Maṇimēkhalai in its Historical Setting

V. Other Views on the Philosophical Systems

The foregoing account of Books XXVII, XXIX and XXX of the Maṇimēkhalai follows faithfully the text of the work, but the exposition of it is entirely my own. It would have become obvious to the reader who has perused the whole of it with any care, that perhaps other views than those expounded above can possibly be urged and other conclusions drawn with very considerable justification. It may be as well that those other lines are [86] considered and my reasons for taking the line that I have taken indicated as a necessary supplement to my exposition of the subject. In regard to this part of the subject, I have had the great advantage of discussion with a scholar of the eminence of Professor Jacobi of Bonn who did me the kindness to look through the manuscript portion relating to the translation of these books and the whole of my exposition thereof. As his criticism is quite typical of the views possible, I set them forth, as far as may be, in his own words, with a running commentary of my own as perhaps the best way of explaining the position.

The first point to call for attention relates to the remarks of the learned Professor regarding Ktakōṭi. Ktakōṭi, it will be remembered, is a name which occurs in the Maṇimēkhalai along with those of Vēdavyāsa and Jaimini among those who were regarded as authoritative expounders of Vaidika Pramāṇas (instruments of knowledge resting upon the Vēda for their authority). Professor Jacobi writes in his letter dated the 28thApril 1927:–

‘Your explanation of Ktakōṭi as the name of the first commentary on the Mīmāṁsa Sutras is of great importance. The vexed question about Bodhāyana and Upavarṣa is brought nearer its solution by your discovery. In connection with it I may be allowed to make the following remarks:

(1) The Vttikāra cannot be equated with Ktakōṭi, if the report of the Maṇimēkhalai may be trusted. For Ktakōṭi taught eight pramāṇas, and the Vttikāra but six; see Śabarasvāmin’s quotation from him ad. I, 1,5 (p. 10, Bibli. Ind. Ed.)

(2) Vēdavyāsa in connection with the pramāṇas cannot be Bādarāyaṇa, since no school of the Vēdāntins is known to have admitted ten pramāṇas, but some acknowledged three, some six. (In the Sūtras of either Mīmāṁsa occur only the three original pramāṇas, as acknowledged by Sāṁkhya. In the Vedanta Sutras arthāpatti does not occur; in the [87] Mīmāṁsa Sūtra, it does occur twice or thrice but the word there denotes something quite different, in no way connected with the pramāṇa arthāpatti). Perhaps Vēdavyāsa should be taken to be the author of the Purāṇas, though the Paurānikas acknowledge eight pramāṇas according to the usual tradition.’

We shall consider the points in this extract seriatim. Ktakōṭi is the name of a person according to the Maṇimēkhalai. But what is said of Ktakōṭi in the Prapañcahdaya passage quoted in extenso shows Ktakōṭi to be the name of the commentary from which the author must have been subsequently named Ktakōṭi. The very formation of the word seems to indicate it as a personal title, though, according to our authority, it is unmistakably the name of the work. The Prapañcahdaya passage quoted makes it clear beyond all possibility of doubt that Ktakōṭi is the author of the commentary as a whole, and Upavarṣa was the expounder of that commentary who, for convenience of teaching and reading, felt it necessary to make an abridged edition of it. That is the position according to this Sanskrit work.

But the point to which the professor takes exception is whether Ktakōṭi can be identified with the Vttikāra as I have taken it in my exposition. The commentators quote largely a commentary as the work of a Vttikāra. They do not, in the great majority of cases, give any other name than that of Vttikāra. But the Prapañcahdaya clearly states that it was Bodhāyana that wrote the commentary Ktakōṭi, and that Upavarṣa’s work was no more than an abridgement. So where it is quoted as distinct from Upavarṣa, sometimes in juxtaposition, the possibility seems to be that Bodhāyana is quoted under the name Vttikāra as the Srī Bhāṣya of Rāmānuja refers to Bodhāyana Vtti as an authoritative work ‘following the [88] text of which closely he writes his own commentary on the Brahma Sūtras’ or Uttara Mīmāṁsa.

The Prapañcahdaya makes the statement clear that Bodhāyana wrote the commentary on the whole of the Mīmāṁsa, Pūrva and Uttara, and Upavarṣa’s abridgment similarly takes into in both the sections of the Mīmāṁsa. But Professor Jacobi’s point of objection is that the Vttikāra formulated eight pramāṇas according to the Maṇimēkhalai. If Ktakōṭi should be taken to be the Vttikāra, the eight pramāṇas are nowhere mentioned in Vēdānta works as formulated by him. He takes the six pramāṇas clearly stated in the commentary on Sūtra I.I.5 of the Pūrva Mīmāṁsa in Sahara’s commentary to be a quotation from the Vttikāra. My point against this is that the position seems to be supported indirectly by what Professor Keith has to say of the quotation from the Vttikāra in this commentary, that the six are the pramāṇas accepted by the commentator as the pramāṇas of the Sūtrakāra, that is Jaimini. As against this, the Professor points out in a later letter dated the 12th July, 1927.

‘You say that the six pramāṇas, mentioned first in the extract from the old Vttikāra, were those formulated by Jaimini, the Sūtrakāra himself. But by the evidence of the Sūtras their author knew only the first three pramāṇas and no more.’

I must admit I had overlooked this point. But then the point admits of an explanation. In that particular chapter, Jaimini is apparently considering the pramāṇas generally admitted as such by other Śāstrakāras, it may be the Naiyāyikas and others, and of these, for the purposes of the Mīmāṁsa Śāstra, the first part of it in particular, he rejects the first three as of no validity and accepts the fourth, Śabda alone, as of valid authority. That position of the Sūtrakāra does not appear to me to [89] be necessarily inconsistent with Jaimini’s regarding six as valid pramāṇas generally from the point of view of the systems of the followers of the Veda as a whole. This is confirmed by the Prapañcahdaya which treats these six as laukika (secular). With such a clear statement as is found in the Maṇimēkhalai any other explanation would make the author absurd, and I believe it would be carrying criticism too far to ascribe to an author an absurdity of this gross kind.

In regard to Vēdavyāsa himself, the Professor’s point is that no Vēdānta work has accepted ten pramāṇas as such. But that is not incompatible with Vēdavyāsa having formulated ten pramāṇas.

In the section under consideration, the Maṇimēkhalai deals with the Vaidika pramāṇas as such, not those of the Mīmāṁsa Śāstra alone. Among those who dealt with the subject of pramāṇas generally the work mentions three, Vēdavyāsa formulating ten pramāṇas as valid, Ktakōṭi eight and Jaimini six. The formulating of these as generally acceptable, is not incompatible with the position that where particular sciences or Śāstras are taken into consideration, these get reduced, for the purposes of the particular Śāstra, to a smaller number. That seems the only satisfactory way of accounting for the varying numbers of the pramāṇas that we find in the different Śāstras. But in regard to Vyāsa himself, the Professor is of opinion that probably he was not Vyāsa Bādarāyaṇa, the author of the Brahma Sūtras.

Indian tradition seems to be uniform in regarding Vēdavyāsa as actually the author of the Brahma Sūtras. No treatise of his on pramāṇas has come down to us, but Vēdavyāsa is regarded as a teacher, and it is just possible to believe he taught, as a necessary preliminary Vaidika pramāṇas as such, and in the course of that teaching [90] formulated ten pramāṇas. This position seems to find confirmation in the passage quoted from the commentary Kāśika on the Ślokavārttika where the names of Ktakōṭi and Vēdavyāsa are brought into connection. Vēdavyāsa is there indicated by the name Pārāśarya, son of Parāśara, and Ktakōṭi is there referred to as one that followed him in point of time and as one that followed him in point of teaching as well, so that Ktakōṭi was one that followed the teaching of Pārāśarya Vyāsa coming later in point of time. It seems therefore understandable that, on the general question of the pramāṇas, Pārāśarya Vyāsa the teacher held ten such as acceptable, while the comparatively late disciple elected to accept only eight of his teacher’s, just as Jaimini, traditionally the pupil of Vyāsa, is said similarly to have accepted only six. The whole position is merely one of classification. While it is possible for one to regard some of these as distinct enough for separate treatment, others may legitimately hold that they are easily capable of inclusion in some of those already considered.

It seems likely there is manuscript warrant for the confusion in the printed texts of the commentary of Sābara on the Pūrva Mīmāṁsa. The Pāṇini translator in his comment on Sūtra I, I, 5 of the Pūrva Mīmāṁsa translates only six pramāṇas as those of the commentator. In his analysis of the first pāda of the work that he prefixes to the translation, he recounts eight pramāṇas without indicating where exactly he obtained the information from? A letter of mine asking for the source of this information regarding the eight pramāṇas, fetched the reply, to my great regret, that the learned translator died on 10th June, 1927. Is it not possible then that this list of eight is the eight of the Vttikāra as quoted by Sābara, and therefore the eight of [91] Ktakōṭi as we find it stated in the Maṇimēkhalai?

That there is some little confusion would become clear from the following extracts from Professor Keith’s Karma Mīmāṁsa. On page 8, he has:– ‘The extract from the Vttikāra proves that an important addition has been made to the teaching of the Mīmāṁsa in the shape of the introduction of the discussion of the validity of knowledge and its diverse forms.’ Further down he has ‘it is not illegitimate to assume that the Vttikāra indulged also in metaphysical discussions’. If the Vttikāra held eight pramāṇas as valid and is quoted as such in commentaries, and if the Maṇimēkhalai said Bōdhāyana Ktakōṭi formulated eight pramāṇas as such, is there not justification for regarding the two as the same person?

The point that the Sūtras actually discuss only four pramāṇas does not materially affect the question as these pramāṇas are pramāṇas of validity for the purpose of Mīmāṁsa Śāstra particularly, while the eight may be held as generally acceptable. The four considered in the Sūtras may be the four generally taken to be those accepted by the Naiyāyikas. That Vēdāntins discuss only the three pramāṇas of the Śāṁkhya would mean no more than that the preoccupation of the Vēdānta writers is to consider or controvert the Śāṁkhya, at any rate, to consider it as perhaps the most influential system obtaining at the time.

Coming to the point that the Vēdavyāsa referred to in the Maṇimēkhalai may be the Paurāṇika Vyāsa, I do not know if there is sufficient justification for distinguishing so many Vyāsas. But it is just possible that the Paurāṇikamata follows the teaching of Ktakōṭi, who, after all, is described as a follower of Vyāsa in point of conviction. In the work Sarva Siddhānta Saṁgraha ascribed to Śaṁkarācārya there is a Vyāsa Siddhānta discussed, but it seems to be a [92] Siddhānta incorporating the teaching of the Mahābhārata as such. This work, Sarva Siddhānta Saṁgraha may be the work of the great Śaṁkara or no. But it apparently was a work considered of some importance and standing as the Vaiṣṇava Ācārya Piḷḷai Lokācārya seems to have considered it as a work of some authority. In any case, there seems to be no room for a decisive negation to the statement in the Maṇimēkhalai that Vyāsa in his treatment of the Vaidika pramāṇas accepted ten pramāṇas as of general validity.

While, therefore, we may not be in a position to support by a decisive authority categorically stated of the position indicated in the Maṇimēkhalai, there seems to be quite enough of circumstantial evidence to support the general position of the Maṇimēkhalai that Vaidika pramāṇas as such received treatment at the hands of three ācāryas of high rank, Vyāsa, Ktakōṭi, and Jaimini, the pramāṇas actually upheld by the three being respectively, ten, eight and six. It further turns out that Ktakōṭi was a follower of Vyāsa in point of teaching, which may mean that he was a teacher of the system formulated by Vyāsa, and this involves his exposition of Vyāsa’s teaching, which took the form of a commentary, such as the Prapañcahdaya actually indicates it to be, the work Ktakōṭi of an author Bōdhāyana. Where Rāmānuja speaks, of his following the text of Bōdhāyana’s vtti in his Śrī Bhaṣya, the ground for identifying Bodhāyana with Ktakōṭi does not seem to be quite without support.

The next point of the professor’s criticism is the relative position of Dignāga and the author of the Maṇimēkhalai, as their teaching of Buddhistic logic is almost identical in point of form. Since the good professor has given me full permission to extract from his letters, I shall [93] take the liberty of expounding his position by quoting his own words:–

‘I now understand what you meant by saying that in the Maṇimēkhalai the logical teaching of Dignāga is anticipated. According to your opinion, there was in South India a school of Logicians headed by Aṟavaṇa Aḍigaḷ where Dignāga learned the system of Logic which later in North India he proclaimed as his own. This is a bold assumption which would require very strong arguments to pass as admissible. A prima facie objection, of which you seem not to be unconscious, is the following:– The Maṇimēkhalai is a romance the scheme of which is laid in the remote past. The events narrated in it are a fiction of the poet (or his predecessors), and so are the persons figuring in it. Why should Aṟavaṇa Aḍigaḷ be an exception? Is he, or his school, well attested by Tamil tradition in the Śangam Literature? In many Jaina romances there is introduced some Yati who gives an exposition of the law, converts the hero, etc., etc., but nobody has taken these teachers for historical persons. They serve the purpose of the poet to give a sketch of the Jain doctrines, or as the case may be to refute heterodox ones. Similarly Śāttaṉār introduced Aṟavaṇa Aḍigaḷ as the exponent of what he himself considered to be the essence of Buddhism. As he is no Śāstrin, no professor of Philosophy, but a poet and grain merchant, he naturally had to gather information from different sources. This accounts for the occasional inconsistencies in his ‘report’, e.g. ‘when first five Avayavas are taught (in accordance with Akṣapāda and Vasubandhu) and afterwards the two last Avayavas are included in dṛṣṭānta so that only three remain (as taught by Dignāga); or when in chapter XXVII the Bauddha system is reckoned among the six systems which are based on the six pramāṇas, and in chapter XXIX Buddha is said to have admitted but two pramāṇas.’

In the passage the first point that emerges is the character of the work Maṇimēkhalai. Maṇimēkhalai is undoubtedly a romantic poem. But I have taken pains to show that it is a romance based on historical occurrences. [94] It at least seems to have been the traditional opinion of the commentators that, while the treatment of the subject in the work is of the character of a romance, the incidents narrated in it are of the character of historical incidents. This is, to some extent, supported by the fact that the author, who is described undoubtedly as the grain merchant of Madura, was a contemporary of the Cēra king, ‘Śenguṭṭuvaṉ’, in whose court he was a much respected figure. He was the particular friend of the Cēra king’s younger brother, Iḷangō, the author of Śilappadhikāram, who more than once, in the body of the work, refers to Śāttaṉār as a friend of his brother, the monarch. Śenguṭṭuvaṉ is a character whose deeds are found described in works, whose character as Śangam works is beyond cavil. The achievements of this monarch are described in identical terms almost, in these two romantic works as well as in a more or less definitely historical poem, the Padirruppattu, by Paranar, a Śangam classic by common consent. Whatever opinion we may form of the works themselves, their character must be governed by this consideration that Śāttaṉār and Śenguṭṭuvaṉ were contemporaries.

The next point is, whether Aṟavaṇa Aḍigaḷ, the author, gives an exposition of Buddhism and Buddhistic logic in this work exactly in the style of the Jain authors referred to, the Jain celebrities described in the work Prabhāvakacarita. The Professor is undoubtedly right in saying that Śāttaṉār was first and foremost a poet and not perhaps a philosopher or Śāstrin. But if the tradition of the Tamil Land could be believed much true philosophic influence has been inspired to professed teachers and founders of systems from sources far less reputable than that of the grain merchant Śāttaṉār.

I have nowhere stated that there was a particular school of logicians, [95] or that Aṟavaṇa Aḍigaḷ was the head of a particular school. All that I meant was there must have been teachers of logic teaching at various centres in the Tamil country. Aṟavaṇa Aḍigaḷ was typical of those teachers and as was usual with these teachers he spent a wandering life here, there, and everywhere, at any rate, in the Tamil country. In the last stage of his life, he was teaching in Kāñcī, as he did in a previous stage at Kāvēripaṭṭiṇam. He may have anticipated the teaching of Dignāga in the enunciation of two pramāṇas, namely that the pramāṇas definitely applicable to the teachings of Buddhism are only two, and that the Avayavas (members of a syllogism) need be only three.

My position that Aṟavaṇa Aḍigaḷ anticipates Dignāga is taken on these two points that at the end of the discussion of Vaidika pramāṇas in Book XXVII, he definitely states that six are the current pramāṇas applicable to the six recognized systems of the time, and according to him one of the six recognized systems was Buddhism.

At the commencement of Book XXIX where he treats of Buddhist logic, he treats of the Pramāṇas, Pratyakṣa and Anumāna, and winds up with the statement that the other pramāṇas are capable of inclusion in the second, i.e., Anumāna. In the Nyāyapravēśa and the Pramāṇasamucchaya of Dignāga, Dignāga solemnly discusses the four pramāṇas of the logicians rejecting the last two and accepting the first two. To me it appears that Aṟavaṇa Aḍigaḷ’s position marks a transition to what ultimately became Dignāga’s teaching. If a man of the reputation of Aṟavaṇa Aḍigaḷ as a teacher taught the system in Kāñcī, one may take it that serious students at Kāñcī had some knowledge of it, and when one of them of the genius of Dignāga systematized the teaching of Buddhist logic, he might have improved upon it by [96] making a deliberate investigation of the whole position including that of the most authoritative school of the time, that of the Naiyāyikas, and laid it down that two are the pramāṇas and not four, while Aṟavaṇa Aḍigaḷ merely says that there are other pramāṇas, but they need not be considered separately, as all of them could be included in the second, the all of them here being apparently, the remaining four out of the six. It does not affect our position even if Aṟavaṇa Aḍigaḷ were a mere creation of the poet, as it would then be that Śāttaṉār incorporates in the Maṇimēkhalai what was the prevalent notion among Buddhists at Kāñcī, or what is perhaps better, in the Tamil country.

In regard to the Avayavas the same transition is indicated almost in the same manner, that is, he discusses the three and leaves the other two as capable of inclusion in the third. That is the reason why I regard it as a transition from the current beliefs of anterior times to the teachings of Dignāga.

The professor follows up his criticism with further remarking,

‘Though it will probably ever be impossible to ascertain all sources from which Śāttaṉār drew his information which lie embodied in theoretical chapters of the Maṇimēkhalai, but still it can be done in one case beyond the possibility of doubt. Indeed on a comparison of the exposition of the fallacies of the Maṇimēkhalai with the corresponding part in the Nyāyapravēśa it will be seen that the number and order of the fallacies, and as you yourself state, “their definitions and illustrations alike are almost identical”. Now the material agreement of two texts amounting practically to identity cannot be set down as a mere chance; it is impossible to expound it in any other way than by assuming that one text is immediately or mediately an abstract of the other; in the case under consideration, it is evident that the author of the Maṇimēkhalai, in this part at least, has borrowed from the Nyāyapravēśa of Śaṁkarasvāmin, [97] nor can it be assumed that both Śāttaṉār and Śaṁkara borrowed from a common source. For we know that Śaṁkara’s source was Dignāga’s Nyāyadvāra of which a Chinese translation has been preserved; but he gave a masterly exposition of his teacher’s logical system improving however on one point by adding four more to Dignāga’s five Pakṣa-ābhāsas. As the same four additional ābhāsas are adopted in Śāttaṉār’s abstract, it is clear that the latter has copied from the Nyāyapravēśa. In this regard the verdict of all unprejudiced scholars will be unanimous. Therefore the posteriority of Śāttaṉār to Śaṁkarasvāmin and a fortiori to Dignāga must be regarded as established. The upper limit of the position of the Maṇimēkhalai may be taken to be A.D. 500.’

In this part the professor takes up the position of Śāttaṉār’s being later than Dignāga on the basis of what is contained in Book XXIX of the Maṇimēkhalai. His position would be less open to objection, and mine perhaps less capable of justification, but for the fact that the position taken by him is not altogether without its own weak points.

We agree in respect of the teaching of Buddhist logic in Book XXIX of the Maṇimēkhalai and that of the Nyāyapravēśa and Pramāṇasamucchaya being almost identical. Our difference is only which is first and which is next. There are two points in the professor’s criticism which challenge consideration. The first is that the Nyāyapravēśa is ascribed to Śaṁkarasvāmin, the immediate disciple of Dignāga himself. There is the further point that the Nyāyapravēśa in regard to Pakṣa-ābhāsas improves upon the teaching of Dignāga by adding four more Pakṣa-ābhāsas thus bringing it into closer identity with the teaching of the Maṇimēkhalai. Therefore the position comes to be that the Maṇimēkhalai copied not the work of Dignāga himself but that of his disciple. But the Nyāyapravēśa [98] is regarded by others, on equally valid evidence of which the principal features have already explained by me above, as the work of Dignāga himself and not of Śaṁkarasvāmin. Therefore the addition of four Pakṣa-ābhāsas by Śaṁkarasvāmin to the teaching of Dignāga would have no basis to stand on.

The question then would be the teaching of Dignāga according to Nyāyapravēśa and the exposition of the Pakṣa-ābhāsas in Book XXIX of the Maṇimēkhalai. Which is anterior and which is posterior is the question. The main features of the argument upon which the particular pramāṇa is ascribed to Dignāga have been indicated above in summary, and the references given to where further information could be had. I requested the good offices of the learned professor to contribute the appended note on the other side of the question of Śaṁkarasvāmin’s authorship of Nyāyapravēśa as some of the sources are not accessible to me. In this context the professor’s position that the Nyāyapravēśa is based on Nyāyadvāra of Dignāga is seriously called in question. Both the Nyāyadvāra and the Nyāyapravēśa are ascribed to Dignāga, and are regarded as separate works altogether. Therefore the position is not quite so clear, and opinion is not so unanimous as to the authorship of these works.

To me it appears that it is a matter of no importance comparatively whether the Nyāyapravēśa is Dignāga’s or Śaṁkarasvāmin’s as Śaṁkarasvāmin is accepted as the immediate disciple of Dignāga. I agree with the learned professor that the two works are so close to each other as to be almost identical that the one must have taken it from the other. But the real point of difference is whether it is impossible that Aṟavaṇa Aḍigaḷ could have taught the subject without formulating it as in a text-book as Dignāga had done it later. [99] Aṟavaṇa Aḍigaḷ was a mere teacher, and not a codifier and controversialist like Dignāga, whose main purpose was to defend Buddhist logic as against those that may be interested in assailing it.

It is unnecessary for this position that Dignāga should necessarily have learned it from Aṟavaṇa Aḍigaḷ or somebody else in Kāñcī, or that he should take himself away to Northern India to misappropriate the teaching, as it were, and publish it as his own. From all that we know of teachers that taught in those times they had freedom to make the alterations implied in this process to be able to do it in their own particular localities. The only characteristic feature that ought to be paid attention to in the case of Dignāga is that he went about controverting, and had therefore to give his teaching a definition which a teacher like Aṟavaṇa Aḍigaḷ did not perhaps quite feel called upon to do from the necessities of controversy.

We cannot in our present state of knowledge be so definite about the position of Akṣapāda and Vasubandhu that these teachers originated the teaching in regard to the five Avayavas, or in regard to anything else. It would be very difficult to ascribe the originating of any very particular item of teaching in these departments to particular authors except to the extent of their having committed these items of teaching to writing in works that have become accessible to us. As a rule it may be taken that these teachings were for a considerable length of time in the floating traditions of the schools before they got entry into written texts, and when they actually reached this state they had necessarily to take a more definite form. Therefore it is quite unnecessary to ascribe any moral turpitude to Dignāga in doing what he actually did, giving to the teaching of the schools before him, of which Aṟavaṇa Aḍigaḷ is a mere [100] representative, the current definition in this formulation of the pramāṇas and the Avayavas.

On the basis of the reasons given by the learned professor, the date A.D. 500 may seem quite reasonable. But Vasubandhu’s date is nowadays taken to be somewhat earlier, and cannot go as far as A.D. 400. It is now taken as proved that Vasubandhu was a contemporary of Samudragupta, and Dignāga therefore could not have been very much later. He might have been somewhat younger than Vasubandhu but could not have been so late as to be capable of being brought to A.D. 400. That from our point of view seems a minor one.

While we are discussing this point, I may as well note here a remark of my friend, Mr. R. Narasimhacharyar, till lately Director of Archaeological Researches in Mysore. He is as well convinced as Professor Jacobi himself as to the Maṇimēkhalai expounding the Nyāyapravēśa of Dignāga. But he put it to me that, having regard to what Book XXIX of the Maṇimēkhalai expounds as Buddhist logic, it would not be unreasonable to argue that Dignāga was an earlier teacher and therefore earlier than the date of the Maṇimēkhalai, that is, the second century A.D. In other words, he would take the date of Maṇimēkhalai to be the second century A.D. and would place Dignāga earlier than that date.

This position of my friend goes to indicate that perhaps the more legitimate historical argument would be to fix the date of the Maṇimēkhalai and shift that of Dignāga, for after all, Maṇimēkhalai’s dating is to be primarily on the historical considerations indicated above rather than on the philosophical systems such as they are in the work. I may, however, note that there are difficulties in the way of accepting that position. Dignāga’s contemporaneity with Vasubandhu would be difficult to call in [101] question unless we are prepared to throw to the four winds all the available evidence of literary tradition completely. Vasubandhu cannot be taken to an anterior date such as this would imply without doing very great violence to accredited Buddhist tradition and Chinese evidence of a definite character. The more reasonable position to take therefore seems to be that which is taken in the course of the exposition of the work above.

The professor’s further point of criticism is in regard to the omission of all references to Mahāyāna. Here is the professor’s position:–

‘Śāttaṉār in his exposition of Buddhism nowhere, as you say, refers to Mahāyānistic ideas. It may, therefore, be assumed that in his time the Mahāyāna was not yet in existence, and accordingly Śāttaṉār must be earlier than Nāgārjuna; but this conclusion can easily be shown to be wrong. For Śāttaṉār refers to Akṣapāda and Nyāya, and as in the Nyāyasūtra, the Śūnyavāda is discussed and refuted, there can be no doubt that in Śāttaṉār’s time, the Mahāyāna was already established long since.’

In this point again, I am sorry that the professor’s argument overshoots the mark. There is no reference to Mahāyāna in the exposition of Buddhism in the Maṇimēkhalai, and the conclusion cannot be that Śāttaṉār did not know the Mahāyāna either to accept it as an orthodox system, or to condemn it as a heretical system. That may be due to Śāttaṉār being anterior to Nāgārjuna and therefore of Śūnyavāda, or of his not knowing it, the teaching not having had sufficient time to have become well known, and reach the Tamil country. The latter is the view that I have taken, and not exactly the former, guided here again by the governing historical considerations. The learned professor, on the contrary, bases himself on the position that the teaching of Akṣapāda [102] and the Nyāyasūtra are identical in every particular.

In regard to the Śūnyavāda that is discussed and refuted there again, there is not that agreement in regard to what exactly the teachings of Akṣapāda were and what additions were made to the Nyāyasūtras since the time of Akṣapāda. This would mean that the discussion of the Śūnyavāda must be proved to be in part ascribable to Akṣapāda, as in fact it cannot be, if according to other scholars who have specialized their studies in the Nyāya, Akṣapāda was far anterior to Nāgārjuna himself.

All I wish to point out is that an argument such as this cannot be held to be decisive in our present state of knowledge of the chronology of these works. The Mahāyāna is not a product of Nāgārjuna’s teaching. The teaching of Mahāyāna can be traced back to the days of Aśoka, if not earlier. But the actual Śūnyavāda in the form in which it has come down to us is still generally regarded as the teaching of Nāgārjuna.

The possibilities are that Śāttaṉār’s teaching embodies whatever was in the opinion of the Tamil country, the orthodox teaching of Buddhism, about the same time as Nāgārjuna was expounding the Śūnyavāda of the Mahāyānistic school in the Āndhra country across, both being the result of the same stir, particularly in the continent of India, that is indicated in the Mahāvaṁsa of Ceylon as the famous Vaitulya controversy. The heretics are located, according to the Mahāvaṁsa, in the coast country set over against Anuradhapura extending northwards into the Āndhra country.

Therefore the time at which Śāttaṉār lived seems to me the time which actually produced Nāgārjuna and Deva and possibly a little anterior. Much as the great Master of the Law, Hiuen-Tsang [Xuan Zang], does not make any reference to Dharmakirti who lived in his time and perhaps was actually teaching when [103] the great traveller was in India, Śāttaṉār fails to mention Nāgārjuna or his Śūnyavāda.

The analogy brought in by the learned professor can hardly be accepted as holding good in this case. His position is:–

‘Similarly in the Vēdānta Sūtra and in the passage from the Vttikāra quoted by Śabarasvāmin ad. I, 1,5 the Śūnyavāda is discussed and refuted. It is true that in the Maṇimēkhalai there is no explicit reference to the Vēdānta philosophy. However the same remark applies also to the Nyāyavārttika, for Uddyotakāra altogether ignores the Vēdānta though at his time it was almost certainly a separate system of philosophy. The same attitude towards Vēdāntism taken up by Uddyotakāra and Śāttaṉār rather speaks in favour of the assumption that both authors were not far removed in time from each other.’

Uddyotakāra is a commentator pure and simple on the Nyāya. It is open to him not to mention the Vēdānta as a system, unless he saw particular reason for doing so, or the actual text that he commented on necessitated a reference. Śāttaṉār stands on a different footing. He lays himself out to discuss what he regarded as heretical systems and then to expound the system of Buddhism that commended itself to him. The difference is vital and of considerable force. It would be therefore difficult to believe that Uddyotakāra and Śāttaṉār were near enough in point of time because of the omission in the works of both of them of any reference to Vēdānta as a system.

Coming to the exposition of Buddhism in Book XXX of the Maṇimēkhalai, the professor’s criticism is as follows:–

‘The translation of chapter XXX of the Maṇimēkhalai is very welcome, though it is rather disappointing being a mere meagre account of Buddhism. I wonder who Śāttaṉār’s authority was in this part. It contains only such teachings [104] as may be acknowledged by all Buddhists, both Hīnayānists and Mahāyānists. It has not any reference to the Sarvāstivāda nor the system of the Sautrāntikas, for the central conception of these two schools is the theory of the dharmas which is not even hinted at in Śāttaṉār’s abstract. I think he merely related what every Śrāvaka was supposed to know.’

It may be stated at once that Śāttaṉār although he does not indicate the authority upon which he relies for the summary of Buddhism in Book XXX refers elsewhere to what his authority is. The Buddhism that he teaches is, ‘the path of the Piṭakas, of the Great one.’ Book XXVI, 1, 66. He is expounding the fundamental teachings of the Buddha and not the teachings of schools of Buddhism which are elaborations and modifications of systems-builders of later times. It is possible to make the inference from this alone that Śāttaṉār was anterior to the growth of definite systems that we know of in Buddhism, particularly the four which are so prominently associated with the Buddhism of a later age. But that argument need not be pushed to any extreme.

A German scholar of the twentieth century, laying himself out definitely to disentangle the teachings of the Buddha from the excrescences of subsequent ages and teachers, inculcates in substance what is the teaching of Śāttaṉār, neither more nor less. The Doctrine of the Buddha by George Grimm, Leipzig, 1926. So Śāttaṉār’s teaching may be regarded as the teaching of ‘the Piṭaka of the Buddha’, and therefore indicates a deference to the authority of the ‘word of the Buddha’ such as it was known to be in his time. In that sense it would be what is called Sthavirāvada and may be regarded as Sautrāntika also, not in the technical sense that the expression acquired, but in a more general sense.

Śāttaṉār’s anxiety is to teach what the Buddha taught. It [105] is just possible on this very ground to claim for him anteriority, though it is equally possible that a later writer could lay himself out to disentangle the actual teaching of the Buddha from its outgrowths. But the claim to Śāttaṉār’s anteriority, according to me, rests not so much on this feature as on the particular feature that the other systems as such did not come in for commendation in the book on Buddhism, and what perhaps is more to be expected, in condemnation along with the heretical systems. One explanation is possible that, while he condemns systems which did not recognize the Buddha, he merely expounds a system taught by the Buddha, and passed over outgrowths from that system with a tolerance which is not unusual in Indian thought.

I have taken it upon myself to make this elaborate criticism of the views of my much esteemed and learned friend, Professor Jacobi, because the importance of the subject and the eminence of the scholarship of the professor alike demand it from me. The stimulus to this line of investigation at this time, came from him to me, and it is but fair to him that I should acknowledge it here, and consider his criticism with the respect which is due to the eminent source from which it comes.

The elaborate criticism and the extensive answer that that necessitated, alike go to show that the line of investigation that was undertaken has shown clearly, though somewhat disappointingly, that this line of investigation cannot by any means lay claim to that finality which, perhaps in the first instance, was expected of it both by Professor Jacobi and by myself. If I could go by this investigation alone I should not have any great difficulty in accepting the position arrived at by the eminent scholar. But once that position is accepted, it is incumbent upon me, as a student of history, to test the [106] position by other lines of enquiry.

Without repeating the details of history, I may merely draw attention here to two facts which stand out. The first is that the author is demonstrably a contemporary of Śenguṭṭuvaṉ Cēra, and of his younger brother, Iḷangō, the author of the Śilappadhikāram. That is one fact of history which it would be difficult to call in question. The second is that at the time to which the work refers which is undoubtedly the time of the author, Kāñcī was not under the Pallavas, nor under the Toṇḍamāṉ chieftain, Iḷaṁ Tiraiyaṉ, but under the princely viceroys of the Cōḻa family.

The history of the Pallavas as such certainly goes back to the age of Samudragupta in the middle of the fourth century as we take it at present. That is a date not far removed from that of Dignāga. The Toṇḍamāṉ chieftains, particularly Toṇḍamāṉ Iḷaṁ Tiraiyaṉ of Śangam fame, must have ruled earlier. The Cōḻa ascendency in the Toṇḍamāṉmaṇḍalam (the country round Kāñcī) must be referred to an age anterior to this. This position had been sought to be got round by the Epigraphists by the assumption of a Cōḻa interregnum previous to the Pallava king Kumāraviṣṇu II who in one of his records is said to have retaken Kāñcī. Leaving aside for the moment criticism of details in connection with this particular statement, it may be said at once that the Cōḻa interregnum such as is postulated must be an interregnum extending over five generations or roundly one century of time.

The Śangam works give evidence of Kāñcī being under the rule of the Cōḻas and then of the Toṇḍamāṉ Iḷaṁ Tiraiyaṉ, and arranging the authors and the patrons referred to therein in the order of succession merely, we get to somewhat like this line of five generations and a length of time of a century. Therefore no conclusion can be accepted which does not satisfy this [107] condition primarily; the age indicated by Professor Jacobi on the line of reasoning that he has adopted, with the philosophical systems of the Maṇimēkhalai as a basis, can hardly satisfy this condition. The alternative suggestion of Mr. Narasimhacharyar has been briefly adverted to.

While therefore acknowledging with gratitude the criticism of the learned professor, I may join in his regret that we cannot come to an agreement on this investigation. [108]