365bet娱乐

The Atharva-Pratisakhya is another phonetic treatise relating to the Atharvaveda, Saunakiya recension It consists of three chapters called prapathakas; the first being divided into three sections (padas) and the second and the third into four each. Each pada consists a number of sutras'. The Atharva-Pratisakhya is comparatively a much shorter treatise and its total extent is about half that of the Caturadhyayika As the Caturadhyayika professedly belongs to the Saunakiya recension, it is likely that one may think that the Atharva-Pratisakhya belongs to the Paippalada recension of the Atharvaveda, Saunakiya recension Such is, however, not the case. There are more than 250 instances 2 where the Atharva-Pratisakhya differs from the Paippalada recension (1) There is no unanimity as regards the number of sutras in the Atharva-Pratisakhya Visvabandhu Shastri mentions the numbers of sutras in the three prapathakas respectively as 1-59; 60-143; and 144-220, whereas Surya Kanta gives them as 1-57, 58-140 and 141-223. (2) Cf. Surya Kanta, Atharva Pratisakhya, preface, pp. 32-33. Surya Kanta has shown that even the sanction of the Caturadhyayika is ignored by the Vulgate MSS. and there are cases where the Caturadhyayika mentions or implies certain words which are not found in the Vulgate. The Vulgate is generally supposed to be the Saunakiya recension (See Bloomfield, The Atharvaveda, pp. 13.14; intro. to Kausika Sutra , Journal of the American Oriental Society XIV, p. xxxvii; and Lanman, Harvard Oriental Series VII, pp. xx ff.) xx ff.) But Surva Kanta has tried to prove that the Vulgate does not represent the Saunaka Shakha. He has further pointed out that APrat cites 3236 passages from the Atharvaveda, Saunakiya recension and there

224 Thus it does not operate on the Paippalada recension and both the Caturadhyayika and the Atharva-Pratisakhya may be generally said to refer to the Saunakiya recension; at least their commentators quote from the Saunakiya recension. The Atharva-Pratisakhya refers to words from the first eighteen kandas of the Atharvaveda, Saunakiya recension , the Caturadhyayika also mentions words from eighteen kandas only and this would prove that the Caturadhyayika and the Atharva-Pratisakhya belong to a considerable ant antiquity³. Out of these two manuals the Caturadhyayika must have been composed earlier, for, the Atharva-Pratisakhya refers to the Caturadhyayika at a number of places". In fact the whole of the third prapathaka f.n. 2 continued) are hardly ten passages where the Vulgate MSS. do not follow the sanction of the Atharva-Pratisakhya , while there are hundreds of passages in the Caturadhyayika (which is SaunakIya), where not a single Vulgate Maitrayani Samhita follows the sanction of this treatise. Thus he has tried to show that a) the Atharva-Pratisakhya belongs to the Vulgate, b) the Caturadhyayika goes with the Saunaka school, and c) that the two, namely, the Vulgate and the Saunaka recension are not one. He has suggested that the Saunakiya recension should be reconstructed by referring to the Caturadhyayika and to the commentary on the Atharvaveda, Saunakiya recension ascribed to Sayana who has expressly stated that he is commenting on the Atharvaveda, Saunakiya recension in the Saunakiya Sakha. (3) "The last two books (of the Atharvaveda, Saunakiya recension ) are practically a khila portion and did not probably form part of the first compilation" - C.V.Vaidya, History of Sanskrit Literature (Vedic period), p. 159. (4) Atharva-Pratisakhya 136 Caturadhyayika 2.5 147 e 2.69,70,73,75,76 147 g 201 2.102 2.62 See Surva Kanta, op. cit., p.61.

225 is, in a way, based on the Caturadhyayika For instance, general rules into s, n into n, for restoring the original for changing form and for inserting the avagraha in the pada text are taken for granted and only the exceptions have been recorded in the Atharva-Pratisakhya Another thing that strikes one is the fact that it refers to about three scores of grammatical or phonetic technicalities but it does not care to define a single term. Perhaps, those technicalities are supposed to have been already known from the Caturadhyayika Thus the Atharva-Pratisakhya does not bear the character of an independent work. It is just a supplementary treatise. The first prapathaka of the Atharva-Pratisakhya begins by stating that pada is the basis of samhita (and hence each pada should be known separately with its accent) (1-3). It then goes on to say that words like Brhaspati (II.13.2) and the word pratyancam (XI.3.26) are doubly accented (4-5). Then it mentions that an udatta a may occur between two a letters, 6 e 5 after a or before i and 9 (6-9). Sutras 10-13 lay down that, in the pada text, the nouns and verbs should be shown separated from their prefixes and prepositions. When there are two prefixes or prepositions, the first one should be arated. Sutras 14 to 28 deal with the accent of verbs (5) Cf. Prat. 2.47; Oldenberg, Zeitschrift der deuschen morgenlandischen Gesellschaft IL, p. 172. (6) e.g. saha:a:ihixaxnah agamam (II.30.5); iha:a: vaha (III. 24.7); dhiya:a:ihi:a:nah (II.5.4); iha:a: etu (I. 15.2) etc. (7) pra:sumatim (IV.25.6); abhi: Vipasyami (XII.1.33) etc. Cf. Caturadhyayika 4.1-2.

226 in relative clauses. The verbs in clauses containing Ca or Va or a relative pronoun get accented. The verb in the antithetical clause, where antithesis is indicated by ca-ca (II.6.2) or uta-uta (VII.5.5), is accented. A verb does not lose its accent when connected with ca, va, anya, na hi, yat, it and aha. A vocative, which is preceded by another vocative with an udatta accent on the first syllable, itself gets an udatta accent on the first syllable 8 The Atharva-Pratisakhya then gives lists of words which have an udatta accent on the first syllable (29-32), words which have an udatta accent on the second syllable (33-34), words which have an udatta accent on the last syllable (35-44), words in which all syllables are anudatta (45-51), words which have svarita accent on the first syllable, words which have only one svarita syllable, words which have a svarita preceded by anudatta (52-56) and words which have the kgaipra accent (57). At the beginning of the second prapathaka, the Atharva-Pratisakhya gives 59 passages without specifying the reason for mentioning them. It then proceeds to state that in the words ending in e (like asme, tve etc.), the e is a substitute for 9 case termination (58-59). The cases are then indicated of (8) jatavedas tanuvasin (I.7.2); rame krene asikni ca (I.23.1) etc. (9) ranayantvasme (IV.21.1); tve kratum api (V.2.3) etc. Cf. Caturadhyayika 1.77.

227 .10 the occurrence of the root saj 11 12 and of the elision m of asmakam (60-61). It is then pointed out that in words like Sakalyesi (1.25.2) the samdhi takes the pararupa and that words ending in locative 1 or u are pragrhya (62-63). The word purvya (IV.1.6) is said to be identical with purva and sura (IV.2.4) with surya (64-66). It is pointed out that the word vreabha (IV.5.1 etc.) designates a divinity and edhi (III.28.3 etc.) is the imperative second person singular of root as, and ihi (II.25.4 etc.) of root i (67-69). Lists of passages where the word parastat (as different from purastat, but with the same meaning) occurs and where ai at the end of a pada gets changed into a due to a following a, i or u are given (70-71). It is pointed out that forms ending in m, having a as penultimate are fem. acc. sing. and those with a as penultimate are masc. acc. sing. (72). Sutras 73 to 82 deal with words ending in a visarga in the pada text. Passages are quoted where the visarga is dropped 13 in the samhita Sutras 83 to 95 give words ending in the vowels a, i etc. and which do not have a visarga. Further (10) a sajantu (V.21.10); tva sajamasi (VII.115.1), sajanty amitran (XI.10.3) and a sajami (XIV.2.48). Otherwise the root is sac. (11) asmakarthava (I.7.6); asmakedam (III.24.4) and asmakoti (VII.77.1) (12) Sakalya + esi should be Sakalyaisi, but it remains Sakalyesi. Similar instances are upetana (III.14.3); upesatu (Vedic Index 67.3) and upesantam (VIII.6.17). CS.Caturadhyayika 3.52. (13) Vasa hi satya (I.10.1); abhraja vataja (I.12.3) etc.

228 it is stated that when pra is seen with a long a (i.e. pra) the following verbal is to be regarded as beginning with a (augment); before a consonant, pra is nowhere long 14. When a word ending in ah is followed by the initial a of the following verbal, the visarra is changed to q; if it is followed 15 by a soft consonant, it is not changed to o In esah and syah, the visarga is elided before consonants. When the preposition is followed by a verb in the past, the verb begins with a (augment) (96-102). Ena, eha are pragrhya (103). Sutra 104 gives a list of passages which have praghya words. Compounds with pati at the end (like bhutapati, II.14.4, etc.) are accented on pa (105). Then follow lists of words having two 1, 1, 1, 2, t and a letters. These are followed by a list of words ending in @ in the pada text, but which is changed to a in the samhita 19 (106-112). Sutra 113 gives a list of words which are seemingly single 18 but actually consist of two separate words and sutra 114 mentions words which appear like two words but are, in fact, single It is stated that verbals ending in a are imp. 20 sec. sing. Ma has udatta accent when it means 'negation' 19 (14) The only exception is pra vocan (II.5.5, VII.26.1). (15) tapo 'tisthat (XI.5.26); sa senam (III.1.1) etc. (16) CS. Caturadhyayika 3.33; Rk-Pratisakhya 2.74. (17) kumbhe: a' bhrtah as kumbha abhrtah (1.6.4) etc. (18) sam yoh (1.6.1); rte na (IV.26.6) etc. (19) abhivara (I.32.4); pravisivamsam (IV.23.1) etc. (20) srja (V.12.10); hara (V.20.9); bhara (V.29.5) etc.

229 and anudatta accent when it means 'to me'. An anudatta, when combined with udatta, itself becomes udatta (115-118). The Atharva-Pratisakhya then gives lists of words which have avagraha after n, n, m,su (as a prefix) and sva. The word (The word) te has an anudatta accent when it means 'your' and and udatta accent when it is a pronoun meaning 'they' (119-123). A list is then given of passages containing words having the root da to give (124.126). It is pointed out that the word idam (in all genders) is antodatta in the first mention or at the beginning of a pada. It is stated that separation by means of an avagraha is to be made with what remains after elision 21 (127-131). The passages containing the causal forms of the roots pat and pad are given and the distinction between these two roots is shown (132-134). Sutra 135 lists occurrences of the word tyam which is the same as tam. Rules regarding the samdhi of consonants, especially before nasals' are then mentioned and lists of passages containing words ending in n, m and t in the pada text are given (136-140). 22 The third prapathaka of the Atharva-Pratisakhya deals mainly with avagraha. At the outset, passages are quoted wherein g is to be repeated in the pada in place of a visarga 23. It is (21) brahmabhih (X.10.23) as brahma' bhih (not brahman); dirghavutvaya (XII.2.6) as dirghayu'tvay (not ayus). (22) Cf. Caturadhyayika 2.5; Rk-Pratisakhya 4.3; Taittiriya Pratisakhya 6.2; Vajasaneya Pratisakhya 4.120. (23) akah (II.25.1) as akar ity akah; dvah (IX.3.22) as dvar iti dvah; etc.

230 stated that a vocative ending in o remains unchanged before 24 a non-Vedic iti (and if it is followed by a vowel in the mantra, o is changed a). A list of passages containing the root dha with api is given; so too of passages in which the same word is repeated twice (amredita). Usually the letters ,,s and long vowels are changed to s, n, visarga and short vowels in the pada text. A list of passages which are an exception to this rule is given (141-147). Passages are also quoted in which a visarga does not change into though 25 followed by k, as well as those in which a short vowel is 26 not lengthened though followed by mat . Normally in copulative compounds each noun has the dual form and its own accent. Suryacandramasa (XI.3.34) and Suryamasayoh (111.29. 5) are pointed out as exceptions. Two passages are quoted wherein the final vowel in prthivi is shortened (148-152). 28 27 Similarly three passages, in which the first vowel is elongated, are pointed out and also a few other passage wherein the elongation takes place before y. It is said that in the two words sapatheyya (V.31.12) and sahaseyyava, e (24) Cf. Vajasaneya Pratisakhya 1.94; Taittiriya Pratisakhya 4.6; Caturadhyayika 1.80-81. (25) ayuh krnoty (VII.32.1,33.1); agnih krnotu (Vedic Index 106.3); and manih krnotu (1.6.31). (26) Virvavan (VIII.5.1); Viryavattara (XVIII.4.38). (27) XVIII.3.49, 4.78. (28) Sauprajastvam (II.29.3); nairbadhyena (Vedic Index 75.1) and saubhagatvaya (XIV.1.50). MEMEDE

231 (and not a) is followed by y. Then a list of passages in which there is sandhi between a and a is given (153-157). It is pointed out that in the words, apavan (XVIII.4.24) and jyotisiman (XVIII.4.14), there is the agama of a and i respectively. Usually the word havisa is followed by vidhema; two passages are quoted where the word is followed Passages wherein the word nu occurs instead of juhomi 29 anu are mentioned and also passages having the root hra with ut. The word mahl occurs sometimes as singular and at other times as dual. Passages wherein the word mahl is singular are pointed out and also the passages in which one t is written in place of two t's (158-165). Sutra 166 refers to optional elision of visarga and the next two sutras quote passages containing the root i in aorist and the root i with pra. It is stated that, in verbs like pretam (I.27.4) the preposition is separated by an avagraha. Similar is the case when pra is joined to the indeclinable itah (166-170). Then passages are given where t occurs between two r letters and occurs between two r letters³1. Passages having conjunct consonants ending in y and h are quoted together with those which have conjuncts of t and m 30 32 g and m, k and n as well as p and n (171-179). Then passages having conjuncts beginning with k and I are quoted. (29) I.15.1; Vedic Index 64.2. (30) vartra (1.3.7) and kartra (X.1.19,30,32) (31) prasarsrana (Vedic Index 39.1). (32) These conjuncts are pointed out possibly to show that there occur & twin-sounds (yama) between them.

232 It is pointed out that in the word asvyena (VIII.3.15) the conjunct ends in y. Conjuncts with h and y as well as m and I are shown and it is stated that in the word tvastrova (XII.3.33) the word tvastra is in instrumental case and that the conjunct ends in r. Passages are quoted in which there are conjuncts of dh and m as well as 1 and m. Then passages having conjuncts beginning with r and n and conjuncts having I and y as well as bh and y are quoted (180-194). Then passages having conjuncts beginning with d are given; 77 passages are quoted wherein a being followed by a 33 拢34 vowel is changed to ; and when it is changed to r with the nasalisation of the preceding vowel. The exceptions, like cikitvan tvam (V.12.1), are pointed out together with seven cases where the elision of visarga occurs (195-200). Passages are then mentioned wherein a visarga is changed to and where it is not so changed. It is stated that a is changed into a when it is preceded by a non-preposition .35 or by the first member of a compound (201-206). Sutras 207-216 are devoted to a treatment of the lengthening of the vowels u, i, a at the beginning, in the middle or at the end of words. Sutras 217-220 point out peculiar cases of compound words and show how they are to be divided into their (33) Cf.Caturadhyayika 1.68. Similar samdhi is found in V I.44.1, 45.1, 47.5, 48.14 etc. (34) Cf. Caturadhyayika 2.29; Prat. 4.69-70; Taittiriya Pratisakhya 9.20; Vajasaneya Pratisakhya 3.140 (35) Cf. Caturadhyayika 2.98.

component parts 36 233 Sutras 221 to 223 conclude the Pratisakhya by stating that, in the sacred texts, letter, gender, accent, case, number etc. are liable to be reversed and that, out of mere conjecture or for grammatical consideration, one should not change the traditional wording of the sacred text. The holy parisat (recension) itself is its true grammar and if any word has not been explained in this book, it should be understood from the parisat itself. 37 that It has been rightly emphasized by Surya Kanta the Atharva-Pratisakhya is chiefly meant to elucidate the pada text rather than the samhita and hence it bears a character slightly different from that of the other Pratisakhyas. Its peculiar contribution lies in the analysis of avagraha between verb and preposition and in the treatment of accentuation which is acute and exhaustive. But, as Surya Kanta has pointed out, even here it has failed to record important notables. It has, for instances, not explained the anamolies where the Atharvaveda, Saunakiya recension (36) For instance, the word samanta in the sense of 'wholly' has udatta accent on the final syllable and it is not to be divided. When it means 'completion' it has udatta accent on the first syllable and it is to be divided as sam anta. The word samana is to be divided only in one case (X.2.13). (37) op. cit., notes, p. 53.

234 38 reads a verbal with accent, but the same cannot be explained. It has been already noted that the Atharva-Pratisakhya depends upon the Caturadhyayika Especially it seems as if the third prapathaka of the Atharva-Pratisakhya were to record exceptions to the general rules of the Caturadhyayika We find that the Atharva-Pratisakhya freely makes use 39 of technical terms like pragrhya, amredita, samapatti (restoration of the original form in the pada text) and upacara (Conversion of visargas before k or pl 40; but to A these terms are not defined and their knowledge is taken for granted. Incidentally, it may be noted that the Atharva-Pratisakhya does not differentiate between two cases of words which carry double accent - 1) archaic genitive compounds like Brhaspati and 2) copulative compounds like Mitravarupa 41 Like the Caturadhyayika the Atharva-Pratisakhya has adopted the gana method. 42 The ganas mentioned herein are Brhaspatyadi (4), vvaghradi (19), ajaradi (34), Sakalyesadi (62), ena eha (38) op. cit., preface, p. 68. The verbals are: pratyetu (III.2.1), askabhayad (IV.1.4), asti (IV.19.2), sahasva (IV.31.2); stha (Vedic Index 21.3), stam (Vedic Index 78.2) and kran (Vedic Index 131.3). (39) Cf. Prat 13.31; Caturadhyayika 4.73: prakrtidarsanam samapattih (40) Cf. Caturadhyayika 2.62; Rk-Pratisakhya 4.41; 13.131. (41) Atharva-Pratisakhya 4: dvir udattam brhaspalyadinam. Panini has enunciated two sutras to distinguish between these two kinds (vi.2.140,141). (42) This gana is found in Vajasaneya Pratisakhya (2.47) where 12 words beginning with Brhaspati have been mentioned.

235 adi (103), bhutadi (105), asmaisatadi (127), and dirghayutvadi (131). By adopting the gana method the Atharva-Pratisakhya has made its extent concise, but this saving is made "at the expense of its reputation and value as an authority for 43 the readings of its text" The Pratisakhyas are not, strictly speaking, grammatical treatises. They are, at the best, concerned with recording phonetic and other kindred peculiarities which characterise the samhitas of their schools. But as Visvabandhu Shastri has pointed out, of all such treatises, those relating to the Atharvaveda, Saunakiya recension are perhaps most grammatical in character, that is, they bear greater affinity to regular grammar and have, to a large extent, adopted grammatical technique 44. Another significant point regarding the phonetic manuals belonging to the Atharvaveda, Saunakiya recension is worth noting. The Pratisakhyas generally refer to divergent opinions of different authorities on grammar. Even Panini does it. It is, however, strange that the Caturadhyayika and the Atharva-Pratisakhya hardly 45 refer to any other authority either for approbation or for general enunciation. These treatises are, indeed, conspicuous in not referring to any contemporary or ancient teacher of grammar and phonetics. (43) Surya Kanta, op. cit., notes, p. 53. (44) The Atharva-Pratisakhya, intro., p. 31. (45) Caturadhyayika 2.24 refers to Sakatayana.

236 There is one more point. Phonetics forms an essential part of the Pratisakhya literature, and is, therefore an original constituent of it. A striking feature which at onee differentiates the Caturadhyayika and the Atharva-Pratisakhya from other kindred treatises is the fact that the former do not contain any general phonetic introduction. For instance, the Caturadhyayika and the Atharva-Pratisakhya do not say anything about the organs of speech. Nor do they deal with the classification of sounds. This lacuna is filled by one of the Paristas 46. GOLEGA MED! (46) Varnapatala, Atharvaveda-Parisistas 53. See the section on Philology in Part II. S