sacrifice, it is of the highest importance to know the deity of a stanza. Sometimes the general principle that a stanza belongs to a deity whose characteristic mark it bears is not applicable, for instance, to a stanza which bears the characteristic marks of more than one deity. In such cases, one has to find out which deity is primarily and which is incidentally mentioned, and this, Yaska implies, can only be done with the help of etymology ; hence the importance- of etymology for practical purposes like the performance of sacrifices.
19. 17.] The last point adduced by Yaska in favour of the science of etymology is that knowledge should be acquired for its own sake. He quotes two stanzas from the $gveda to show that it is through knowledge that one is purged of his defects, attains an unassailable position, and all the joys and bless- ings of this world and the next. His arguments in favour of etymology may be summarized as follows : (1) Etymology is the complement of grammar and is therefore essential for understanding the meaning of the Vedic texts. (2) It is essential for an accurate analysis of words into their constituent elements. (3) It is of great importance for the performance of sacrifice, for it enables one to find out precisely the deity of a particular stanza, and to recite the appropriate texts while offering oblations to various gods. (4) Lastly, knowledge for its own sake is commended and ignorance con- demned. 18.] Roth thinks the whole of the eighteenth section is an interpolation. The second stanza is quoted by Patanjali in the Mahabhasya.1 If Roth is right, the interpolation is an old one. 2O. 10.] The paragraph traces the origin and the necessity of compiling the list of words called the Nighantu. The hymns were revealed by direct intuition to the primeval bards. They handed them down to their successors by oral instruction. Later generations being devoid of direct intuition, and being weary of oral instruction, compiled with a desire to facilitate the study of the Vedas, 1 See foot-note in the text.
20. lo] EXEGETICAL AND CRITICAL NOTES 221
    the Nighantu, the Veda, and the Vedahgas. The last sentence attributes the com pilation of the Nighantu, the Veda, and the Vedahga to the same period. Yaska, however, makes it clear that by compilation of the Veda he does notmean the composition of the Vedic hymns, which, he says, were revealed bydirect intuition. Yet what he intended to convey by the ’ compilation’ of the Veda is not clear. According to Durga it refers to the growth of the branches(6dkha) of the Veda. But it more probably refers to the constitution of theSamhita text and the arrangement of the hymns of the Rgveda into ten books. Yaska several times refers to the ten books of the Rgveda with the termdaatayrisu ; and he expressly states that it was the hymns and not the Samhitatext that was revealed to the primeval seers. This definition of Samhita, more over, indicates that he regarded it as a later production. Thus it is likely thatby’ compilation of the Veda ’ Yaska means the constitution of the Samhita text. The use of the singular number here is significant. It would be equally interesting to ascertain what Yaska meant by theVedangas. He could not possibly mean the six traditional Vedangas for the simple reason that some of the Vedangas, like Panini’s Astadhyaifi, are much later than Yaska and therefore could not have been known to him.Secondly, though Yaska’s Nirukta itself is a Vedahga now, he himself couldnot have been so presumptuous as to regard his own commentary as aVedahga. Nor did he regard the Nighantu as a Vedahga, for he refers to it as ’the list’, and its separate mention in the sentence; ’later generations. . . compiled the Nighantu, the Veda, and the Vedahga ‘, suffices to show thatto Yaska the Nighantu meant something different from the Vedangas. Yaskaquotes from many Brahmanas, and he was obviously familiar with most ofthem. The large number of Brahmana quotations in the Nirukta indicates thatin Yaska’s time the Brahmanas were popular treatises, not only for the per- formance of sacrifices, but as handbooks of the common stock of wisdom. Fromtheir auxiliary character in the study of the Veda, it might be argued that inYaska’s time the Brahmanas were included in the Vedahgas. This argumentacquires some plausibility from the fact that in the above sentence Yaska saysnothing about the compilation of the Brahmanas, thus implying that they areVedahgas. This inference is, however, invalidated by Yaska’s practice ofregularly mentioning these works by their proper name, i. e. Brahmana. Nordoes he anywhere give any hint whatsoever that their study is essential for theproper understanding of the Vedas. But on the other hand he does say thatwithout etymology it is not possible to understand the meaning of the Veda.Etymology, therefore, is an auxiliary science for the study of the Veda. InYaska’s time there was an established school of etymologists, whose opinionshe quotes on more than twenty occasions. The standard treatise of this schoolwould therefore constitute one of the Vedahgas. Yaska further says thatetymology is the complement of grammar. In his view grammar is thereforeequally important for understanding the Vedas. Besides individual grammarians, 222 EXEGETICAL AND CRITICAL NOTES [20. 10 Yaska on several occasions also refers to the school of grammarians. Their standard work would thus be another Vedanga. In his discussions Yaska moreover quotes the opinions of the two schools of ritualists, the older and the younger. Their most authoritative work or works would form another Vedahga. Two other schools, those well versed in legendary lore and in Vedic metres, are also quoted. Their standard works would form additional Vedangas. These would give rise to the later Vedangas and subsequently, after the time of Panini, become stereotyped in the most authoritative surviving work of each school. Astronomy would, as less essential, have been added later, and the number of Vedangas finally accepted as six. 2O. 12.] This gives a general description of the contents and the main divisions of the Nighantu. The first part deals with the synonyms both of nouns and verbs ; the second with homonyms ; the third with deities, whether primarily or incidentally mentioned. Then comes the explanation of the primary and incidental mention and the definition of the daivata. CHAPTER II 1.] The MSS. of the longer recension and the commentator read gunena, while those of the shorter one vikarena. Both are consistent in their reading, for they have their respective reading in N. 1. 14, the greater part of the first sentence in which and also its counterpart in section 12 is identical with the passage under discussion. The authority of the commentator no doubt throws greater weight on the side of gunena, which, on that ground, was adopted by Samasrami, who, finding that four of his MSS. read vikarena,’, remarks in the foot-note,1 param vrtti-viruddhah, and ignores it without any further comment, as if that fact alone were sufficient for its condemnation. Roth does not seem to have any knowledge of the variant vikarena, for he does not mention it in his list of Abwetehende Lesungen der Mrzeren Recension? The variant vikarena, however, gives a better meaning. If we adopt the reading gunena, the transla- tion of the sentence would be the following :’ With regard to those words, the accent and grammatical formation of which are regular, and which are accom- panied by an explanatory radical element, (we say that) they should be explained in the ordinary manner.’ To a follower of the school of etymologists the phrase underlined would be meaningless, for, according to their fundamental doctrine, every word is accompanied by its radical element. Yaska, in this case, would be placed on the horns of a dilemma: either he must give up the fundamental doctrine of his school, or recognize the absurdity of his sentence. The meaning is considerably improved by adopting the reading 1 Bil. Ind., vol. ii, p. 147. * pp. 105-12.
21. TO] EXEGETICAL AND CRITICAL NOTES 223 viMrena ; the translation of the sentence in this case is the following :* Withregard to those words, the accent and grammatical formation of which are regular, and which arc accompanied by an explanatory radical modification, (we say that) they should be explained in the ordinary manner.’ Y&ska in this case is saved from the above dilemma. Further, in the immediately following sentence vikara actually makes its appearance and is coupled with pradctika also. In myopinion rikarena is therefore the original reading. 2.] The meaning of Yaska, when he says that some Vedic nouns are derived from the roots of the classical Sanskrit and vice versa, is not quite clear. Atfirst sight he would here seem to regard the Vedic and the classical Sanskrit as two distinct languages with two different sets of roots, which in some cases havemutually influenced each other. But from his remarks in the first chapter it is evident that he is conscious of the close relationship between the Vedic and the classical languages. To Kautsa’s criticism that the Vedic hymns are meaning less, Yaska replies’ that they are significant because their words are identical with those of the spoken language. He notices 2 the uses of prepositions andparticles common to the two languages. He seems to realize as well that the one is historically the outgrowth of the other when he points out that the meaning of a particular word in the classical Sanskrit is only an extension fromthat of the Vedic language. In fact, as expounder of the Nighantu and the commentator of about 600 Vedic stanzas, he could not have failed to observe the close affinity of the Vedic and classical Sanskrit. Considering these facts, the distinction made by Yaska between the roots of the Vedic and the classical Sanskrit would seem inexplicable. With a view to reconcile these apparently contradictory views, I propose to interpret the passage as follows : the statement that Vedic nouns are derived from classical roots means that the roots fromwhich those particular words are derived do not occur in verbal forms in the Veda, but survive in the classical language only. Agni is called damunas in the Vedic hymns. Yaska would no doubt have derived the word damunas from the root damt’ to become tame ‘. He would say that although the root dam is still used in the sense of * becoming tame ’ in classical Sanskrit, it does not occur in this sense in the Vedic language. Similarly, when he says that classical wordslike usnam, ghrtam, &c., are derived from Vedic roots, all that he means is that the roots us and ghr are used in their respective meanings* to burn ’ and * to drip’ in the Vedic language only, and that they have lost these meanings in the classical language.
22. 10.] Yaska here notices the dialectical difference of the spoken Sanskrit, or what otherwise may be called provincialisms. He divides people into those who employ primary forms and those who employ secondary forms. Accordingto this distinction the Eambojas and the Easterners use primary and the Aryasand the Northerners derivative secondary forms. Yaska differentiates the Aryas from the Easterners and the Northerners. This shows that the Easterners 1 See I. 14. 2 See 1. 3-4. 224 EXEGETICAL AND CRITICAL NOTES [2. 10 and the Northerners were not Aryas at least, were not regarded as such by Yaska although they must have been brought under the influence of the Aryas to such an extent as even to adopt their language. The same distinction in almost identical words is made by Patanjali in the Mahabhasya.1 Roth’s 2 denial of the correctness of Yaska’s statement that the Aryas use &ava in the sense of motion is groundless, because Yaska is corroborated by a grammarian of such eminence as Patanjali. Moreover, Roth seems to forget that Yaska’s statement is made with regard to the spoken language only as distinguished from the written. At the same time Roth’s suggestion that the whole passage : U^fd’Sfa ….. 1J^ ff?f is spurious and added by a more learned grammarian than Yaska cannot be accepted, because Patanjali quotes the sentence : y although the etymological explanation of the words kambojas and kambala looks suspicious in itself, besides being very clumsily thrust between ^flfd^f<1<4l and fo^u v**SH<| &c., which are in fact the two component parts of a complete sentence. Moreover, the passage in the Mdhabhasya does not contain these etymological explanations. It is thus very likely that they were added later on.
23. 18.] Yaska derives the word danda from the root dad,’ to hold ‘. In order to show that the root dad is not pure invention on his part, Yaska gives an example illustrating the use of the root dad in the sense of * holding ‘. The commentator remarks that dad is used in this sense even in Yedic Sanskrit also, and cites fq^l^i* *Jfi\ W^n in support of his statement. He further adds that Akrura was the king of Vrsnyandhaka and held a jewel called syamantaka. This story (of the syamantaka jewel and King AkrQra) is related in the Bhagavata and Brahma puranas, the Mahabharata, and Hemacandra.3
24. 21.] Yaska gives three derivations for the word kaksa. Durga interprets the word kaksa in three different ways, probably in order to match the etymological explanation. Durga’s interpretations can be deduced from the etymologies of kaksa : it is possible that Yaska himself intended to express the different meanings of the word. If so, it would be an illustration of his principle that the derivations should be different when meanings are different. Kaksa means (1) the region of the girth, hence kaksya means girth, i. e. the belt carried round the region of the girth ; (2) cords used in churning milk. In this sense it is derived from the root gah, ’to churn’, with the suffix A. Durga remarks: Sl I tu^lR^ ff ^%T fq<nWqffi ^ (3) Armpit, as the mosi concealed part of the human body, and not therefore to be made known. In this sense it is derived from the root khya, ’to make known’, in two ways: (a) from the reduplicated form of the root 1 Patanjali, Mahabhatya 1. 1. 1, p. 9. ’ Both, Nintkta : JfrUuferuttp, p. 17. 3 See Bib. Ind. ii. 164, foot-note.
25. 13] EXEGETICAL AND CRITICAL NOTES 225 khya, i. e. kakhyah>kaksah, the reduplication being useless ; (b) from the root khya with Mm, i.e. kirn, khyah>kaksah. The meaning would be: What is there to be made known in the armpit ? Nothing. Being hidden, it should not be exposed. Durga’s comment is: TJ (4) Armpit, as the most rubbed part of the human body, either (a) on account of the motion of the arms, or (6) in order to allay the itching sensation caused by perspiration. In this sense, it is derived from the root kas ’ to rub ‘. Durga remarks : f*t|h|5T Yaska’s attempt to derive kaksa from gah and khya is futile. It is impossible to derive kaksa from gah or khya. The nearest approach to a phonetically accurate etymology is the derivation of kaksa from the root kas. 1 It is, how ever, doubtful whether all the etymologies given above are genuine. As shown elsewhere, many are interpolated.
26. 24.] Durga explains d< W I Kiel by ^pffaw nTWlc{. There is nothing in the text to justify Durga’s explanation of 7W(. Further, it confuses the sense of the text, Durga’s comment being: ^jsjfa^f tnMu^[ fj’lli^ ifa fft * on account of its similarity with the female armpit the male armpit V^C!) |, is also called kaksa (armpit).’ The words nwi+n*n<^ come immediately after ^M^RT. Yaska’s meaning seems to be the following: The word kaksa is derived from the root kas (to rub). From this similarity (of being rubbed) the human armpit is so called (i. e. kaksa).
27. 13.] Roth 2 translates the passage as follows: ‘But even simple (einfache) words one should explain neither to a non-grammarian, nor to a stranger, nor to any one else (sonst einem) who is unfit (untauglich) for this science.’ The literal rendering of Yaska’s words would be :i He should not explain simple words (ekapadani), not to a non-grammarian, nor to a stranger, or to one who does not know (anything about) this (science).’ Yaska goes on to say :’ But one should explain (the ekapaddni) to one whom he knows, or to one who is capable of understanding, or to a wise and pious man.’ By ekapaddni Yaska evidently means primary (nominal) derivatives, the explanation of which in Nirukta 2. 2 he contrasts with that of secondary derivatives (taddhitas) and compounds (samasas), while in this passage he states to what kind of people (secondary derivatives and compounds, as well as) even simple words (i.e. primary derivatives), should not be explained. Durga has the following comment : fnsr The commentator quotes 3RfT in support of his remarks and indicates that on 1 Cf. Zend kasa, Lat. coxa, O.H.G. hahsa. 2 See Erltiutemngen, p. 18, note on ii. 3. 10. P 226 EXEGETICAL AND CRITICAL NOTES [3. 13 examining the word WfT in its isolated state, it cannot be said whether it is derived from the root han, ’to kill’, or M, ’to abandon. He also quotes a stanza from the Rgveda and shows that the character of ma, whether it is a pronoun or a particle, can be ascertained by context only. The argument is that in order to give the etymological explanation of a word, one should know its meaning, which can be fully realized with the help of the context only, hence the warning against explaining isolated words. 4.] Both 1 thinks that the verses quoted in the section are interpolated. It is remarkable that they are accented. The acceiit, however, is not marked with strict grammatical accuracy. Roth 1 thinks that this is an example of unskilful admixture of the grammatical and the musical accentuation. The verses are quoted from the Samhitopanisad.- With the exception of the third verse, they are identical in subject-matter with Manu ii. 114, 115, 144 ; Yasistha ii, 8, 9, 10 ; Visnu xxix. 9, 10, xxx. 47 ; Apastamba i. 1, 14. The fourth section has no connexion whatsoever with the preceding or the following section and interrupts the otherwise logically harmonious order of the text. 5.] From here begins the explanation of the words of the Nighatitu in the order in which they occur in that list. All the synonyms enumerated in the first three chapters of the Nigliantu are disposed of in the remaining part of the second an*1 Ke third chapters. Every synonym is not explained. Only a few are selected for this purpose, others are passed over. Yaska contents himself with indicating the general method of explanation. The commentator charac terizes the work of Yaska on the three sections of synonyms as follows : i I What, then, is the characteristic of this commentary ‘? We reply that this commentary on the three sections of synonyms is that which explains a state, synonym, analysis, number, doubtful derivations and quotations, as well as their different interpretations. Examples are as follows: (1) State the word gauh is a synonym of earth; (2) Synonym the declaration of an obscure by a well-known substantive, as the word gauh means the sun ; (3) Analysis the word gauh means the sun because it moves (gacchati) in the atmosphere ; (4) Number there are twenty-one synonyms of earth ; (5) Doubtful derivation the word nirrtih is derived from the root ram with ni, or from the root r. In the former case it means the goddess of death, in the latter distress ; (6) Doubtful quotation ‘He who made it’, &c. ; (7) Different interpretations the ascetics hold that a man of large progeny comes to grief, &c.
28. 7.] According to Yaska, the sun is called gauh also. In order to illustrate this meaning of the word gau/i, he quotes RV. vi. 56. 3. The word gauh occurs in the locative singular in the stanza, which, according to Yaska, means ’ in the sun ‘. But Durga explains Tf% by Wnffer j^ntiiq$nf«fi. Roth trans-

• See Erlautmmgen, p. 18. 2 Burnett’s edition, pp. 29-32.8. i] EXEGETICAL AND CRITICAL NOTES 227
  lates TTO by* through the course of clouds ’ (durch den . . . Wolkemug). In his explanation Durga remarks : “sniW^fn efc^fH W?i<ni[i: | The joints (parusc) are the days and nights according to Aupamanyava. He explains ^fi as (1) the wheel-shaped brilliant disk that is constantly revolving ; (2) as the cycle of time with its divisions and subdivisions into year, seasons, month, fortnight, day, night, hour, minute, second, and the twinkling of the eye.

6. 11.] Roth thinks that in the sentence : ‘Bfafa fl^tifl refers to the moon. His argument is that if the sun is called gauh, the moon, to which a ray of the sun brings light, can also be so called. And he attributes the same meaning to the word goh in the stanza RV. i. 84. 15 quoted by Yaska. This explanation is misleading and does not suit the context. The word in the sentence does not refer to the moon, but to that particular ray of the sun which illumines the moon. This is a case of extension of meaning. The sun is called gauh, then each of its rays is called gauh. This meaning becomes clear from the sentence which follows the one under discussion, wherein Yaska says : ^ff jfi| ^iq\ 1Tf ^^ri, ‘all the rays are called gavah also.’ This shows that *J in the previous sentence refers to one ray and not to the moon. Durga interprets the passage rightly. He remarks : ^ftif
7. 1.] Durga explains TT^t by vwl in accordance with the meaning given to it by Yaska. Roth translates it by cattle (Einder). As to Durga’s explanation of the dual TT see Roth’s note on ii. 7, Erlauterungen, p. 19.
8. 1.] The stanza RV. i. 164. 32 is quoted to explain the meaning of Nirrtih. According to the interpretation of the ascetics, nlrrtih means distress. The stanza in that case would mean : The man. who causes impregnation (T$ R^rf?f), does not know the reality, for he acts either in a moment of passion or with a desire to get a son. When he comes to know of it, he still does not realize his responsibility, for the foetus is hidden from him. But the child nourished in the mother’s womb is born in course of time. Thus multiplying, the poor man, unable to make adequate provision for the bringing up of his children, comes to grief. Durga’s words are: 3ff3J<3![ ^FU^TT ^i[^<2?t <3[fTs[* This passage foreshadows the Malthusian doctrine. Another different inter- pretation of the followers of the school of ascetics is that the man who causes impregnation becomes himself involved in transmigration, and being born again and again, comes to grief. This is an admonition to lead a celibate life. According to the interpretation of the etymologists, the word nirrtih means earth. The meaning of the stanza then would be : The cloud which discharges the rain-water does not know anything about it, as to where it comes from. It is hidden from the atmospheric cloud which sees it falling, because in the atmosphere it exists in the form of vapours, while the rain-water, swelling in the atmosphere and increasing in various ways, falls to the ground. Roth tkinks it refers to lightning which quickly disappears in the clouds and P2 228 EXEGETICAL AND CRITICAL NOTES [8. i
   leaves an abundant progeny in the form of showers of rain (Regengiisse) whichfall on the earth. See Erliiuterungen, p. 20, note on ii. 8. 1.
9. 11.] The legend of Sakapuni, together with the whole of the ninth section, has no bearing on the subject-matter of the chapter and is altogether out of place. It certainly produces the impression of an interpolation. See Koth, loc. cit.
10. 2.] The second half of the second hemistich is interpreted by Durga as referring to the re-evaporation of the rain-water. As atmospheric deity in the form of lightning, it brings rain down. As celestial deity in the form of thesun, it takes the water up by evaporation. Thus the deity has the doublecharacteristic. This explanation seems to fit in with Yaska’s sentence
11. 4.] The sentence is a little puzzling with regard to its context. Accordingto the normal construction, the word 1?T<^ in the sentence should refer to theword ^uf^(Zn ; but this does not suit the meaning of the passage, for in theRgveda it is neither applied rarely to the sun, nor precluded from being the receptacle of offerings. Durga explains the passage satisfactorily by construingTpTr^ with the last derivation in the preceding sentence ; the meaning thenwould be that the epithet ‘aufc^fta is rarely used in the Rgveda with reference to the sun and has only one hymn addressed to him, while oblations are notoffered to him under this appellation. Durga also refers to disagreement among the commentators with regard to the interpretation of this sentence. See Roth, Erlauterungen, p. 21, note on ii. 13. 19. 5.J Dawn is explained by Yaska as the time subsequent to night andantecedent to sunrise. So figuratively it may be said that the night gives birth to the dawn, and the dawn to the sun. The passage : ^u H^ru ^fffaff M4Hic|is not clear. The sequence of the genesis of the dawn and the sun does not proceed in any order. The words underlineddisturb the regular succession of the birth of the dawn and the sun. The text in this particular place seems to be corrupted. The meaning will be relevant if the word ^TT ….. be substituted for TTf^ I The translation in that case would be: ‘just as being born she gives birth to the sun, i.e. the dawn to thesun, so the night has left place for dawn.’ Durga’s comment is the following :
12. 4.] The sun is called the calf of the dawn from two analogies : (1) Thecalf always goes with the mother. The sun and dawn also appear simultaneously.On account of this companionship the sun is described as the calf. (2) The calf enjoys the privilege of taking milk from the udder of the mother cow ; the sunalso drinks the dew which falls at early dawn and is looked upon as the calf sucking the milk. See Roth, op. cit., p. 22.
13. Roth thinks that Yaska has wrongly introduced the sun in his explanationof the second hemistich of the stanza RV. vi. 9. 1. At first sight it may appearfar-fetched, but- the simile is apt and justifiable. The text of the Veda is the
14. 4] EXEGETICAL AND CRITICAL NOTES 229
    following :* Being born the Vaisvanara fire has overcome darkness with its light like the king/ Yaska’s explanation of the same is the following :’ Being born, i. e. being kindled, the Vaisvanara fire has dispelled darkness with its light like the sun who is the king of all luminaries/ The dispelling of darkness by Agni could aptly be compared with that of the sun alone, hence Yaska’s explanation is far from being unduly far-fetched. 24.] See Roth’s note on fott<sii, op. cit. It may be pointed out that the meaning’ digger of a lotus stalk’ is more appropriate for the force of comparison. Durga also explains the stanza by interpreting sarasvati as atmospheric speech, and by paraphrasing ‘peaks of mountains’ by ’tops of clouds’, ‘waves’ by ’ thunders ‘,* sweeping what is far and near’ by ’ sweeping heaven and earth ‘. 28.] The stanza is translated by Professor Macdonell in J.R.A.S. See vol. xxx, pp. 439, 471-2. Durga has the following comment : ’ The horse hastens forth in spite of being bound in several places. Anything else bound in a single place would not have been able to move, much less to hasten forth.’ See Roth, op. cit., p. 23. CHAPTER III
15. 4.] Durga’s argument for looking upon the daughter as equal to the son is that the daughter’s son is a grandson. A sonless man can have no grandson ; hence if a man had a grandson, he has a son ipso facto. Consequently the daughter would be equal to a son. This argument is not very sound. For it will be applicable in case a man has both a son and a daughter, will give to the daughter a status equal to that of the son, and will entitle her to all the rights and privileges of a son. Historically this is not true of the Aryan family. The daughter never enjoyed equality with a son as far as succession and inheritance were concerned, for on marriage she passed out of the patria potestas of the paterfamilias. Durga’s second argument for the equality of a son and a daughter is that the sacrificial rites performed on the birth of a son are identical with those per- formed on the birth of a daughter ; the sacred texts used in the celebration of the garbhadliana ceremony are the same ; and finally the physical and physio logical processes involved in procreation are without any distinction what soever in begetting a male or a female child. It will be superfluous to add that Durga’s comment gives the argument employed in the text in an amplified 1 14. 9. 4. 8. 8 Cf. ix. 180. 230 EXEGETICAL AND CRITICAL NOTES [4. 4

/ state. Metrical passages from the SatapatJia Brdhmana l and Manu 2 are cited in support of the view stated above, while the Maitrayanl Samhita 1 is quoted to corroborate the ideas of the opposite school. Both pros and cons are thus placed side by side.

4. 15.] The passage shows that women were sold and abandoned. Durga explains sale and abandonment as marriages by purchase and capture. It may also refer to slavery. Abandonment in fact survived among some Indian com munities, noticeably the Rajputs, who exposed female children. This led to infanticide, to which the British Government has put a stop only in modern times. Roth 2 thinks that the passage beginning with “^‘T^T^ and ending with frfij^f Tj^rret is an interpolation for the following reasons : (1) the explanation of the two hemistichs of the stanza (RV. iii. 31. 1) quoted in the fourth section is separated in a forced manner, (2) the length of the intervening passage ; (3) the looseness of its connexion ; (4) the designation of the metrical passage &c. as a stanza of the Rgveda (rk) a term which Yaska never uses and which is evidently a mistake in the present case, as the verses do not belong to the Rgveda, nor possibly could belong to any other of its recensions, on account of their form and contents : (5) finally, the deviation from the purpose of the chapter is remarkable. Roth, however, is unable to say whether the whole or a part of the intervening passage is an interpolation. The only justification which he finds for this irrelevancy is that scholars were particularly busy with the controversy concerning the laws of inheritance. All external evidence i. e. of the MSS. of both recensions and of Durga, who has preserved a critical text of the Nirukta in his commentary is against Roth’s conclusion. As to the internal evidence, the argument of violent separation of the explanation of the two hemistichs and of the looseness of connexion is exaggerated, for the passage up” to f <|f^n^ ^[&% is an amplified exposition of the second pada of the first hemistich. As to Roth’s fourth argument, the very fact that the word fk is not used to denote a stanza of the Rgveda in the Nirukta as Roth himself says shows that it did not then express the meaning attributed to it by Roth. On the contrary, the word nigama is always employed by Yaska to indicate a Vedic passage. The author using a different word for a metrical Brahmana passage shows that he was fully conscious of the difference between the Vedic and non-Vedic passages. Hence it is not a mistake. As to the deviation from the subject proper of the chapter, it may be said that ancient authors had nothing which could correspond to the modern system of foot-notes ; they were obliged therefore to put any discussion arising out of cognate or allied subjects within the text itself. No doubt the intervening passage is a long one and has no other parallel in 4. 6. 4 ; 4. 7. 9. 3 See op. tit., iii. 4, p. 24. 5] EXEGETICAL AND CRITICAL NOTES 231
   the Nirukta, yet to regard it as an interpolation without some positive proof would be going too far. The assumption would be more or less conjectural. 4. 16.] The followers of a third school, representing a compromise between the two other schools which champion the rights of the daughter and the son respectively, hold that it is the brotherless daughter who has a right to inherit the patrimony. Durga states the case for a brotherless daughter as follows : The brotherless girl alone inherits the patrimony and not one who has a brother. The male agnates, who have the right to offer the funeral cake to the deceased, and are alive, preclude a woman from inheritance, because she con tributes to the growth not of her father’s but of a different family, i. e. of her husband, hence she has no right to claim a share in her fathers property. But a brotherless daughter is different. In that case there is no one to perform the funeral rites for the deceased, and because the daughter’s son discharges these duties through the relationship of his mother, his mother, i. e. the brotherless daughter, is entitled to inherit the property of her late father. A stanza is quoted from AV. (i. 17. 1). The MSS. of the shorter recension and the Sib. Ind. edition of the-Nirukta cite the second hemistich only. The MSS. of the longer recension and Durga according to the Bombay edition of the Nirukta give the stanza in full. From the fact that in the text of the Nirukta the second hemi stich is alone explained, one may conclude that originally the quotation was limited to the second half of the AV. stanza only. Durga remarks that the stanza is used in the preparation of medicines for a woman suffering from excessive menstruation. The charm signifies : Let all the blood-vessels come to a standstill, like brotherless daughters who find their path obstructed. See Roth, op. cit., pp. 25-6.

5. ] Roth seems to have misunderstood the significance of the first pada of the stanza RV. i. 124. 7, quoted in the fifth section. He thinks that a brotherless maiden becomes homeless after the death of her father. Being homeless and destitute she approaches men boldly. This is not what is meant. The meaning is that a daughter, although given away in marriage and therefore usually lost to the parental family, comes back to discharge the duties of a son because she is brotherless. Both dawn and the brotherless daughter are conceived as going away from their natal home, and the point of comparison lies in their return ; the one comes back next morning in accordance with the law of nature, the other by the law of society. The simile is rather crude. There are four similes in the stanza. Roth seems to have noticed three only. Na, the particle of comparison, is used four times, and Yaska also says that there are four similes. They are as follows: (1) dawn comes back to men like a daughter who has no brother ; (2) she goes to obtain wealth like one who mounts the platform in the gambling-hall ; (3) like a well-dressed wife desiring her husband ; and (4) like a smiling maiden, she discloses her beauty. *rfr^j is explained in the Nirukta and by Durga as meaning a southern woman who goes to the gambling hall to obtain wealth. rf: is interpreted 232 EXEGETICAL AND CRITICAL NOTES [5 as WFIW, which, according to Durga, means ^^faqmuMi^f, i. e. the seat or place where the die is thrown. The word T& is derived from the root gf (to invoke) and /acquires this sense because it is ^JERJTC’, which is explained by Durga as follows : ^‘ft^ ff THf ^l11^*1 -4 Mfd Durga does not seem to have a very clear idea of the peculiar custom to which he refers. At one place, he says :’ If a sonless woman goes there,, she obtains wealth. The gamblers give her wealth. This is the custom of the southern people.’ At another place, he says: ‘The woman who has lost her son and husband mounts it, i. e. takes her seat on it in the midst of the gamblers. Then she obtains wealth from the relatives of her husband, i. e. her share of the property, settled upon her by her husband.’ Whether the wealth was given by the gamblers or by the relatives of the deceased husband is not made clear ; the precise nature of the custom remains therefore doubtful. In this connexion it may be mentioned that among the orthodox Hindus of Northern India, tliere still prevails a custom called jholibharana, i. e.’ filling the lap ‘, according to which the members of a woman’s paternal family fill her lap with money on her becoming a widow. This attempt to explain a passage of the Rgveda with regard to the customs of Southern India indicates that the expounder himself belonged to the south. It will be going too far to read the peculiar customs of the south in the Rgveda. This confused and altogether far-fetched explanation of the stanza, together with a number of irrelevant derivations which follow the explanation, and the use of the epithet nigama with regard to a passage which is apparently a Brahmana quotation, make the authenticity of the passage doubtful. J|<f|f^c([ should be interpreted to mean,’ one who fights in a chariot ‘; the simile then would be natural. ‘Like a chariot-fighter, the dawn conies to obtain wealth/

6. 10.] Roth paraphrases ^m,<$i von dem dberen unbehauenen Theft, i. e. unhewnupper part. It is more likely that the lower part is meant, which, according to Durga, should be covered with earth and grass. The meaning seems to be that the lowenpart of the sacrificial post should be buried in the ground and not exposed to sight. Durga says that the unhewn part of a sacrificial post is called upara. I think the part meant is lower and not upper, for the upper part remaining above the surface of the earth will still be exposed even if it be smeared over with mud and grass. Further, the part to be buried under ground is likely to be left unhewn.

7. 19.] This is the explanation of the second hemistich of RV. iii. 31. 1, quoted in the fourth section. According to Durga the daughter here meansthe brotnerless daughter. He remarks : This is the second half of the stanza left over; This is explained to support the theory of ’ the brotherless daughter ‘. The first hemistich should be similarly interpreted. The sonless man has the

8. 9] EXEGETICAL AND CRITICAL NOTES 233 daughter’s son as his grandson only when the daughter is brotherless and the rite of ’ appointment’ is performed. But he has no such claim on the sons of those daughters who have brothers. Otherwise all the husbands will be deprived of their sons and marriage itself will be useless and a burden. Or every male child will have a double pedigree. Moreover, every wife being the daughter of somebody may be called upon by her father to remain in her paternal home. This will throw society in confusion. Therefore the son of that daughter only who is duly appointed in accordance with law belongs to her father, but not the son of a ’laughter who has a brother. According to Durga the texts which advocate equality between a son and a daughter refer to a brotherless daughter. ^7f:%^l is explained as ^TTCF: 4)ftli ^fl ^fi^R, ^f* fa^ffl rf WnRTrrnC*l> i- e. husband. Durga explains Um^fa ’ he arranges, or selects, or makes him approach ‘, rln, i. e. free from the anguish of sonlessness. 6.] The meaning of the word *rfj[ is obscure. Yuska followed by Durgaexplains it by *ffeT in the fourth section and by *f^ in the sixth. Moreover, he seems to take ^fj[i as a copulative compound standing for both man and woman. 8.] Durga makes the following remarks on the stanza RV. x. 53. 4 : Thersiship is that of Sautika Agni, and this is his dialogue with the Viwdevas.1 Roth’s criticism of it is, that although a mythological element is found in the two preceding hymns, it is entirely lacking in this hymn. Durga takes ^TT as a vocative, which is not only against Yaska’s explana tion, but is impossible on account of the accent. The translation of the stanza according to Durga would be as follows :’ I know that most excellent strength of speech with which we shall overcome the evil spirits. O gods, partakers of food and accomplishers of sacrifice, and ye five-tribes, favour my sacrifice.’

9. 9.] Two explanations are given of ’the five-tribes’, in the Ninikta. According to one view,* the five-tribes’ are the gandharvas, manes, gods, evil spirits, and demons ; according to Aupamanyava, they are the four castes and the Nisadas. The former is highly improbable, because the five-tribes are called upon to participate in the sacrifice and an invitation to demons and to evil spirits to share the sacrifice i? inconceivable on the part of any sacrificer. The evidence of the stanza itself is against this view. The first hemistich mentions the overthrow of the evil spirits in clear and unmistakable terms. Further, on account of the eternal hostility between the gods, the guardians and promoters of the sacrifice, on one hand, and the evil spirits and demons, the disturbers and destroyers of the sacrifice, on the other, it does not seem possible that both gods and demons could be thus asked to enjoy the sacrifice. A gathering of this kind coulc} never be harmonious. The latter explanation 1 See Roth, op. tit., pp. 27-8. 234 EXEGETICAL AND CRITICAL NOTES [8.9 is undoubtedly the better one. Yaska quotes KV. viii. 63. 7 to show that the five-tribes 1 of the stanza under discussion are meant to be human, and thus supports this view by implication. Durga, of course, agrees with Yaska. This means that the four castes and the Nisadas shared the sacrifice in common. This would imply some sort of inter-dining among the various castes, and would show that the water-tight compartments into which the various castes are divided, and separated by rigid barriers of mutual exclusiveness, did not exist in Yaska’s time. Further, the epithet’ holy’ (yajhiyasah) is applied to ‘five-tribes’. If we accept Aupamanyava’s view, all the four tribes and the Nisadas would be holy. This would mean, even if the Nisadas, who according to the etymologists are sinful beings, are excluded, that the lower castes were not despised by the upper ones, especially the Brahmanas, which would again imply greater equality for the former and less tyranny on the part of the latter. This shows that, unlike the moderns, the ancient Brahmanas could not have regarded themselves as polluted by the mere sight or the touch of the lower castes, otherwise the latter would neither have been invited to partake of the sacrifice, nor called holy ; i. e. the problem of the pariah and the ’ untouchables ’ had not then arisen, or rather the pariah and the ’ untouchables ’ themselves did not exist. 1O. 9.] The etymological explanation -SlfeJU^Idtft qiWJ-flqdlfa Tf I is not found in Durga’s commentary. As shown in the introduction, the passage therefore is an interpolation. The spurious character of the passage is indicated by the internal evidence also : Yaska has already explained the word ambu-da (cloud) as the giver (from the Vda to give) of water (ambu). The passage in question is, therefore, superfluous. Moreover, it does not explain the etymology of the word ambu-da. Besides, here the derivation of ambu-da is of secondary importance ; it is the word arbu-da that Yaska wants to explain. For this purpose he finds it necessary to explain arnam first, and explains ambu-da incidentally. It is reasonable to expect that in such a case he would not attach too much importance to ambu-da. This evidence, together with Durga’s omission of the passage, leads me to the conclusion stated above. 12.] According to Durga, the meaning of the stanza i. 163. 21 is as follows : ’ The rays of the sun which fall on a bright object whose darkness is removed, or which are bright when they fall having obtained a share of the immortal water from the earth, and being accompanied by it, join all creatures in invoca tion ; or, having seized the fluids, respectfully go back to the sun. They have the knowledge of their duties. The lord, wise guardian of the entire universe, i.e. the sun, approached me, whose intellect is immature.’ With regard to the soul, the meaning is as follows : the senses, well placed in their several organs, having obtained a share of the immortal knowledge, stimulate the consciousness of the inner person with regard to external 1 See Macdonell and Keith, Vedic Index, under panca janah.

10. 9] EXEGETICAL AND CRITICAL NOTES 235 objects, or approach the faculty of discrimination, bearing the knowledge of external objects in order to bring about their cognition. The senses themselves have no power of apprehension. They are but the unconsciouso instruments of the conscious purusa, who is the doer, and whose nature is knowledge. The lord, guardian of all senses, i. e. the soul, the wise one, has entered me, the immature one.

11. 4.]’ Devarah is so called because he is the second husband.’ This refers to the custom called niyoga. It prevailed among the Jews 1 and was also an Indo-European2 practice. It is also found among the Purans, a Borneo jungle people of very primitive type, among the Bayaka in Africa, among the Aleuts, and Thlinkeats, and Koloshes.’

12. 13.J Roth translates TJTT fapff WT as noch elie sie gleichsam ans Leben greift, i.e. ‘before it, so to say, attacks life’. Durga explains the same words as : ^TOT <fliH3*si *JTT 1JT t!iii<^t3n3q ^f\qV fi|fa[ II The passage is quoted in order to illustrate the use of yatlia as a particle of comparison.

13. 9.] Roth says that, according to Durga, the inaccurate diction of Yaska has given rise to the divergent reading ^MlH^lfl^U^ffa^ I (op. tit.,ui. 20. 7, pp. 34-5), Roth probably has misunderstood Durga. At any rate, Durga does not say anything about the ’ inaccurate manner of expression of Yaska ‘. It is also not clear to me in what Ysska’s inaccuracy consists. Moreover, Roth’s quotation from Durga’s commentary is defective ; Durga’s sentence runs thus : ^MlH*|fa$Vfa^ cTOTfa <Ti<nC I …. *Ri: TOct qrefTT I TFJtffiT ^7f *ref?T I The words underlined are mutilated and iqfa is omitted in Roth’s quotation. Durga only wants to say that some MSS. of the Nirukta have the variant ^Mltf<3. And in his paraphrase of the two Vedic quotations, RV. v. 75. 7 and yii. 104. 21, he explains both f?n[: and ^Jrf: not by the meaning attributed to them by the reading J4lH<3 but by that of the variant. His remark is as follows : 1 See Deuteronomy xxv. 5; St. Matthew Grote, ii. 6, p. 520, and note at the end of xxii. 24. vol. ii. 3 See Xenophon. Rep. Lac. i, 9, cited by 3 See Hobhouse, Evolution ofMorals, pp. 34-5. 236 EXEGETICAL AND CRITICAL NOTES [2. IT CHAPTER IV Durga, after the usual benedictory stanza, quotes two verses without indicating their source, to the effect that a seer, after greatly expanding knowledge, should expound it briefly, for in the world an adherence to prolixity and conciseness is desired of learned men (according to the exigencies of the occasion). He takes this dictum as a test of sound scholarship and applies it to the commentary of Yaska. He shows that it is followed by Yaska, who is concise and also prolix as the occasion demands. According to him, brevity is the characteristic of the jsecond and the third chapters of the Nirukta, commenting on the synonyms which should be and are explained briefly, i.e. by giving the necessary informa tion about a particular word, its synonyms, their number, quotations to obviate ambiguity of meaning, and explanation of the same. All this exposition is brief. He then notices omissions on the part of Yaska in the explanation of synonyms in the Nlrukta. ’ In this part of his commentary of the Nirukta ‘, he remarks, ’ one single meaning indicates many words and many words denote one single meaning.’ He adds that the different shades of the meanings of these words are not explained. In order to make his meaning clear he gives the following example : 102 words beginning with ^f^ and ending with ^‘STO^t are explained to have the general meaning of motion. But a particular word expresses a particular kind of motion only to the exclusion of others ; e. g. the verb Icasati denotes hopping only and not flying ; ^Tld^ is used in the sense of rolling only in connexion with a ball, for instance ; ^fld^ expresses’ trickling’ and is used in connexion with liquids only. Thus the specific forms of the general meaning should be investigated. Further, the usage of different provinces is nut explained. In some places the common words are used to denote technical expressions and vice versa. These arguments of Durga may be supplemented by drawing attention to the fact that only a few of the synonyms are explained the restjare passed over. With this section of Yaska’s commentary Durga contrasts his commentary on homonyms. He points out that the whole list of the homonyms is explained word by word. Different meanings of one word are fully explained and illus- trated with Vedic quotations. The characteristic of Yaska’s commentary on homonyms is therefore prolixity.

14. 11.] Many meanings are attributed to the word ftim+i. On account of its different interpretations by Yaska and his contemporaries, and the several theories put forward by Durga, the word seems to have lost its precise meaning at an early period. It is evident that the word means a particular part of the sacrificial animal. Durga’s one theory is that the body of an animal can be divided into : (1) external parts, as shoulders, hips, &c. ; (2) internal parts, as tongue, heart, &c. He thinks that the two words preceding IVfll4 m tne18] EXEGETICAL AND CRITICAL NOTES 237
    passage quoted by Yaska refer to the external parts of the body, i. e. flanks andhips, hence ftjfnT would naturally also refer to some external part. This is the argument adduced by him while explaining Yaska’s sentence that the wordftjTTTR means * forearm ’ (dos). The word dos is derived by Yaska from the root dru (to run),’ because ‘, says Durga, ^it is with the strength of the forearm that animals run. Durga’s second theory is that the two words preceding in the passage referred to above describe successive parts of the body ; therefore would mean a part which comes next to hips, i.e. arms. It is for this reason that he paraphrases fff%: by iffl and derives fafaff: from the root fa\3 cl. to pervade ‘, and explains ftftcft H^f?f as **TR: H vjO^t^l *lf?l t W Vf^fd, i.e. it is full of ordure, … or its flesh is loose. I myself derive (qfarH from Jt with fa and translate it as ‘it is open ‘. The various meanings attributed to R|<ni are the following : (1)’ forearm ‘, byYaska ; (2) yonih (uterus) explained as equivalent to anus, according to Durga, bySakapuni ; (3)’ liver’ on account of its dark colour (tyamari), by Taitlki ; (4) fat, lit. white meat (titi-mamsa), by Galava. 1 Durga remarks, Thus it is clear that the grammatical form and the meaning of the word ftTfTW are not (precisely) known ‘. 11.] Durga explains RV. i. 115. 4 as follows : ’ The divinity and the greatness of the sun is that, ignoring all workers in the midst of their work, he rolls upthe net of light which is difficult to be gathered by others, but which the sunrolls up in an instant, without much trouble and without any other help. Hehas yoked the bay steeds i. e. rays which draw up fluids from the stable, i. e. the earth. The earth is the sadhastham, for it is hence that rays extract fluids.’ The meaning is that when the rays are withdrawn from the earth, night spreads her garment over all. 15.] Roth translates the word Rfi*fcRT as ‘image’, and Langlois as ‘marionette’. According to Yaska, however, it means a ‘maiden’. Durgaexplains it by ^n^WdrRfT, i.e. an image made of the sala tree. 17.] The quotation ^f *pTT V^: &c., has not yet been traced. In both theBib. Ind. and the Bombay editions of Durga’s commentary the stanza is givenin full. But Durga explains the fourth verse only. His remark, that the rest is to be discovered (p3fU^ Ifa: )> indicates that he did not know the other verses of the stanza. To include the first three verses in the above-mentioned editions of Durga’s commentary is therefore a mistake. 18.] Yaska quotes one Vedic passage only to illustrate his meanings of the two words ^<JMK^1 and ^TT^%. The order in which these two words occur in the Nighantu is reversed in Yaska’s quotation. According to Yaska, the word^MK^ means ‘sun, ocean, and the tortoise’. Durga paraphrases the wordby ^fcMd^UU^T, i. e.’ complete without any flaw ‘, and further explains the1 See Roth, op. cit., iv. 3. 6, pp. 38-0. 238 EXEGETICAL AND CRITICAL NOTES [18
    term as that which may be sufficient in this, and glorious in the next world. Durga’s derivation of lH I <,<g is the following : I He says : ^R <pf\R< Durga notices the difference of the order of the two words in the Nighantu and the Nirukta, and concludes that the Nighantu and the Nirukta are the works of different authors. See Roth, op. cit., iv. 18, p. 45. The word J<plK is explained as’ moving swiftly ‘, i. e. derived from the root </tuj, or as ‘having good offspring’. On the latter Durga remarks, ‘The word tuk is a synonym of offspring. The offspring of Agni is intended to be golden. And Agni himself is called of golden seed, and so on. Durga paraphrases JH<flfK by 3^T%, i. e.’ of noble breed ‘, or the horses who have noble foals, for it is the noble horses only who can breed noble foals, i. e. the fact of breeding noble foals implies praise of the horses, and the possession of such horses implies praise of Agni. 19.] Durga’s explanation of fkjia^qiJ does not indicate any reference to phallism. There is no evidence to show that Durga or Yaska was even aware of its existence. According to them the phrase denotes profligate persons whose sole or chief end in life is to gratify their sensual desires. But the phrase is a possessive (Bahuvnhi) compound, and can be translated accurately only as ’they whose god is the phallus’. It may be that the cult of the phallus, inasmuch as it originated from the aborigines, was not known to Yaska or Durga. In ancient times it was a widespread cult, and in one form or another survived in many Roman Catholic countries, like Belgium, France, and Italy, down to the middle of the eighteenth century. Phallic remains are discovered all over the world, notably in Ireland. According to some, the maypole and the cross are also phallic survivals a very doubtful conclusion. The worship of Siva as the lingam may remotely be connected with it. See Hastings, Encyclopaedia of Religion and Ethics,’ Phallism ‘. Cf. also phallic objects, monu ments, and remains. 21.] Durga explains the stanza RV. i. 164. 33 as follows :’ Heaven is my father, the progenitor, and the great earth is my mother, because the rain descends from heaven to earth, helps the various herbs and plants to grow, which nourish the body and endow it with the seed of future generations : heaven and earth are therefore the primaeval cause of life. The intermediate space is the womb, i. e. the source of rain-water. The father has bestowed the life-germ on the daughter, i.e. the cloud has bestowed the rain-water on the earth. The earth is here called the daughter (du-hita) because it is placed at a great distance from the cloud (dure-hita).’

15. 10.] Yaska takes Ipsfr as a compound and explains 7T?ff and 4|l4i* This analysis is in agreement with that of the author of the Padapatha. Durgn remarks that Yaska has divided one word having the ablative or genitive ending 25.25] EXEGETICAL AND CRITICAL NOTES 239
    into two words of the accusative case, and in order to connect the sense of the passage has supplied the words ’ diseases ’ and ’ dangers’ in the Nirukta. Roth agrees with Yaska’s explanation. Of. op. cit., p. 48. 23.] Durga explains imi (mother) by ^rf^Tff^ST^, i. e. fashioner of all beings ; ^f: as 4^uH ^3HT: trraiTH*!?) I, i. e. one who saves from a great sin. According to the interpretation of the etymologists the meaning of the stanza would be as follows: ‘Heaven is unimpaired, the intermediate space is unim paired. He is the mother, the father, and the son. The Vttvedevas are unimpaired, the five-tribes are unimpaired, and all that is born, or is yet to be born.’ 24.] Roth’s translation of the second hemistich is not clear. The word 3njfT*^is explained as i<at,by Yaska, which is paraphrased by Durga as <^^ ‘bound’. He says: ….. Tf ^l^Vr-g^T f T^ |lf^ if* Ttfcft TnjT * ff I . . . T &cT5: \K< | There is a contradiction in Durga’s explanation : a falcon that is bound cannot pursue its prey ; ^RP^ derived from the </***(. means ’liberate’, ‘free’, ’let loose’. Of. Roth, op. cit., iv. 24, p. 48.

16. 12.] Durga explains ^cjifl^H. as one who is addicted to self-enjoyment and who does not sacrifice to the gods.

17. 19,] Yaska does not cite any passage to illustrate BIU in the sense of ’ separate ‘, &c. Durga supplies the quotation RV. vi. 40. 5 for this purpose. He says that some commentators interpret the passage (VS. 8. 20) quoted by Yaska in two ways, i. e. taking alternately the two meanings of ^>n<^ ’ separate ‘, and ’ prosperity ‘. Of. Roth, op. cit., 14, p. 50.

18. 25’.] Roth thinks that the text is corrupt and the word ^oii^ is super fluous. I do not agree with Roth, for with a proper punctuation the sentence is quite clear. It is to be read thus : ^Tl^fd ^mum^ I ^WTO I ^f5TT ^RTT: I ’ The author calls Pfisan goat-teamed. Goat-teamed, i. e. goats are the coursers.’ Cf. Roth, op. cit., 18, p. 51. 240 EXEGETICAL AND CRITICAL NOTES [3
    CHAPTER V Durga explains SflRi^ as ‘wrapped up in waters, flowing on all sides, or well-washed, i. e. a cloud ‘. He explains!WT% asl in the atmosphere ‘. and i^liit> as “‘of waters’. According to Durga, the sense is as follows: Indra found the cloud in the atmosphere and noticed its strength, so he bored holes into it for the waters to flow down. The waters then proclaimed Indra to be Gandharva. Durga adds,’ it is heard that Indra is G-andharva and the Maruts are his nymphs ‘. 3.] Durga explains Mf!^ by ^^rff3fH!W, i. e. manifest, and paraphrases ^HsR^ by 4^VIH. after Yaska. According to him, lightning remains invisible during eight months, appearing in the rainy season only, hence the epithet ^facn^sji is used with regard to it. Cf. Roth, op. cit, 15, p. 55. 4.] According to Durga, the groups of atmospheric gods here refer to the Maruts, for the word group has been used in the plural number and it is the Maruts only who have several groups, each consisting of seven members. Hethen quotes the Brahmana passage : % <H<IH i'dl W I He further says that Rudra is called varaka also, and quotes RV. i. 114. 5, in support of his statement. Cf. Roth, op. cit., 5, p. 56. 7.] Yaska’s explanation of RV. i. 150. 1 is confused. He explains “Vf^l by ^fa’q: But a worshipper cannot be unfriendly to the god whom he worships. Roth connects ^ffT with ^^u and deduces the meaning,’ strange, stranger, guest, and a hostile person ‘. The other meaning, lord ‘, attributed to ^rft: by Yaska is equally unsatisfactory as far as the present passage is concerned. Durga makes an attempt to explain ^rfT as competent to offer oblations and sing praises ‘. But by so doing he ignores the interrogative particle feif^ . Yaska’s explanation of the third pada is obscure. Durga derives S|<% from the root J to injure, and paraphrases it by f^f% . According to him the sense is that just as many streams flow into some wide cave, and notwithstanding the large number of streams the holding capacity of the cave is not affected thereby, so many oblations are poured into fire, without affecting the latter’s capacity to consume them. Having seen this phenomenon, the seer proclaimed, <f^ir? KUJ W 31, i. e. (oblations flow into fire as waters) into the cavity of a great well. Cf. Roth, op. cit, p. 59. 7.] Muir 1 attributes the sentence, ’the words Sipivi?tq and Vimu are two names of Visnu ‘, to Aupamanyava. This, however, is not correct. The first 1 Cf. Sanskrit texts, vol. iv, p. 505. 26] EXEGETICAL AND CRITICAL NOTES 241
    sentence contains Yaska’s own remark, while the second gives the view of Aupamanyava. Durga puts the second sentence only in the mouth of Aupa manyava, and his interpretation is supported by internal evidence also. Durga remarks that of the two names of Visnu, the first aione is illustrated, because the second belongs to the daivata kdncla. Cf. Roth, op. cit. , v. 8, p. 59. 12.] Durga remarks that some think the stanza x. 89. 5 to be chiefly addressed to Indra, and explains the third pada as follows : i. e. Indra pervades everything as Soma does all plants and trees. Following Yaska, Durga suggests two alternative interpretations : he ascribes (1) the first hemistich and the fourth pada to Indra, and the third pada to Soma, (2) the first three padas to Soma and the last to Indra. The meaning would be, (1) Indra, who is infused with energy, who rushes to the attack, the shaker of enemies, the impetuous, the great hero armed with the thunderbolt him all counter-measures do not deceive ; they perish even before they reach him ; may he and Soma which surpasses all plants and trees favour us ; (2) may Soma which infuses energy, which flows quickly, the shaker of vessels, active, exhilarating, foaming, and surpassing all plants and trees, and Indra whom all counter-measures do not deceive ; they perish even before him ; favour us. In both cases, Durga thinks this to be a joint panegyric to Indra and Soma. Cf. Rcth, op. cit., v. 12, p. 62. 24.] Yaska explains cpd<3 by &fl4$|. This shows that he noticed this example of Prakrtization in the Rgveda. Whether he deduced any general principle of Prakrtization, or whether he was even aware of the existence of this phenomenon, is doubtful. Durga mechanically repeats Yaska’s words. Apparently the paraphrase of JZ^ by eid^ did not strike him as unusual. 25.J Durga explains the second hemistich of the stanza x. 44. 6 as follows : ’ Men, who were unable to obtain thy favour and who could not ascend the ship of sacrifice, being chiefly addicted to sensual pleasures, committed vile deeds and consequently obtained bodies appropriate to their deeds in accordance with the law of Karma: He then quotes a passage without indicating its source : ^TT I’ Now there is a prospect that they, whose deeds are vile, will obtain a vile form of existence, i. e. that of a dog, or of a pig, or of a low-born man ’ (Chand. Up. v. 10. 7). 25.] Durga remarks that some commentators explain the stanza x. 50. 6 with reference to Agni, i. e. Agni is called the son of strength (suno sahasah) as he is produced by strongly twirling one stick on another. This explanation, according to him, is inaccurate, for it is contrary to the context, as the hymn (x. 50) is addressed to Indra. He thinks that the epithet,’ son of strength ‘, here refers to Indra, who is so called because he is the son of prana. Cf. Roth, op. tit. 7, p. 70. 26.] Durga explains the stanza x. 101. 7 as follows :’ Refresh these horses Q 242 EXEGETICAL AND CRITICAL NOTES [26 with water and fodder as the battle is near at hand. Win a good victory with the refreshed horses. A victory, when one’s dear friend, brothers, sons, &c., are killed, is not good. Having a noble car, pour down men into this well of battle as if they were water/ &c. He explains ^qaJ as3JTO: T f^f CHAPTER VI Yaska divides the word aSu-Su-ksanih into three parts and says, au- and $u are synonyms of quick ‘. The latter part, ksanih, is derived from the root ksan (to injure). According to him the word therefore means, ‘injuring very quickly.’ Durga has not rightly understood Yaska and consequently mis construes the sentence. He thinks that the word u occurring in Yaska’s statement does not refer to the second part of the word au-u-ksanih, but to an altogether different word which means ’ space ‘, and which has been used by Yaska incidentally only : ^J- ^SfK^ HmfWi . . . < He says that of the three parts au~$u-kanih, Yaska explains the first (au) and the third (ksanih) but passes over the second (&), adding that 6u is derived from the root Sue,l to shine ‘. Durga’s explanation of Yaska’s division and derivation of au-u-Jc8anih is quite wrong, and it is surprising to note that he could thus misunderstand Yaska’s very clear statement. After deriving each individual part of a$u-$u foanih, Yaska explains the meaning as -411^ ^J^T ^rtPtfa- It appears that Durga has connected these three words with each part of a&u-Su-ksanih respec tively and hence his mistake. It may be again pointed out that Yaska very clearly says that the words a&u and u are synonyms of ’ quick ‘, and he nowhere derives u from uc. Yaska offers another etymological explanation of a-&u&u– ksanih. The first part a is the preposition and the second part is formed from the desiderative form of the root inc. Grassmann agrees with Yaska’s second derivation. Cf. op. cit., p. 188. Roth probably was not aware of the contra- diction in Durga’s explanations of Yaska. 2.] Durga explains RTT by waters stored up in the interior of the cloud. When the cloud is pierced, waters flow down without any other obstruction. Running by downward channels they protect reservoirs like lakes, tanks, &c. He paraphrases <j^fli by ^^R. He gives a second interpretation of the same as follows : The rain-water coming down from a pierced cloud is invoked by the people. They shout with joy,* Oh ! how fine ‘. Words like these coming out of people’s mouths protect the rain-water. He takes quifl* in the sense of * waters’ in the first case, and in the sense of ’ shouts of people’ in the 8] EXEGETICAL AND CRITICAL NOTES 243 second, and derives TUft: from the root ^f (to flow) in the former and from ^^ in the latter sense. 3.] Durga explains^311 as’ from every side ‘. The meaning, according to him, is,’ Uproot them from every side so that the enemy, even if they try their best, may not be able to know from which side they are being uprooted. Or uproot them without a trace.’ Durga’s last sentence is strikingly modern. His words are : 4.] In his paraphrase of the stanza v. 54. 6 Durga explains ^\ra: in two ways: (1) as an epithet of worms, Le. the worms which penetrate a tree and consume its sap ; (2) aa an epithet of the Maruts, i. e. O wise Maruts, you steal the waters of a cloud as worms the sap of a tree. Durga’s second interpretation is the correct one ; ^V: is in the vocative case as the accent indicates, and can only be connected with the Maruts. 6.] According to Durga, a mother is called nana (from Vnam) because she stoops in her various acts of kindness, such as giving suck to the baby. Adaughter is called nana also, because she stoops down while she attends on her father. He remarks that if the word tatas is taken in the sense of * a father ‘, then nana would mean ’ a mother’; but if the former signifies’ a son ‘, then the latter would mean ’ a daughter ‘. He paraphrases bhimk by brahma and remarks that brahma is called bhisak because it is he who prepares remedy (bhesaja) for sacrifice when the disease of expiation becomes manifest. He then cites a passage which looks like a Brahmana quotation :’ Verily is this sacrifice cured where there is a Bralma who knows so much.’ He adds that” bhisak means a physician also. 6.] The passage containing the explanation and illustration of the word ksonasya, and placed within square brackets, is omitted by the MSS. of the shorter recension. The omission however is not justified. The genuineness of the passage is beyond doubt, for the word ksonasya occurs in the list of homonyms enumerated in the fourth section of the Nighantu, and must there- fore be explained and illustrated, as every word in the above-mentioned list is so treated according to Yaska’s plan. If the passage in question is omitted, ksonasya would be left unexplained. Thus a gap would be created which would make the otherwise complete commentary of Yaska on homonyms incomplete. The unjustifiable character of the omission is further proved by the evidence of Durga, who does not question the authenticity of the passage. 8.] Yuska’s explanation of the stanza viii. 99. 3 is very unsatisfactory. He suggests two alternative interpretations. He ignores the particle iva in his first, and id and na in both his interpretations. Durga has not rightly followed his second interpretation, wherein the meaning of the particle iva is explained. Durga mistakes the explanation of iva for that of id, passed over by Yaska in both cases. Durga justifies Yaska by saying that iva, id, and na are expletives. This is I think the only instance where na has been taken as an expletive. 12 244 EXEGETICAL AND CRITICAL NOTES [8 Both Yaska and Durga paraphrase Wiaksata by a participle. For the form bhafaata, see Professor MacdonelPs Vedic Grammar for Students, p. 401. 14.] Yaska explains f^IT^W by X^^KlRU!: Durga reads and gives lpR[TfTW as a variant. None of the MSS. which have been collated so far, except one utilized by the editor of the Nirukta in the Sib. Ind. and specified by him as >ff, have the reading ^^q^iRffl, although Durga’s reading acquires a certain amount of plausibility as to its correctness on account of the similarity of sound, for f^uji^tlJ and vii<^ifon are more similar to each other than filil^W and ^Sl^lfi^n Durga’s explanation of the word ft ff ^i|lfii fittnql^qfa <T^ cT W^pTRJ^f’TT | is far-fetched. He divides ri^ad-asah contrary to Yaska, who seems to take it as ri$a-dasah, i. e. destroyers of the enemy. A more natural way of explaining the word would be ri$a- adah, i. e. devourers of the enemy. Grassmann explains it as destroyers of violent acts ‘. Roth’s quotation of Durga’s comment is inaccurate. He makes the sibilant in dasinah palatal, while it is dental. Cf. Roth, op. tit., vi. 14, p. 80. 15. Yaska construes ^Jfl as the nom, pi. and explains it by or qirnuaii as an epithet of the atmospheric gods. Evidently he derives n^j^ from ^fg and 1^ a poor etymology. In the text of the Rgvedic stanza (x. 82. 4) ^fpf is immediately followed by ^. The former is the opposite of the latter, and both are in the same case, yet according to Yaska the former is the nom. pi. while the latter is the loc. sing. ; the former is an epithet of the gods, the latter of the region. This interpretation is very unsatisfactory and destroys the contrast intended to be brought about by the use of two opposite terms. Durga mechanically follows Yaska. The natural way would be to take ^f$H and JJfl both as loc. sing, and to connect them with T^ffa . This will bring out the contrast. p is derived from ^ (to move), i. e. trodden, and “^^as its opposite would mean ’ untrodden ‘. I translate the Vedic passage as follows: ‘seated in the region which is trodden and which is not trodden,’ &c. Grassman translates ^f^ as’ not illumined ‘,’ dusky’ ; see Warterbuch, p. 157. 15.] Yaska paraphrases ^UAHfa by ii^ifa, i.e. he derives it from the root jan (to be born). But grammatically it is ao. pass, of the root If (to grow old), and the sense will suit the Vedic quotation better than that expressed by jaw, i. e. Agni has been made old by sacrifices as a father of cows by his many children. It occurs only once in the RV. Durga remarks: 16.] Yaska remarks that pacata is used as a noun in x. 116. 8, that it occurs in the singular and dual numbers, and quotes VS. 21. 60 and RV. iii. 28. 2 to support his statement. Durga goes still farther and says that pacata is used in the plural also: li<dSh^^i f^q^‘i ^pq-q! qT Hqt?f As if to illustrate his remark, he construes pacata in the above-mentioned stanza 33] EXEGETICAL AND CRITICAL NOTES 245 with liavlmsi : M^m MauT*ufl ffYfo. In the hemistich verb precedes object, as ^rf% before f^fffa and <faui before Tj^n ^7f jftwi; The translation should therefore be :* accept graciously the cooked viands and the soma ‘. 19.] Durga gives three explanations of the word udhas : (1) The hand-press full of soma-juice, taking it as a noun ; (2) below, taking it as an adverb. The meaning then would be : release the soma-juice in various vessels and cups below the skin used for straining purposes. . . . ; (3) below or above ; the meaning is : release the straining skin ‘vhich is below or above the sacrificial car: 22.] Durga takes*j<^ts(fa to be Yaska’s explanation of tflfejtjft He is evidently wrong, for Yaska clearly says that both ^t^T^TRF and ^Jl7l are intensive forms, paraphrasing the former by <^[c^ and the latter by 28.] The author of the Egvedapadapatha analyses vayah into va and yah. Yaska rightly objects to this analysis, for the relative pronoun would introduce a dependent clause whose finite verb, according to the rules of accentuation, should have the acute accent. And because the finite verb has the grave accent, it shows that the clause is not dependent and therefore Sakalya’s analysis is not correct. Yaska’s other objection to the analysis is that the meaning will be incomplete. Yaska takes vayah as one word, meaning the young of a bird. For the accent of the verb in a dependent clause, see Pa. viii. 1. 66 ; Professor Macdonell’s A Vedic Grammar for Students, p. 467. 3O. ] Durga reads utfir: instead of cf^^r: and explains : %f nfa $eifi I Famine is personified. On account of starvation, the sight of famine-stricken people becomes dim, therefore famine is called one-eyed. On account of insufficient nourishment, people totter on their legs, therefore famine is spoken of as having a crooked gait. Famishing people scream, and so famine is called screaming. It is called barren because there are no crops, or because people are no longer liberal in their gifts. 33.] The entire section, including the stanza together with the commentary, seems to be spurious. Yaska never cites more than one Vedic quotation to illustrate the same meaning of a word. The commentary on this stanza bears the stamp of a -different commentator. As the style has affinities with comments of the parisistas, it is likely that it has been added by the author of these. The explanation of rdupe and rduvrdha is not satisfactory. Durga has also noticed this doubtful explanation and remarks: Cf. Roth, op. cit., p. 98.

    APPENDIX AN ALPHABETICAL LIST OF STORIES RELATEDIN THE N1RUKTA Akrara and the Jewel .-../ 2. 1 Agastya and Indra . . ‘. . . . . i/l. 6-6 Angiras’ birth 3. 17 Atri’s birth . 3. 17 Asvins (birth of Asvins) . . .12.10 Asvins, Usas, and Aditya 5. 21 Aditya, Usas, and Asvins . 5. 21 Aditya, SaranyQ, and Asvins .12.10 Indra and Agastya ../! 6-6 Indra and the Seers (means of livelihood in a famine) . . . 6. 5 Usas, Aditya, and Asvins … . 5. 21 Urvasi and Mitravaruna (birth of Vasistha) . . . 5. 13 Trita cast into a well …. . . . . 4. 6 Devfipi and 6antanu . . . y2. 10-12 Panis and Sarama . 11. 25 Brahmana anil Vidya . <.‘2. 4 Bharadvaja’s birth . . 3. 17 Mitravaruna and Urvasi (birth of Vasistha) … 5. 13 Bhrgu’s birth 3. 17 Mudgala and his victory . . . . . . . 9.23-24 Yama and Yami . . . . . . . . . . 11. 34 Lopamudra’s love . 5. 2 Vasistha and his fetters …. . 9. 26 Vasistha and the frogs . . 9. 6 Vidya and Brahmana …… 2. 4 Visvakarman and the Universal Sacrifice . . . . . 10. 26 Vaikhanasa’s birth …’… . 3. 17 Visvamitra and the rivers 2. 24-27 6antanu and Devapi …. 2. 10-12 Sakapuni and a deity …. 2. 8 SaranyQ, Aditya, and Asvins …. . 12. 10 Sunahsepa . .. 3. 4 Sarama and Panis 11. 26 Surya, Savitr, and Soma ….. 12. 8 Savitr, Surya, and Soma (marriage of Stlrya) . 12. 8 Soma, Savitr, and Sttrya 12. 8INDEX OF AUTHORITIES CITED IN THE NIRUKTA Agrayana . 10.8 Agrayana . 1. 9 ; 6. 18 Acaryah Eke “. 8, 4, 5 7.22 Aitihasikah Audumbarayana

19. 8 ; 7. 13 ; 8. 212. 16 ; 12. 1, 10 1.1 Aupamanyava . . 1. 1 ; 2. 2, 6. 11 ; 8. 8, 11, 18, 19 ; 5. 7 ; 6. 30 ; 10. 8 Kathakam ….. 10. 5
    Aurnavabha 2. 26 ; 6. 18 ; 7. 15 ; 12. 1, 19 Katthakya 8. 5, 6, 10, 17 ; 9. 41, 42Kautsa. 1.15Kraustuki 8. 2 Gargya 1. 3, 12 ; 8. 18Galava 4. 8 Taitiki … . . ….4. 3 ; 5. 27 Carma&ras . . . “. 3. 15 Duhitr-dayadyah 3. 3 Naidanah 6. 9 ; 7. 12Nairuktah 1. 12 ; 2. 14, 16 ; 3. 8, 14, 19 ; 4. 24 ; 5. 11 ; 6, 1, 3, 11 ; 7. 4, 5 ; 8. 14 ; 9. 4 ; 11. 19, 29, 31 ; 12. 10, 41. Parivrajakah 2. 8 Parve yajnikah . . . 7. 23Brahmanam 1. 16 ; 3. 20 ; 6. 31 ; 7. 12, 18, 17, 23, 28 ; 8. 4, 22 ; 12. 8, 14, 41Yajnikah . . 5. 11 ; 7. 4 ; 11. 29, 31, 42, 43 Varsyayani .1.2Vaiyakaranah . . . … . . . . 1. 12 ; 9. 5 Satabalaksa 11. 6 6akatayana 1. 3, 12, 18Sakapani 2. 8 ; 3. 11, 13, 19 ; 4. 3, 15 ; 5. 3, 13, 28 ; 7. 14, 23, 28 ; 8. 2, 5, 6, 7, 10, 14, 17, 18, 19 ; 12. 19, 40. Sthaulasthlvi … . . . 7. 14; 10. 1 Sakalya . ’ 6. 28 Haridravikam . 10. 5 A LIST OF QUOTATIONS OCCURRING IN THE NXRUKTA, ARRANGED IN THE ORDER OF THE SAMHITAS.
    Rgveda. BOOK QUOTATIONS OCCURRING IN THE NIRUKTA 249 KV.
    250 LIST OF QUOTATIONS BOOK
    n. in. OCCURRING IN THE NIRUKTA 251 V.
    VI. BOOK 252 RV. VIL VIII. LIST OF QUOTATIONS OCCURRING IN THE NIRUKTA 268 BV. 254 LIST OF QUOTATIONS RV. OCCURRING IN THE NIRUKTA 255 Atharva Veda. Stanzas occurring in the $igveda are indicated by an asterisk, AV. 256 LIST OF QUOTATIONS AV.OCCURRING IN THE NIRUKTA 257
    Sdma Veda. sv. 258 LIST OF QUOTATIONS VS.
    OCCURRING IN THE NIRUKTA VS.
    *29. 39 *29. 40 *29. 41 *29. 42 *29. 43 *29. 48 *29. 51

20. 52 *29. 55 *31. 16 *33. 13 *33. 23 *33. 31 *33. 32 *33. 37 *33. 41