However far the Kārikās of Īsvara-Kṛṣṇa may go back, they are what they are, a metrical work in the style of a later age, an age that gave rise to other Kārikās like Bhartṛhari’s (about 650 A.D.) Kārikās on grammar. Everybody has wondered, therefore, what could have become of the real Sāṃkhya-Sūtras, if they ever existed; or, if they did not, why there should never have been such Sūtras for so important a system of philosophy as the Sāṃkhya. There is clearly a great gap between the end of the Upanishad period and the literary period that was able to give rise to the metrical work of Īsvara-Kṛṣṇa. In what form could the Sāṃkhya-philosophy have existed in that interval?

To judge from analogy we should certainly say, in the form of Sūtras, such as were handed down for other branches of learning by oral tradition. The Kārikās themselves presuppose such a tradition quite as much as the much later Sūtras which we possess. They are both meant to recapitulate what existed, never to originate what we should call new and original thoughts. When we see the Kārikās declare that they leave out on purpose the Ākhāyikās, the illustrative stories contained in the fourth book of our Sūtras, this cannot prove their posteriority to the Sūtras as we have them; but it shows that at Īsvara-Kṛṣṇa’s time there existed a body of Sāṃkhya-philosophy which contained such stories as we find in our modern Sūtras, but neither in the Kārikās nor in the Tattva-samāsa. Besides these stories other things also were omitted by Īsvara-Kṛṣṇa, comprehended under the name of Paravāda, probably controversies, such as those on the necessity of an Īsvara.

Under these circumstances I venture to say that such a work in Sūtras not only existed, but that we are in actual possession of it, namely in the text of the much neglected Tattva-samāsa. Because it contains a number of new technical terms, it has been put down at once as modern, as if what is new to us must be new chronologically also. We know far too little of the history of the Sāṃkhya to justify so confident a conclusion. Colebrooke1 told us long ago that, if the scholiast of Kapila2 may be trusted, and why should he not? the Tattva-samāsa was the proper text-book of the Sāṃkhya-philosophy. It was a mere accident that he, Colebrooke, could not find a copy of it. “Whether that Tattva-samāsa of Kapila be extant,” he wrote, “or whether the Sūtras of Pañcasikha be so, is not certain.” And again he wrote: “It appears from the Preface of the Kapila-bhāshya that a more compendious tract in the form of Sūtras or aphorisms, bears the title of Tattva-samāsa, and is ascribed to the same author,” i.e. to Kapila.

I admit that the introductory portion of this tract sounds modern, and probably is so, but I find no other marks of a modern date in the body of the work. On the contrary there are several indications in it of its being an earlier form of the Sāṃkhya-philosophy than what we possess in the Kārikās or in the Sūtras. When it agrees with the Kārikās, sometimes almost verbatim, it is the metrical text that seems to me to presuppose the prose, not the prose the metrical version. In the Sūtras themselves we find no allusion as yet to the atheistic or non-theistic doctrines which distinguish the later texts of the Sāṃkhya, and which are still absent from the Samkhya-Kārikās also. The so-called Aisvaryas or superhuman powers, which are recognised in the Tattva-samāsa, might seem to presuppose the recognition of an Īṣvara, though this is very doubtful; but the direct identification of Purusha with Brahman in the Tattva-samāsa points certainly to an earlier and less pronounced Nirīṣvara or Lord-less character of the ancient Sāṃkhya. It should also be mentioned that Vijñana-Bhikshu, no mean authority on such matters, and even supposed by some to have been himself the author of our modern Sāṃkhya-Sūtras, takes it for granted that the Tattva-samāsa was certainly prior to the Kapila-Sūtras which we possess. For why should he defend Kapila, and not the author of the Tattva-samāsa, against the charge of Punarukti or giving us a mere useless repetition, and why should he have found no excuse for the existence of the Kapila-Sūtras except that they are short and complete, while the Tattva-samāsa is short and compact?3

Not being able to find a MS. of the Tattva-samāsa Colebrooke decided to translate instead the Sāṃkhya-Kārikās, and thus it came to pass that most scholars have been under the impression that in India also this metrical version was considered as the most authoritative and most popular manual of the Sāṃkhya-philosophy. This is the way in which certain prepossessions arise. We have learnt since from Ballantyne4 that at Benares, where he resided, these Kārikās were hardly known at all except to those who had seen Professor Wilson’s English edition of them, while the Tattva-samāsa was well known to all the native assistants whom he employed. Nor can we doubt that in the part of India best known to Ballantyne it was really an important and popular work, if we consider the number of commentaries written on it,5 and the frequency of allusions to it which occur in other commentaries. The commentary published by Baltantyne is, if I understand him rightly, anonymous. It gives first what it calls the Sāṃkhya-Sūtrāṇi, and then the Samāsākhya-sūtra-vṛttiḥ. Hall, l. c., p. 13, quotes one commentary by Kshemānanda, called Sāṃkhya-kramadīpikā, but it is not quite clear to me whether this is the same as the one published by Ballantyne, nor have I had access to any other MSS.

We must not forget that in modern times the Sāṃkhya-philosophy has ceased to be popular in several parts of India. Even in the sixteenth century Vijñana-Bhikshu, in his commentary on the Sāṃkhya-Sūtras (v. 5), complains that it has been swallowed up by the sun of the time, and that but a small part of the moon of knowledge remained; while in the Bhāgavata Purāṇa I, 3, 10, the Sāṃkhya is spoken of as Kāla-vipluta, destroyed by time. Professor Wilson told me that, during the whole of his intercourse with learned natives, he met with one Brāhman only who professed to be acquainted with the writings of this philosophical school, and Professor Bhandarkar (l. c., p. 3) states that the very name of Sāṃkhya-pravacana was unknown on his side of India. Hence we may well understand that Sāṃkhya MSS. are scarce in India, and entirely absent in certain localities. It is possible also that the very smallness of the Tattva-samāsa may have lowered it in the eyes of native scholars, and that in time it may have been eclipsed by its more voluminous commentaries. But if we accept it as what it professes to be, and what, up to the time of Vijñana-Bhikshu at least, it was considered to be in India, it seems to me just the book that was wanted to fill the gap to which I referred before. By itself it would fill a few pages only. In fact it is a mere enumeration of topics, and, as such, it would agree very well with the somewhat puzzling name of Sāṃkhya, which means no more than enumeration. All other derivations of this title seem far-fetched6 as compared with this. According to Vijñana-Bhikshu in his commentary on the Sūtras (pp. 6, 110, ed. Hall), both the Sāṃkhya-Sūtras and the Yoga-Sūtras are really mere developments of the Tattva-samāsa-Sūtras. Both are called therefore Sāṃkhya-pravacana, exposition of the Sāṃkhya, the latter adding the peculiar arguments in support of the existence of an Īṣvara or Supreme Lord, and therefore called Seṣvara, in opposition to the Sāṃkhya, which is called An-īsvara, or Lord-less.

And here it is important to remark also that the name of Ṣaṣṭi-tantra, the Doctrine of the Sixty, which is given by Īsvara-Kṛṣṇa, or at all events by the author of the 72nd of his Kārikās, should occur and be accounted for in the Tattva-samāsa, as containing the 17 (enumerated in 64 and 65), and the 33, previously exhibited in 62 and 63, together with the 10 Mūlikārthas or fundamental facts which together would make up the sixty topics of the Ṣaṣṭi-tantra. At the end of the 25 great topics of the Tattva-samāsa we find the straightforward declaration:

Iti tattva-samāsākhya-sāṃkhya-sūtrāṇi

Here end the Sāṃkhya-Sūtras called Tattva-samāsa.

At first sight, no doubt, Samāsa seems to mean a mere abstract; but Samāsa may be used also in opposition to Bṛhat, and there is no other work in existence of which it could be called an abstract, certainly not either of the Kārikās or of the modern Sūtras, such as we possess them.

The whole arrangement is different from the other and more recent treatments of Sāṃkhya-philosophy. The three kinds of pain, for instance, which generally form the starting-point of the whole system, are relegated to the very end as a separate topic. We meet with technical subjects and technical terms which are not to be found at all in other and, as it would seem, more modern Sāṃkhya works. The smallness of the Tattva-samāsa can hardly be used as an argument against its ever having been an important work, for we find similar short, yet old Sūtra works, for instance, the Sarvānukrama and other Anukramaṇis described in my History of Ancient Sanskrit Literature.7 However, in matters of this kind we must avoid being too positive either in denying or asserting the age and authenticity of Sanskrit texts. All I can say is that there is ho mark of modern age in their language, though the commentary is, no doubt, of a later date. What weighs with me is the fact that Indian Pandits evidently considered the Tattva-samāsa-Sūtras as the original outlines of the Sāṃkhya-philosophy, while the idea that they are a later spurious production rests, as far as I can see at present, on no real argument whatever.

If then I venture to call the Tattva-samāsa the oldest record that has reached us of the Sāṃkhya-philosophy, and if I prefer to follow them in the account I give of that philosophy, I am quite aware that many scholars will object, and will prefer the description of the Sāṃkhya as given in the Kārikās and in the Sūtras. Both of them, particularly the Kārikās, give us certainly better arranged accounts of that philosophy, as may be seen in the excellent editions and translations which we owe to Professor Garbe, and I may now add to Satish Chandra Banerji, 1898. If, as I believe, the Tattva-samāsa-Sūtras are older than our Sāṃkhya-Sūtras, their account of the Sāṃkhya-philosophy would always possess its peculiar interest from a historical point of view; while even if their priority with regard to the Kārikās and Sūtras be doubted, they would always retain their value as showing us in how great a variety the systems of philosophy really existed in so large a country as India.

These Samāsa- Sūtras, it is true, are hardly more than a table of contents, a mere Sāṃkhyam or Pari-saṃkhyā, but that would only show once more that they presuppose the existence of a commentary from the very first. What we possess in the shape of commentaries may not be very old, for commentaries may come and go in different schools, while the Sūtras which they intend to explain, would remain unchanged, engraved on the memory of teachers and pupils. How tenacious that philosophical Paramparā was we can see from the pregnant fact that the Ākhyāyikās or stories, though left out in the Kārikās, must surely have existed both before and after the time of Īsvara-Kṛṣṇa, for though absent in the Tattva-samāsa and in the Kārikās, they reappear in our Sāṃkhya-Sūtras. Where were they during the interval if not in Sūtras or Kārikās, now lost to us?

The commentary on the Tattva-samāsa, the publication of which we owe to Ballantyne, begins with an introduction which sounds, no doubt, like a late tradition, but reminds us in some respects of the dialogue at the beginning of the Chinese translation of the commentary on the Sāṃkhya-kārikās. But though it may sound like a late tradition, it would be very difficult to prove that it was so. Chronology is not a matter of taste that can be settled by mere impressions.

A certain Brāhman, we are told, overcome by the three kinds of pain, took refuge with the great Ṛṣi Kapila, the teacher (not necessarily the originator) of the Sāṃkhya,8 and having declared his family, his name, and his clan in order to become his pupil, he said:

“Reverend Sir. What is here on earth the highest (the summum bonum)? What is truth? What must I do to be saved?”

Kapila said, “I shall tell thee.” Then follows the topics which are twenty-five in number:

Here follows the Tattva-samāsa-Sūtras

1. Essays, I, p. 244

2. Sāṃkhya-pravacana-bhāshya, pp. 7, 110.

3. Sāṃkhya-pravacana-bhāshya, Introduction.

4. Drift of the Sāṃkhya, p. 1.

5. Five are mentioned by Hall in his Preface, p. 33.

6. They are mentioned in the Preface to Hall’s edition of the Sāṃkhya-pravacana-bhāshya, 1856. Some of them are mere definitions without any attempt at etymology.

7. M.M., India, p. 362.

8. In the Bhāgavata-purāna I, 3, 11, Kapila is said to have retired the Sāṃkhy (Sāṃkhya-Sāra, ed. Hall, p. 7, note).