The same inference may also be drawn from another series of facts, viz. the peculiarities of the language of his Sutras. The latter are very considerable and very remark- able. They may be classed under four heads. In the Apastambiya Dharma-sutra we have, first, archaic words and forms either occurring in other Vedic writings or formed according to the analogy of Vedic usage ; secondly, ancient forms and words specially prescribed by Pa#ini, which have not been traced except in Apastamba's Sutras ; thirdly, words and forms which are both against Vedic usage and against Pa«ini's rules, and which sometimes find their analogies in the ancient Prakrits ; and fourthly, anomalies in the construction of sentences. To the first class belong, kravyadas, I, 7, 2i; 15, carnivorous, formed according to the analogy of ris&das; the frequent use of the singular dara, e.g. II, i, i, 17-18, a wife, instead of the plural dara^; salavr/ki, I, 3, 10, 19, for salavr/kt ; the substitution of'/ for r in plehkha, I, u, 31, 14; occa-
Ind. Stud. V, 65-74 ; XIII, 443, 444. As regards the explanation of Katy4- yana's and Pata%-ali's words, I side with Kaiya/a and Professor Goldstiicker. But I am unable to follow the latter in the inferences which he draws from the fact, that Katyayana and Pata^f^ali declare Y>favalkya and other sages to be as ancient as those whose Brahmawas and Kalpas are designated by the plural of adjectives formed by the addition of the affix in to the names of the promul- gators. Though Pftwini asserts, IV, 3, 105, that only those Brahmawas which are known by appellations like BhallavinaA, KaushitakinaA, &c., have been proclaimed by ancient sages, and though Katyayana and the author of the Great Commentary add that this rule does not hold good in the case of the work called Ya^avalkani Brahmawani, it does not necessarily follow, as Professor Goldstiicker thinks, that an extraordinarily long interval lies between Pawini and Katyayana — so long a period that what Pawini considered to be recent had become ancient in Katyayana's time. Professor Weber has rightly objected to this reasoning. The difference between the statements of the two grammarians may have been caused by different traditions prevailing in different schools, or by an oversight on the part of Pamni, which, as the scene of Ya^flavalkya's activity seems to have been Videha in eastern India, while Pa«ini belonged to the extreme north-we^t, is not at all improbable. As regards the two dates, I place, following, with Professor Max Miiller, the native tradition, Katyayana in the fourth century B. c., and Pata%ali, with Professors Goldstiicker, Kern, and BhaWarkar, between 178-140 B.C.
Xliv APASTAMBA.
sional offences against the rules of internal and external Sandhi, e.g. in agrrhyamanakara7/a/^, I, 4, 12, 8; in skup'tva. I, u, 31, 22, the irregular absolutive of skubh or of sku ; in paduna, I, i, 2. 13 ; in adhajana-rayin,
I, i, 2, 21 ; and in sarvatopeta, I. 6, 19, 8 ; the neglect of the rule requiring vrzddhi in the first syllable of the name Pushkarasadi, I, 10, 28, 1; the irregular instru- mental vidya, I, n, 30, 3, for vidyaya, and ni/^rreyasa,
II, 7, 16, 2, for ni//^reyasena ; the nominatives dual avam, L 7,20, 6, for avam, and kru#£akrau«£a, I, 5, 17, 36 for °krau££au ; and the potentials in it a, such as prak- shalayita, I, i, 2, 28 ; abhiprasarayita, I, 2, 6, 3, &c.
Among the words mentioned by Pa//ini, but not traced except in the Dharma-sutra, may be enumerated the verb str/h, to do damage, I, n, 31, 9; the verb srinkh, to sneeze, from which jr/hkhanika, I, 5, 16, 14, and ni/*- jr^'hkhana, II. 2, 5, 9, are derived; and the noun veda- dhyaya, I, 9, 24, 6 ; II, 4, 8, 5, in the sense of a student of the Veda. Words offending against rules given by Pawini, without being either archaic or Prakritic, are e.g. sar- v an n in, I, 6, 18, 33, one who eats anybody's food, which, according to Pa,vini V, 2, 9, should be sarvannina; sarpa^irshin, I, 5, 17, 39 ; annasawskartr/, a cook, II, 3,6, 16; dharmy a, righteous, for dharmya, I, 2, 7, 21, and elsewhere; divitrt, a gambler, II, 10, 25, 13, for devitri, the very remarkable form prajwati, I, i, 4, i, for pra^nati, finds an analogy in the Vedic jnyaptre for jnaptre1 and in Pali, pawha from prajwa for pra^na; and the curious compounds avahgagra, I, i, 2,38, paran- gavr/tta, II, 5, 10, u, where the first parts show the forms of the nominative instead of the base, and pratisurya- matsya//, I, 3, n, 31, which as a copulative compound is wrong, though not without analogies in Prakrit and in later Sanskrit 2. The irregular forms caused by the same ten- dencies as those which effected the formation of the
1 Wackernagel, Altindische Grammatik, vol. i, p. xxxiii.
3 See ZHtschr. d. Deutschen Morg. Ges., vol. xl, p. 539 seq. ; Epigraphia
Indica, vol. i, p. 3.
INTRODUCTION. xlv
Prakrit languages, are, aviprakramiwa, II, 2, 5, 2, for aviprakramawa, where an # standing in thesi has been changed to i; samvrtttl/i, II, 3, 6, 13, sawvartete, II, 5, if, 20, and paryanta, I, 3, 9, 21, and I, 3, n, 33 (compare Mara//n awt for anta^), in each of which a standing before a nasal has been lengthened ; awika, I, 6, 19, i, the initial a of which stands for ri, if it really has the meaning of r/wika. as some commentators asserted ; anulepawa, 1,3, if, 13; I, n, 32, 5, with the Prakritic change of na to wa; vyupa^-ava, I, 2, 8, 15, with va for pa ; rz'tve for rit vye, wherej seems to have been absorbed by the following e; apa-fjrayita, I, n, 32, 16, for ap^^ra- yita, and bhatrzvyatikrama, I, fo, 28, 20, where r has been assimilated to the preceding, or has been lost before the following consonant. The irregularities in the construction are less frequent. But in two Sutras, I, 3, 10, 2, and I, 3, i j , 31, some words which ought to stand in the locative case have the terminations of the nominative, and it looks as if the author had changed his mind about the construction which he meant to use. In a third passage II, 10, 26, 20, .n.ma£/£^edana;tf savr*sha«asya, the adjective which is intended to qualify the noun .mna has been placed in the genitive case, though the noun has been made the first part of a compound.
The occurrence of so many irregularities1 in so small a treatise as the Dharma-sCitra is, proves clearly that the author did not follow Pa//ini's grammar, and makes it very unlikely that he knew it at all. If the anomalous forms used by Apastamba all agreed with the usage of the other Sutrakaras, known to us, it might be contended that, though acquainted with the rules of the great grammarian, he had elected .to adopt by preference the language of the Vedic schools. But this is by no means the case. The majority of the irregular forms are peculiar to Apastamba. As it is thus not probable that Apastamba employed his peculiar expressions in obedience to the tradition of the
1 Many more may be collected from the other divisions of the body of Siltras. See VYinteraitz, op. tit., p. 13 seqq. ; Gurupq^akaumudi, p. 34 seq.
APASTAMBA.
Vedic schools or of his particular school, he must have either been unacquainted with Pa;/ini or have considered his teachings of no great importance. In other words, he must either have lived earlier than P£«ini or before Pawini's grammar had acquired general fame throughout India, and become the standard authority for Sanskrit authors. In either case so late a date as 150 B. C. or the first century B.C. would not fit. For Pataw^ali's Mahabhashya furnishes abundant proof that at the time of its composition, in the second century B.C., Pawini's grammar occupied a position similar to that which it holds now, and has held since the beginning of our era in the estimation of the learned of India. On linguistic grounds it seems to me Apastamba cannot be placed later than the third century B.C., and if his statement regarding Svetaketu is taken into account, the lower limit for the composition of his Sutras must be put further back by 150-200 years.
But sufficient space has already been allotted to these attempts to assign a date to the founder of the Apastambiya school, the result of which, in the present state of our knowledge of the ancient history of India, must remain, I fear, less certain and less precise than is desirable. It now is necessary to say, in conclusion, a few words about the history of the text of the Dharma-sutra, and about its commentary, the U^vala Vrztti of Haradatta. The oldest writer with a known date who quotes the Apastam- biya Dharma-sutra is Sarikara&irya \ c. 800 A.D. Even somewhat earlier Kumarila, c. 750, refers repeatedly to a law-book by Apastamba 2. But it is improbable that he had our Dharma-sutra before him. For he says, p. 138, that Apastamba expressly sanctions local usages, opposed to the teaching of the Vedas, for the natives of those dis- tricts where they had prevailed since ancient times. Now, that is just an opinion, which our Dharma-sutra declares to be wrong and refutes repeatedly 3. As it seems
1 See Deussen, Vedanta, p. 35.
a Tantravirttika, pp. 138, 139, 142, 174, 175, 179, Benares ed.
3 Ap. Dh. I, i, 14, 8, 9-10; II, 6, 14, 10-13; II, 6, 15, i.
INTRODUCTION. xlvtt
hazardous to impute to a man, like Kum&rSla, ignorance or spite against Apastamba, I am inclined to assume that the great Mimawsaka refers to some other work, attributed to Apastamba, perhaps the metrical Apastamba-smr/ti which Apararka quotes very frequently l. Among the commen- tators on Smrztis the oldest, who quote the Dharma-sutra, are Medhdtithi, the author of the Manubhashya, and Vi^flfanejvara, who composed the Mitakshard, the well- known commentary on Ya^wavalkya's Dharma-jastra during the reign of the A'alukya king Vikramdditya VI, of Kalya«a towards the end of the eleventh century. From that time downwards Apastamba is quoted by almost every writer on law. But the whole text, such as it is given in my edition2, is vouched for only by the com- mentator Haradatta, who wrote his U^fvalS. VWtti, at the latest, in the fifteenth century A. D. or possibly 100 years earlier3. Haradatta was, however, not the first commen- tator of the Dharma-sutra. He frequently quotes the opinions of several predecessors whom he designates by the general expressions anysJt or apara/*, i. e. another (writer). The fact that the \Jggva\& was preceded by earlier commentaries which protected the text from cor- ruption, also speaks in favour of the authenticity of the latter, which is further attested by the close agreement of the Hirawyakeji Dharma-sutra, mentioned above.
As regards the value of the tJggvala" for the explanation of Apastamba's text, it certainly belongs to the best com-
1 Ap. Dh., Introd., p. x.
* Apastamblya Dharma-sutram, second edition, Part i, Bombay, 1892 ; Part ii, Bombay, 1894.
3 It seems not doubtful that Haradatta, the author of the U<ggval&, is the same person who wrote the Anakula Vre'tti on the Apastamblya Grrhya-sfitra, an explanation of the Apastamblya Grihya-mantras (see Burnell, Ind. Ant. 1, 6), and the Mhakshara Vr/tti on the Dharma-sutra of Gautama. From the occurrence in the latter work of Tamil words, added in explanation of Sanskrit expressions, it follows that Haradatta was a native of the south of India. I am not in a position to decide if our author also wrote the Padamaw^ai! Vr/'tti oa the Kirika of Vamana and (7ayaditya. This is Professor Aufrecht's opinion, Catalogus Catalogorum, p. 754 seq. See also my remarks in the Introd. to the second eoL, p. viii.
xlvili APASTAM1JA.
rnentaries existing. Haradatta possessed in the older Vr/ttis abundant and good materials on which he could draw ; he himself apparently was well versed in Hindu law and in Sanskrit grammar, and distinguished by sobriety and freedom from that vanity which induces many Indian commentators to load their works with endless and useless quotations. His explanations, therefore, can mostly be followed without hesitation, and, even when they appear unacceptable, they deserve careful consideration.
