Pramāṇa · Search · About · API

A History of Sanskrit Literature c31.s3: 3. The Authenticity of the Arthaçāstra

A History of Sanskrit Literature c31.s3

Language
English

3. The Authenticity of the Arthaçāstra

c31.s3.1

The current belief¹ which ascribes the *Arthaçāstra* to Cāṇakya or Viṣṇugupta or Kauṭilya, minister of Candragupta, rests on the verse just cited, on statements at the end of i. 1 and ii. 10 where Kauṭilya—the variant Kauṭalya has no value, being obviously a correction—appears as the author, and in the latter of which he claims to have gone through all the sciences and to have had regard to practice (*prayoga*), while a verse added at the very end, after the last colophon, says that Viṣṇugupta composed both the text and the comment, apparently because he noticed that in other cases there was discrepancy between these two important elements of a scientific work. These statements are taken to offset the fact that, by using the phrase *iti Kauṭilyaḥ* to give normally the deciding opinion in discussions, one would conclude that the work was not by the author, but was the product of a school which followed his views, as in the case of Jaimini or Bādarāyaṇa in the philosophical Sūtras. It must, however, be noted that under the explanation of Apadeça in the last book is cited one of Kauṭilya's sentences, from which the prima facie conclusion is that Kauṭilya is cited as an authority, not as the author. The case, therefore, must be solved by considerations of general probability based on what we know of Cāṇakya, and what we find in or are told about the work.

c31.s3.2

¹ Jacobi, SBA. 1911, pp. 732 ff., 954 ff.; 1912, pp. 832 ff.; ZDMG. lxxiv. 248 ff., 254, and the editors other than Jolly. Against this view see Winternitz, GIL. iii. 518 f.; Bhandarkar, POCP. 1919, i. 24 ff.; Keith, JRAS. 1916, pp. 130 ff.; 1920, p. 628; EHR. 1925, pp. 420 f.; JCL. vii. 275 f.

c31.s3.3

It is significant that, though we hear of Cāṇakya in the Purāṇas and later texts as the minister of Candragupta, and though the *Mudrārākṣasa* makes an interesting figure of him, we have not the slightest reference there or elsewhere to his literary activity. Doubt has even been cast on his historic character, for Megasthenes, the ambassador of Seleukos who spent a considerable time at the court of Candragupta, does not mention him; but, owing to our fragmentary knowledge of Megasthenes, this argument cannot be stressed. Nor can we make much progress by discussing the probability whether an Indian statesman would write memoirs like Bismarck, for, while the indifference to morality and the insistence on distrust as a quality of a wise king are common to both, there is all the difference in the world between the detailed accounts of real events in which he figured given in Bismarck's *Gedanken und Erinnerungen*¹ and the absolutely general and very pedantic utterances of the *Arthaçāstra*, which never anywhere hints that its author had any knowledge of the overthrow of the Nandas and the wars which brought Candragupta his empire and the cessions made by Seleukos. His sovereign's name, his family, what is still more amazing his country, his capital, are passed over in absolute silence by this alleged ancient statesman meditating in his days of retirement on the maxims of policy. The rules laid down are those which might be valuable for a moderate-sized state, and ignore entirely the issue of the government of an empire such as that of Candragupta. So complete does the impossibility of such silence appear to be that one critic,² accepting the genuineness of the ascription, explains the book as written before Candragupta acquired the empire. This is a candid admission but really serves to prove that the claim is absurd.

c31.s3.4

¹ Stuttgart, 1898.

c31.s3.5

² Smith, EHI. p. 146.

c31.s3.6

Efforts have naturally been made to find at least striking resemblances between the account given in the *Arthaçāstra* and the fragments of Megasthenes. The effort is a complete failure;¹ coincidences there are many in number, but on matters which hold good of India generally in the period before and after Christ. The vital resemblances of important detail are absolutely lacking, even when we put aside all those statements of the Greek author which rest doubtless on misunderstandings or are obscurely reported. The *Arthaçāstra* knows nothing of the wooden fortification of Pāṭaliputra but provides for stone work; it ignores the boards of town officials without any head of each, but engaged in co-operation which Megasthenes specifies; it knows nothing of the commander-in-chief of the fleet, and a regular navy such as Candragupta must have used, but which was probably of minor account in many states. The care of strangers, escorting them to the border, seeing after their effects if deceased, are unknown to the *Arthaçāstra*, which does not provide for the registration of births and deaths, while the work of Megasthenes' board in selling old and new manufactured articles contrasts strikingly with the highly developed commercial and industrial conditions envisaged by the *Arthaçāstra*. Megasthenes' statement as to the king's ownership of the land is supported by other Indian evidence; it is not the view of the *Arthaçāstra*; Megasthenes describes a knowledge of minerals far less advanced than that of the *Arthaçāstra* which knows much of alchemy; the taxes of Megasthenes are simple as compared with the numerous imposts of the text, and, while Megasthenes ignores writing, the *Arthaçāstra* is full of rules on registration, the preparation of royal documents, and recognizes passports.²

c31.s3.7

¹ Stein, *Megasthenes und Kauṭilya*, SWA. 1921.

c31.s3.8

² The metre of the work is not early, and its grammatical irregularities are not primitive; Keith, JRAS. 1916, pp. 136 f.

c31.s3.9

If we abandon the unhappy identification, the date becomes difficult to settle. We may, however, note that Patañjali does not know the work, that the knowledge of alchemy suggests acquaintance with Greek science,³ and that the term *suruṅgā*, mine, is doubtless borrowed from Greek syrinx, probably not until after the Christian era.⁴ Moreover, it seems most probable that the *Arthaçāstra* knew and used the Smṛtis of Manu, Yājñavalkya, and Nārada at least; in the case of Yājñavalkya the case appears to be proved;¹ where that text penalizes operations on boils, the *Arthaçāstra* sensibly excludes operations on dangerous boils, and in other cases it takes up the language of the Smṛti. The identity of the exhortation to the soldiers with a verse of Bhāsa may denote borrowing, but, as Bhāsa's date is uncertain, this does not help much to a definite result. The text was doubtless known to Daṇḍin who mentions its length, 6,000 Çlokas (i. e. sets of thirty-two syllables), and considers it as recent, unless we put this down to dramatic propriety in his notice; Bāṇa recognizes it, and Kālidāsa's remarks on hunting were perhaps taken from it. This accords well with the fact that the *Kauṭilīya* and Cāṇakya are known to the *Nandīsūtra* and *Anuyogadvāra-sūtra* of the Jain canon in the middle of the fifth century A.D., and that Varāhamihira in his *Bṛhatsaṁhitā* has parallel matter, while Caraka's medical treatise enumerates thirty-six special devices as compared with thirty-two of the last book of the text. Further, the work is before the *Kāmasūtra*, whose date, as will be seen, may be the fourth century A.D., before Vātsyāyana's *Nyāyabhāṣya*, and before the *Tantrākhyāyika* or *Pañcatantra*, perhaps of the same period. That the work was a product of *c.* 300, written by an official attached to some court, is at least plausible, if it cannot be proved. Whether anything goes back to Cāṇakya is an insoluble question. The author may have lived in the south, since he refers to the pearls, diamonds, shells, and gems of that part, and South Indian and Ceylonese gems bulk largely in the chapter on the examination of gems, but this is conjecture, for the fact that manuscripts exist only in the south is not of much importance.

c31.s3.10

³ Jolly's ed., pp. 42 f., against Ray, *Hist. of Hindu Chemistry*, ii. 31; R. V. Patvardhan, *POCP.* 1919, i, p. clv.

c31.s3.11

⁴ Stein, ZII. iii. 280 ff.; Winternitz, IHQ. i. 429 ff.

c31.s3.12

¹ As shown by T. Gaṇapati Śāstrī, TSS. 79, pp. 8 ff. A defence of the antiquity of the work is given by Narendranath Law (*Calc. Review*, Sept. Dec. 1924) and K. P. Jayaswal (*Hindu Polity*, App. C), but neither of these authors explains why the author knows nothing of an empire or Pāṭaliputra. ‘Credo quia impossibile’ is still, it appears, not obsolete.

c31.s3.13

The literature known to the text included Vedas, Vedāṅgas, epic, didactic and narrative, Purāṇas, Itivṛttas, Ākhyāyikās, and probably a large number of texts on special sciences such as examination of jewels, agriculture, military matters, architecture, alchemy, veterinary art, and other topics. The theory that the information given in the text was merely derived from fellow experts is contradicted by the express assertion of the text and all probability. The text also knew Jain legends, Jain gods and technical terms, while its version of epic legends is by no means always derived from the great epic, but may be paralleled in the Vedic and Buddhist literature. All this coincides with the date above suggested as likely.

Linked texts (15)