Much has been said on the relation of `Sa^mkaraacaarya to the Buddhists, and the views which are current on this topic differ as much as black differs from white. The more enthusiastic of `Sa^mkara's followers claim that he is chiefly responsible for driving the Buddhists out of India. Their sectarian opponents, on the other hand, have claimed that far from opposing Buddhism `Sa^mkara secretly accepted its doctrines and introduced as many of them as he could into the Vedaanta tradition. Scholars outside of India have also been far from agreement in their opinions, for some have emphasized the practical element of `Sa^mkara's doctrine, which is certainly opposed to Buddhism, while others, by emphasizing the idealistic and acosmic elements, have reduced the differences to a minimum. In Japan, where the Buddhist teachers of the past are held in great respect, scholars have followed still a different path, arguing that `Sa^mkara failed to understand Buddhism, whatever his attitude to it may have been.
In this conflict of pinions, no one, to the best of my knowledge, has availed himself of two methods of research which I think may minimize prejudice and lead to a more rational and more widely acceptable solution. These two methods are the comparison of the commentaries of `Sa^mkaraacaarya and Bhaaskara on the Brahma-suutra and the comparison of the arguments `Sa^mkara uses against the Buddhists in his Brahma-suutra-bhaa.sya with those he uses in commenting on the B.rhadaara.nyaka Upani.sad. I do not say that these methods will solve all problems, but they will help, and for the following reason.
It becomes very clear from a comparison of `Sa^mkara and Bhaaskara that the major part of `Sa^mkara's commentary on the Brahma-suutra is not original with `Sa^mkara but is repeated from what commentators had written in the past. What we have in the Brahma-suutra-bhaa.sya is the accumulated philosophy of a millennium. It is true there are original elements, some of them very important, and, as I hope to show, these elements can sometimes be
1. The substance of this paper was delivered in the form of a lecture before the Bombay Branch of the Royal Asiatic Society, November 25, 1952.
recognized as original, but it is tradition that forms the mainstay. Now, to determine `Sa^mkara's exact attitude toward Buddhism, it is essential that we attempt to disentangle what is original with `Sa^mkara from what is not. Let me point out briefly the means at oar disposal.
`Sa^mkara's is the most ancient commentary been preserved on the Brahma-suutra, that collection of laconic utterances which summarize the philosophy of the Upani.sads. `Sa^mkara wrote this commentary probably early in the eighth century A.D. The next most ancient to survive that of Bhaaskara. This has been published in only one edition, based on few and corrupt manuscripts, and badly edited.[2a] Despite this, it is usually possible to make out Bhaaskara's meaning with the help of `Sa^mkara and other texts. This commentary of Bhaaskara is referred by Vaacaspati Mi`sra in about A.D. 850. Bhaaskara, in torn, refers to arguments which seem to have been raised first by `Sa^mkara's pupils, Sure`svara and Padmapaada. He should, therefore, be at least two generations younger than `Sa^mkara. If we place the date of his commentary at A.D. 800 we shall not be far wrong.
These are the two most ancient commentaries that have been preserved. However, we know that many commentaries had been composed before this time. Bhaaskara is particularly valuable in furnishing evidence of this. In the course of his comment he refers frequently to the opinion of earlier
2. The dates A.D. 788-820, which have been widely accepted for `Sa^mkara in the past, must be pushed back. A detailed and scholarly treatment of the subject will be found in the first volume of Hajime Nakamura's Japanese work Shooki no Vedaanta Tetsugaku (Tokyo: Iwanami Shoten, 1950), pp 63-121. The early limit for `Sa^mkara is the date of Dharmakiirti, whom `Sa^mkara quotes in the Upade`sa-saahasrii, K.r.s.na `Saastrii Navare, ed. (Bombay: Jagadishvara Press, 1886), XVIII.142. Dharmakiirti rose to fame between the visits to India of Hsuan Tsang and I Ching, that is, between A.D. 634 and 673. The later limit is given by two sets of facts. (A) `Sa^mkara's pupil Sure`svara is quoted by Vidyaananda, who must have lived slightly before A.D. 800. See Pathak in "Bhart.rhari and Kumaarila," Journal of the Bombay Branch of the Royal Asiatic Society, XVIII (1894), 225-229. (B) There must be at least two generations between `Sa^mkara and Vaacaspati Mi`sra, who wrote the Nyaaya-suucii-nibandha in A.D. 841. The generations, on Nakamura's showing (op. cit., p. 89 and p. 98, note 12), are: `Sa^mkara, `Sriivatsaa^nka. Bhaaskara, Vaacaspati. Nakamura bases this at least in part on Yamuna's Siddhi-traya. Chowkhamba Snaskrit Series Work No. 10(Benares: Chowkhamba Sanskrit Book Depot, 1900), p. 6. One can prove the same result perhaps more surely by taking the following sequence: `Sa^mkara, Padmapaada, and Sure`svara, Bhaaskara, Vaacaspati. For evidence that Bhaaskara is later than `Sa^mkara's pupils Padmapaada and Sure`svara, see note 4 below.
2a. Brahmasuutra with a Commentary by Bhaaskaraachaarya. Chowkhamba Sanskrit Series Work No. 20 (Banares: Chowkhamba Sanskrit Book Depot, 1903-1915).
3. See, for example, Bhaamatii and Kalpataru on II.iii.9; III.iii.29; III.iv.27. See The Brahma-suutra `Saa^nkara Bhaa.sya with the Commentaries Bhaamatii, Kalpatara and Parimala (Bombay: Nirnaya Sagar Press, 1938). Vaacaspati's references to Bhaaskara were first brought to notice by M. Hiriyanna in the introduction to his edition of the I.s.tasiddhi. Gaekwad Oriental Series No. 65 (Baroda: Oriental Institute, 1933). A more complete listing of the references will be found in Nakamura, op. cit., pp. 91-97.
4. Thus, Bhaaskara fives the following argument (op. cit., p. 19, lines 7 ff.): "If you say that ignorance (avidyaa) is beginningless, then there can be no release, for it could be no more destroyed then brahma. Or, if it has beginning, whence does it arise? And if it does arise, since it is than an effect, it must be a real thing; whereby your thesis is destroyed that ignorance is unreal (avastu), for no unreal thing like a hare's horn arises. Again, ignorance cannot belong to brahma, for the nature of brahma is pure consciousness and incomparable bliss; nor can it belong to the individual soul, for you admit the existence of no individual soul different from brahma." Similar arguments are again put forward by Bhaaskara (op. cit., p. 95 lines 26 ff.).
Now, `Sa^mkara, to the best of my knowledge, nowhere says that ignorance is unreal. See my article "`Sa^mkara on the Question: Whose is Avidyaa?" Philosophy East and West, III, No. 1 (April, 1995), 69-72. It is Padmapaada who says this. See Hacker, Untersuchungen uber das fruhere Vedanta (Mainz: Akademie der Wissenschaften und der Literatur, Abhandlungen der Geistes- und Sozialwissenschaftlichen Klasse, 1950), pp. 2017-2018. Again, `Sa^mkara never gives any locus of ignorance (avidyaa). Sure`svara, however, held the locus to be the individual soul (Hacker, ibid., p. 1973) and Padmapaada held the locus to be brahmaatma (ibid., pp. 2026-2027).
I formerly believed that tradition might be correct in making Bhaaskara a contemporary of `S^mkara. But the above facts make it fairly certain that he was at least two generations later.
teachers. Unfortunately, he never mentions their names. He introduces their remarks by "ke cit tu" (but some people_, "apara aahu.h" (other say), and similar expressions. It is possible, however, by a painstaking comparison of these remarks, to join certain ones together and to make out some idea of the major trends in Brahma-suutra interpretation that preceded both Bhaaskara and `Sa^mkara.
Historically, perhaps the most important of these previous interpreters is a philosopher referred to by Bhaaskara and by `Sa^mkara's followers as the v.rttikaara, that is, the commentator. To distinguish him from others I shall call him the "Protocommentator." Now, a most interesting fact appears from a comparison of `Sa^mkara and Bhaaskara. Whenever we are told by the supercommentators that `Sa^mkara is departing from the views of the Protocommentator, we find Bhaaskara upholding these very views and passionately objecting to `Sa^mkara's departure. I think it is quite clear that Bhaaskara's commentary is based in large part on this pre-`Sa^mkaran v.rtti (commentary).
A second interesting fact is that a very large part of Bhaaskara's comment is substantially the same as `Sa^mkara's. Yet, we cannot suppose that Bhaaskara often copied from `Sa^mkara, for `Sa^mkara was his archenemy. In the first two books of Bhaaskara's commentary, out of forty-five references to earlier interpreters, seventeen are to `Sa^mkara and many of these are vitriolic. Referring to specific doctrines of `Sa^mkara, he says, "No one but a drunkard could hold such theories" (I.i.4; p. 20, line 23). With reference to the whole
5. For example, on Brahma-suutra I.i.19 Govardhana and Aanandagiri attribute `Sa^mkara's first exegesis of the aanandamayaatma (bliss-self) to the Protocommentator. The Brahma-suutra Shaa^nkarabhaa.syam with the Commentaries Ratnaprabhaa, Bhaamatii and Nyaayanir.naya (Bombay: Nirnaya Sagar Press, 1904), p. 125. This is the interpretation followed by Bhaaskara, who finds fault with `Sa^mkara's second and preferred interpretation, saying, "The proper way to interpret this suutra is the traditionally handed down way" (Bhaaskara, op. cit., p. 29, lines6-7).
On Brahma-suutra I.i.23, `Sa^mkara, according to the supercommentators, is impugning the Protocommentator when he says, "Some adduce under this suutra 'praa.nasya praa.nam etc.,' but this is wrong." Bhaaskara, however, clearly accepts the adduction.
Compare also `Sa^mkara and Bhaaskara on Brahma-suutra I.i.25; I.ii.23; I.iv.26.9
school of `Sa^mkara, "they destroy the meaning of the suutras and lead students into error" (III.ii.3). Again, Bhaaskara speaks of "the maayaavaadins hanging on to Buddhist doctrine" (II.ii.29), and of "this despicable, broken-down maayaavaada that has been chanted by Mahaayaana Buddhists" (I.iv.25).
I do not think Bhaaskara often borrows directly from `Sa^mkara. We are, rather, to explain the similarity by supposing that `Sa^mkara, as well as Bhaaskara, follows in the main the old Protocommentator. He may depart from him on crucial points, but the Protocommentator furnishes the framework of `Sa^mkara's Brahma-suutra-bhaa.sya also.
Now, if this is so, we have at last an instrument for determining what is original and what is traditional in `Sa^mkara's philosophy. Where `Sa^mkara and Bhaaskara agree, the text is traditional and goes back to the Protocommentator. Where Bhaaskara criticizes `Sa^mkara, the probability is that the doctrine does not go back to the Protocommentator. It may, of course, go back to some other Vedaanta tradition. Of that we cannot be certain without further evidence, such as is sometimes afforded by Gau.dapaada or the Vaakyapadiiya. Nor can we reach any conclusion where `Sa^mkara and Bhaaskara disagree without Bhaaskara's actually criticizing `Sa^mkara. The instrument not perfect. It is, however, a great help.
There is, as I have indicated, one further instrument may be used: a comparison of `Sa^mkara's doctrines as given in his Brahma-suutra-bhaa.sya with what he gives us elsewhere. Especially valuable in this regard is commentary on the B.rhadaara.nyaka Upani.sad, for we know definitely that this commentary is by `Sa^mkara, it is the longest of all Upani.sad commentaries attributed to him, and it includes the most philosophical discussion. Such a comparison leads to interesting results. First, and this is Kathe Marschner's discovery, there are scarcely any outright contradictions between the two works. Second, and I believe this has not been pointed out, emphasis of the two commentaries is very different. For example, some `Sa^mkara's greatest arguments in the B.rhadaara.nyaka are directed against the bhedaabheda (doctrine of identity in difference), and these arguments found throughout that work. Similar arguments are rare in his Brahma-suutra-bhaa.sya and with one exception are limited to a few lines. Or, to come
6. This is noticed by Hiriyanna, "Bhart.rprapa~nca: An Old Vedaantin," Indian Antiquary, LIII (1923), 84, where it is also noticed how `Sure`svara tries to effect a compromise of bhedaabheda with his master's view.
7. The exception is II.i.14. For brief remarks see II.iii.43; II.iii.48; III.ii.29. On the other hand `Sa^mkara occasionally uses bhedaabheda terminology in his Brahma-suutra-bhaa.sya, e.g., in speaking of kaaryaavastha and kaara.naavastha (II.i.9), and his general attitude is to regard bhedaabheda as a provisionally valid point of view.
to the proper subject of this paper, the arguments against Buddhism, we shell have occasion to note that the two great arguments which `Sa^mkara does not share with Bhaaskara re-occur at length in the B.rhadaara.nyaka commentary, while of the arguments he does share a large number are there omitted.
The explanation of this state of affairs, I think, is this: `Sa^mkara's B.rhadaara.nyakopani.sad-bhaa.sya is a far more original piece of writing than his Brahma-suutra-bhaa.sya. In commenting on the Brahma-suutra `Sa^mkara is very careful not to depart from tradition. On the other hand, in commenting on the B.rhadaara.nyaka it appears to be his intention to break with tradition. He disagrees time and again with Bhart.rprapa~nca, who was probably the most famous interpreter of that text before `Sa^mkara's time.
Let me now apply these criteria of originality to the most notable anti-Buddhist arguments in `Sa^mkara's Brahma-suutra-bhaa.sya. We shall see whether this new method leads to significant results.
The anti-Buddhist section of the Brahma-suutra consists of fifteen sentences: II.ii.18 through II.ii.32 in the enumeration of `Sa^mkara. This section falls into two adhikara.nas (topics), of which the first is directed against the Buddhist realists or, in the words of `Sa^mkara, against those Buddhists who admit the existence of the external world. In this topic there are differences of arrangement between `Sa^mkara and Bhaaskara, but almost complete agreement with regard to the actual arguments employed.
Five specifically Buddhist doctrines are attacked: the doctrine of the aggregates, the chain of causation, the doctrine of momentariness, the Buddhist definition of space (aakaa`sa), and the theory that origin comes only from destruction. These doctrines are well known and I need describe them only briefly.
All entities, according co Buddhist realism, are collocations of atomic particles. Of these particles, there are a great number of types and these are categorized in either two or five great aggregates (samudaaya or skandha). The dual division is into material and mental aggregates. Against this doctrine both `Sa^mkara and Bhaaskara argue that these aggregates are impossible, for there is no conscious agent to cause their aggregation. Or, supposing that they did exist, being wholly independent, there would be no reason for them ever to cease. Essentially, the Vedaanta argument is one of theism against atheism. The necessity of a conscious agent is a favorite Vedaanta
8. Bhart.rprapa~nca's commentary, we are told, was "even more voluminous than `Sa^mkara's." Cf. Hiriyanna, op. cit., p. 77. In this article one will find a lucid account of Bhart.rprapa~nca's views, based on quotations from `Sa^mkara's commentators. The quotations themselves are gathered in Hiriyanna, "Fragments of Bhart.rprapa~nca," Proceeding of the Third All-India Oriental Conference, 1924, pp. 439-450.
argument and is not limited to attacks on Buddhism. It is employed against the Saa^mkhya by the suutras and at great length by the commentators (II.ii.1 ff.). It is used even against the Vai`se.sika (`Sa^mkara and Bhaaskara on II.ii.12), which seems in the days before Pra`sastapaada to have been atheistic. This is obviously one of the oldest arguments of the Vedaanta.
"Chain of causation" is, perhaps, not a paper translation of pratiityasamutpaada, but it has long been used. The chain is a sorties of twelve members which attempts to explain suffering and therefore life. Sensation, desire, taking, becoming, birth, deathóweach member arises in dependence on the preceding member and the chain id unending, like a snake with its tail in its mouth. `Sa^mkara and Bhaaskara use the same criticism: The most that this chain can do is to state that the preceding link is the cause of the succeeding one. It remains to give a cause of the aggregates or of the chain as a whole.
`Sa^mkara elaborates this argument more than Bhaaskara but adds nothing substantially new. Near the end of his comment on II.ii.19 he presents a dilemma. How does one collection give rise to another: like to like, or anything to anything? If the first, men would never become animals or go to hell; if the second, they might become elephants or gods at any moment. A similar ani.s.ta-prasanga (absurd consequence) is found in Bhaaskara six suutras father on.
Each atomic particle is considered by the Buddhists to be momentary. They are said, however, to be capable of forming a continuum on the analogy of the flame of a lamp. `Sa^mkara and Bhaaskara claim that these two statements are contradictory. If the particles are truly commentary, then the action of the former particle must expend itself before the rise of the latter. Like the first of these anti-Buddhist arguments, this must be considered old, for it is closely related to the argument of satkaaryavaada (doctrine of pre-existence of the effect in the cause) which is a basic doctrine of the earliest Vedaanta. The effect must pre-exist in the cause, and the cause must post-exist in the effect.
9. The word "God" is mentioned only once in the Vai`se.sika-suutra and there in the argument of an opponent. In Pra`sastapaada, however, it is God who gives the first impetus to the formation of the atoms. Pra`sastapaada-bhaa.sya. Kaa`si Sanskrit Series No. 3 (Benares: Chowkhamba Sanskrit Book Depot, 1923), p. 20. Since `Sa^mkara takes no account of this in describing the Vai`se.sika system and since his description is at variance in other respects with Pra`sastapaada (e.g., there is nothing in Pra`sastapaada about catura.nukas; cf. Brahma-suutra-bhaa.sya II.ii.11), one may infer that `Sa^mkara is describing a more archaic system. The Ratna-prabhaa on II.ii.11 refers to a Raava.na-bhaa.sya on the Vai`se.sika-suutra. This appears to be now lost but may have been `Sa^mkara's source. See S. K. Belvalker, The Brahma-suutra of Badarayana (Poona: The Oriental Book Agency, 1923), Chap. II, notes, p. 31.
10. This point is argued, with differences of detail, by both `Sa^mkara and Bhaaskara on II.i.18. Perhaps the most elegant exposition of the Vedaanta point of view is in `Sa^mkara on B.rhadaara.nyaka Upani.sad I.ii.1. Anadasrama Sanskrit Series No. 15 (Bombay: Nirnaya Sagar Press, 1939), pp. 21 ff.
The Sarvaastivaadins defined aakaa`sa (space) as aavara.naabhaavamaatram (simple absence of obstruction or covering). Both commentators object on the ground of Vedic scripture. The fact is that the Vedaanta has no concept of space, and the word aakaa`sa does not mean space in the Vedaanta; it refers to a substance and must be translated "air" or "ether." On this question, `Sa^mkara adds a clever quibble on the word aavara.na, but it is not sufficiently important to detain us.
In only one suutra on this topic does `Sa^mkara add an argument that is both original and of basic importance. This is on II.ii.25: anusm.rte`s ca. Indeed, `Sa^mkara's argument here is so keen and so obviously superior to Bhaaskara's that if the latter were copying from `Sa^mkara rather than from the Protocommentator he could hardly have failed to use it.
The laconic words of the suutra are to be translated thus: "and because of memory." Both commentators start out by saying that the Buddhist doctrine of momentariness must imply momentariness of the perceiver as well as of the perceived, an implication which the phenomenon of memory proves to be impossible. If the man who remembers is different from the man who apprehended we would never have such notions as "I saw it."
From here on, Bhaaskara and `Sa^mkara diverge. Bhaaskara proceeds by way of an argumentum ad hominem.
"The continuum of consciousness could not pass to the next world, for the last moment of consciousness could not jump like a frog into an intermediate body at the time of death. . . ."
Buddhist: "Yes it could, by force of karma."
Bhaaskara: "There is no proof that karma has such force."
Buddhist: "The Buddha, who is omniscient, has said that it has."
Bhaaskara: "There is no proof that he was omniscient."
We are here to suppose that the poor Buddhist remains silent, for Bhaaskara continues, "And since there could thus be no going to heaven or hell, there could be no return into the womb. So, the whole of life would be impossible."
`Sa^mkara's argument is considerably longer, but may be condensed as follows:
Buddhist: "As regards the notion
'I saw it,' this may be due to similarity."
`Sa^mkara: "No; the statement 'This is like that' refers to two entities. If you will not admit a single apprehender of the two entities, the statement will be nonsense. Furthermore, in cases such as the notion 'I saw it,' our cognition is in the form of 'this is that,' not in the form of 'this is like that.' Again, we sometimes doubt of an external thing whether it is this or that;
we do not doubt 'Is this the same I who saw that?' "
This is `Sa^mkara at his best, literally unfolding the meaning of scripture for from two words "anusm.rte`s ca" he has, in the coarse of refuting a Buddhist objection, greatly strengthened his favorite metaphysical doctrine, the permanent reality of the self. It is worth nothing that the Buddhist objection here is idealist rather than realist. This doctrine that what we call identity is in fact similarity is not pressed by the Sarvaastivaadins. On the other hand, it is basic to the logic of Dharmakiirti and the Buddhist idealist. That `Sa^mkara's counterargument is his original contribution is rendered the more likely by the fact that he uses it again in two passages of the B.rhadaara.nyaka commentary, once in almost exactly the same form against the Buddhists (on IV.iii.7) and again briefly in setting forth the doctrine of satkaaryavaada (on I.ii.l).
The last argument in this section devoted to Buddhist realism is treated similarly by both commentators. The Buddhists claim that bhaava (existence) arises from abhaava (non-existence) or, to put it somewhat differently, origin comes only from destruction. The grass springs only faun the perished seed.
The criticism used by `Sa^mkara and Bhaaskara is that nothing can come from nothing. This is the maxim of Epicurus, and the reasons adduced by the Indians are the same as those of the Greek. They are, after all, only common sense. Non-being or non-existence is without distinction. That is, we cannot speak of different kinds of non-existence. If we allow birth from non-existence there would be no reason why the glass should come from the perished seed rather than from perished milk or from a hare's horn. Some centuries later, the Naiyaayikas, holding the Buddhist fortress in this conflict, were to forge a powerful weapon against the Vedaanta by analyzing so-called non-existence into a series of negative relations that could be distinguished according to the abhaviiya-pratiyogitaavacchedaka (limiter of the absential counterpositiveness)  but neither `Sa^mkara nor Bhaaskara yet needed to worry about this polysyllabic menace. Of the seed, they said, it is not the perishing condition which causes the grass, but the non-destroyed parts of the seed which form parts of the grass.
To summarize the first anti-Buddhist topic of the Brahma-suutra: `Sa^mkara's arguments against the realists contain substantially nothing new. The one apparent exception is the important argument on II.ii.25, but this, as I have
11. Compare Epicurus apud Diogenes Laertius X.38: "Nothing comes into being out of what does not exist, for [if it did,] anything could come into being out of anything, having no need of seeds [peculiar to itself]."
12. See Ingalls, Materials for the Study of Navya-nyaaya Logic. Harvard Oriental Series Vol. 40 (Cambridge, Mass: Harvard University Press, 1951), Chap. II, Sec. 30.
pointed out, is not actually directed against the realists. It is, rather, a sudden preliminary thrust at Buddhist idealism. This traditionalism in `Sa^mkara's anti-realist arguments is just what one might expect from a general view of the history of Indian philosophy. Buddhist realism had died in India centuries before the time of `Sa^mkara. The Vedaanta arguments against this school go back to an ancient past. One might say the whole question had become academic by the time of `Sa^mkara and that it is only his literary skill that enables him to impart warmth to the conflict.
When we come to the second topic, dealing with Buddhist idealism, we face a different situation.
The topic of Buddhist idealism begins with Brahma-suutra II.ii.28: naabhava upalabdhe.h (the world is not non-existent, because we apprehend it). This and the four succeeding suutras have long been the subject of scholarly debate, both as to their original intention and as to `Sa^mkara's interpretation of them. A comparison of `Sa^mkara and Bhaaskara on these five suutras shows very striking differences. Bhaaskara even omits two of the suutras, considering that they belong rather with the commentary than with the basic text. And for the rest, the major part of his comment consists of criticisms of Dharmakiirti and `Sa^mkara. Since these cannot derive from the Protocommentator, I am led to believe that the Protocommentator had much less to say on this topic than on the last.
It may be worth while to digress for a moment to say that inferences as to the date of the Brahma-suutra based on these five anti-idealist suutras seem to me to rest on very shaky ground. There is no part of the Brahma-suutra on which tradition is so vague. The Buddhist arguments that are severally attacked by the various commentators, so far from being `Suunyavaada, which Jacobi claimed was the original enemy, are not even early Vij~naanavaada. They are chiefly derived from Dharmakiirti, who lived in the seventh century. No other section of the Brahma-suutra is under such strong suspicion of being an insertion. Except for these five suutras, I can see no reason for placing the Brahma-suutra later than the time of Christ. To shift the date five centuries or so on the basis of five Suutras, two of which are not even universally recognized, seems to me to be opposed to all historical method and common sense.
To return to `Sa^mkara. He begins his comment on II.ii.28 with an elaborate idealist puurvapak.sa (statement of opposition) consisting of five arguments designed to show the unreality of the outer world. The arguments are taken
13. Bhaaskara omits suutras 31 and 32, Raamaanuja, `Sriika.n.tha, and `Sriipati omit 31 but not 32.
14. I refer in particular to the attempt of Jacobi, "The Dates of the Philosophical Suutras of the Brahmans," Journal of the American Oriental Society, XXXI (1911), 1-29.
chiefly from Dharmakiirti and are answered seriatim.
One of the arguments runs as follows: Although cognition remains the same wherever it arises, it does so always with a certain particularization, that is, with a bias toward a particular object. Our cognitions are always cognitions of something. This particularization being, so to speak, completely enclosed in the knowledge, there is no need of our positing anything outside the knowledge. The particularization need actually be only of the cognition, not of the thing. The pertinent verses of Dharmakiirti are actually quoted by Bhaaskara, who refers to the author scornfully as viprabhiksu, i.e., the Brahmin monk ( Bhaaskara, op. cit., p. 123, line 15). One of the verses occurs in Dharmakiirti's Pramaa.na-vaarttikaa, but both verses actually occur together in the same author's Pramaa.na-vini`scayaa, which is now preserved only in the Tibetan. This work is undoubtedly the source on which `Sa^mkara and Bhaaskara drew.
A second argument derives from Dharmakiirti's verse:
Sahopalambha-niyamaad abhedo niila-tad-dhiya.h
Bheda`s ca bhraanti-vij~naanair d.r`syetendaav ivaadvaye.
(Blue and the cognition of blue are not different entities, for the on invariably occurs with the other. One should recognize their difference as due to false cognition, like [the double moon seen by an astygmatic] in the moon, which is single.)
The doctrine of the simultaneity of cognition and content gains strength from the phenomenon of apperception, that is, cases of reflective knowledge in such forms as "I know that I see the post." It is said that in apperception the object and the knowledge are never separate. This can be only because they are identical in nature.
15. Apratyak.sopalambhasya naarthadr.s.ti.h prasiddhyati. (a)
Avibhaago 'pi buddhyaatmaa viparyasita-dar`sanai.h;
Graahya-graahaka sa.mvitti-bhedavaan iva lak.syate. (b)
The half-verse (a) occurs in the Pramaa.na-vini`scaya (Narthang edition of Tanjur, Mdo 115, folio 274b). The complete verse (b) occurs in the same text, folio 273b, and also in Dharmakiirti's Pramaa.navaarttika II.354. I owe the reference for verse (a) to the kindness of my friend Jain Muni Jambuvijay. Verse (b), which `Sa^mkara also uses in his Upade`sa-saahasrii XVIII.142, and which is quoted in the Sarvadar`sana-sa^mgraha, Bibliotheca Indica Work No. 21 (Calcutta: Asiatic Society of Bengal, 1858), p. 16, has been traced to its source by L. de la Vallee Poussin, Museon, New Series II (1901), 182, note 86. See also Nakamura, op. cit., p. 106, note 3.
16. This is the verse quoted by Vaacaspati Mi`sra in his Bhaamatii on II.ii.28 (op. cit. in note 3, p. 3, p. 544). It is again quoted in Sarva-dar`sana-sa^mgraha, p. 16. The two halves did not originally belong together. They are traced by de la Vallee Poussin (op. cit., p. 182, note 85) to Pramaa.na-vini`scaya, Tanjur Mdo 115, folio 274a, and to Pramaa.na-vaarttika, ibid., folio 239b. In the recently recovered Sanskrit text of the Pramaa.na-vaarttika the second half-verse forms part of III.388 (Dharmakiirti, Pramaa.navaarttikam with Commentary of Manorathanandin, Raahula Sa^mkratyaayana, ed. Patna: Bihar and Orissa Research Society, 1938-40).
`Sa^mkara's particular answers to these Buddhist pronouncements are not so important as his general criticism. In particular, he says that if knowledge takes the form of the object from its, so to speak, containing it, this fact could not be so if there existed no object. Again, the fact that idea and object occur together in apperception shows no more than that they stand to each other in the relation of means and object.
More important are the general criticisms. No one apprehends a post or a wall as an apprehension, but only as an object of apprehension, namely, a post or a wall. This is almost the same as the argument in favor of true identity that I have pointed out above and that I tried to show is original with `Sa^mkara. When we recognize something, we recognize that A is B, not that A is like B.
Finally, `Sa^mkara uses the following argument, and this is his heaviest artillery. He uses it again in commenting on the B.rhadaara.nyaka (IV.iii.7), but it is significantly absent from Bhaaskara.
There must be something beyond the cognition, namely, a cognizer.
The Buddhist is then made to object that to allow the cognition to be grasped by something outside itself leads to an infinite regress. Something still further must then grasp the grasper. It is to avoid this that he stops at the cognition, which he regards as self-luminous, like a lamp.
But `Sa^mkara counters by saying that this cognition could not be reached by any means, nor could it have anyone to understand it. It would be like a thousand lamps set in a wilderness. For illumination the lamp needs an eye. A witness is necessary in order to have a cognition. And there is no logical necessity (aakaa^nk.saa) for something to grasp the grasper. The witness stands self-proved.
Buddhist: "But in urging against me the self-validity of the knower you are merely using my theory under a different name."
`Sa^mkara: "No, because you claim that cognition is momentary and multiple."
I ask the reader to bear these final words in mind, for I shall refer to them again in summing up `Sa^mkara's attitude.
Another of the idealist arguments set forth in II.ii.28 receives its answer in the next suutra. Here Bhaaskara and `Sa^mkara agree even to the point of verbal similarity and the argument must go back to the Protocommentator. `Sa^mkara, as we shall see, has been accused of hypocrisy in repeating the traditional view.
The Buddhist argument runs thus: Just as our ideas of dreams and mirages have the form of perceiver and perceived although they lack external objects,
so also do our ideas in the waking state. Or, to put the matter more simply: The ideas we have in dreams are false; therefore, the ideas we have when awake are false, because they are ideas.
The answer is based on common sense. Dreaming is different from waking. The ideas we have in dreams are different from those we have when awake. The first are sublated by awakening; the second are not.
Bhaaskara is infuriated by `Sa^mkara's conservative piety. Referring to "those maayaavaadins who hang on to Buddhist doctrine," be claims that the present suutra shows their philosophy to be entirely wrong. I shall return to this question later. For the present let me point out two facts. `Sa^mkara does not use this argument again against the Buddhists in the B.rhadaara.nyaka, but he nowhere contradicts it. It is `Sa^mkara's followers who slip over into the Buddhist position, not `Sa^mkara. The author of the Viveka-cuu.daama.ni uses exactly this Buddhist argument to prove the unreality of the external world. But the author of the Viveka-cuu.daama.ni was not `Sa^mkaraacaarya.
Let me complete my survey of the anti-idealist topic. Under II.ii.31 `Sa^mkara considers briefly the aalaya-vij~naana of the Buddhist idealists. Aalayavij~naana or reservoir-consciousness is a concept set up by the Vij~naanavaadins to provide something permanent in the constant flux of momentary particles. It is a consciousness or cognition as abstracted from all terms of the relations in which cognition occurs. That is, it is pure consciousness, not consciousness of anything. Metaphysically it is similar to `S^mkara's brahma, which is knowledge devoid of all the terms in which knowledge occurs, in other words, pure knowledge, not knowledge of anything. But there is a psychological and historical difference between the two terms. First, let me paraphrase `Sa^mkara's words on II.ii.31.
"The reservoir-consciousness that you set up, being momentary, is no better than ordinary consciousness. Or, if you allow the reservoir-consciousness to be lasting, you destroy your theory of momentariness."
The reservoir-consciousness appears to `Sa^mkara as the last-minute thought of a school which has spent most of its effort in nihilism, a notion pulled in by the hair to save the system from becoming sheer nonsense. His own brahma as pure knowledge, however, he regards not as the outcome of nihilism, but as the quintessence of positive reality.
Finally, `Sa^mkara casts a glance at the `Suunyavaada, which he has elsewhere
17. Viveka-cuu.daamani 170. See Swami Madhavananda, Vivekachudamani of Sri Sankaracharya, Text, with English Translation, etc. (5th ed., Mayavati, District Almora, Himalayas: Advaita Ashrama, 1952).
18. See my article "The Study of `Sa^mkaraacaarya," Annals, Bhandarkar Oriental Research Institute, XXXIII (1952), 7.
neglected. "The `Suunyavaada," he says, and he repeats himself word for word on B.rhadaara.nyaka IV.iii.7, "being contradictory to all valid means of knowledge, we have not thought worth while to refute." He then adds the important statement: "Common sense (loka-vyavahaara) cannot be denied without the discovery of some other truth."
`Sa^mkara's arguments against the Buddhists in his Brahma-Suutra-bhaa.sya end at this point. Having reviewed them in detail, let me now offer some general considerations.
In the first place, we see that the majority of these arguments are traditional. But, granted this basic fact, there is still a noticeable difference between `Sa^mkara's arguments against the Buddhist realists and those against the idealists. The arguments against the idealists show more originality. This is entirely in keeping with what we know of the history of philosophy. In India of the eighth century, Buddhist idealism was still a living system of thought although its influence was rapidly declining.
Two important arguments against the idealists appear to be `Sa^mkara's original contribution: the argument for identity and the argument for the witness. Both of these are found elsewhere in `Sa^mkara and are noticeably absent from Bhaaskara. Both of them are positive arguments in that they go far beyond the simple refutation of an opponent's theory. Both arguments are integral to `Sa^mkara's system of metaphysics, for the witness is the center of all `Sa^mkara's philosophy. It is the light by which everything is seen, the light of which the sun and moon are pale reflections. It is not only real but so egregiously real that the workaday world fades into mist beside it. And the only fact that enables us to realize this truth is that it is the witness within us. We realize it by realizing an identity: "tat tvam asi" (that art thou); or "brahm'smi" (I am brahma). This is no watered similarity as if we should say, "I am something like brahma", it is an absolute identity and this is ultimately proved simply by psychological experience. As `Sa^mkara states at the beginning of his Brahma-suutra comment: Everyone has the notion "I am"; no one can deny the self, for it is the self even of the denier. Both these arguments, that of the witness and that of identity, are at the very center of `Sa^mkara's system of Vedaanta.
Let me next consider Bhaaskara's charge of hypocrisy against `Sa^mkara. The same charge has been brought forward by a few scholars in modern times also. It may be put most strongly as follows. In arguing against the Vij~naanavaada, `Sa^mkara seems to argue as a realist. Certainly he seems to insist on
19. See `Sa^mkara, Brahma-suutra-bhaa.sya (edition given under note 3) I.i.1 (p. 81); I.i.4 (p. 134); II.iii.7 (P. 585).
the reality of the external thing. On the other hand, when developing his own system of philosophy, he claims not once but a hundred times that the world is unreal, as unreal as the foam on water, as the trick of a magician, as a mirage, as a dream. If not hypocrisy, this is at least a logical contradiction. Or is it?
Notice, first, that in arguing with the Buddhist idealists `Sa^mkara's emphasis is not so much on the necessity of the external world as it is on the necessity of something beyond cognition. He has at least psychological justification for this so long as his brahma is conceived as the cognizer rather than as cognition itself. When it comes to logical definition, of course, there is small difference between the Buddhist and the Vedanta concepts, but `Sa^mkara admits that brahma is not logically definable. And then he leaves himself a logical loophole. Remember his words: "Common sense cannot be denied without the discovery of some other truth." Indeed, Bhaaskara, in condemning `Sa^mkara, seems to have understood `Sa^mkara's point of view perfectly. "Perhaps you reason thus," he says. "So long as there is no other, permanent entity, pure consciousness is to be denied. But when there is the true aatma, then we can get rid of the world."  Bhaaskara goes on to ask, "What sort of logic is this?" But to answer his question would require a large treatise.
I myself would judge the evidence in this way. If we are to adopt a meta-physical and static view of philosophy, there is little difference between `Sa^mkara and Vij~naanavaada Buddhism, so little, in fact, that the whole discussion seems fairly pointless. The central reality of both systems is a qualityless, changeless unity, and from this point of view there is much justice in the Buddhist objection that I have quoted: "But in urging against me the self-luminousness of the knower you are merely using my theory under a different name."
But if we try to think psychologically and historically, that is, if we try to think our way back into the minds of the philosophers whose works we read, there is a very real difference between these antagonists.
To `Sa^mkara and to many of his followers there appeared almost no similar between Kevalaadvaita and idealistic Buddhism. `Sa^mkara did not begin by denying the reality of the workaday world; he was forced into this position in order only to explain the unchanging and eternal and universal brahma. This appears very clearly when we review the passages in which `Sa^mkara defends his theory of avidyaa (ignorance) and maayaa (illusion). One may sum up the defense under three headings. First, the inner truth (see, for
20. Bhaaskara, op. cit., p. 124, line 19.
example, I.i.1). Given the immediate knowledge of the constant self, we cannot tightly attribute to this self the variations of the external world. Then, the ultimate truth (see, for example, II.i.27). Given the concept of a particles brahma, which is universal, we cannot explain the external world as parts of brahma, nor can we explain changing circumstances as varying states of brahma. The external world must be simply an appearance. Thirdly, scripture (for example, II.i.14). Naturally, `Sa^mkara pits this first and derives his theory from his interpretation of such passages as tat kena ka^m vijaaniiyaat (then [viz., when duality has ceased] by what means world one know what? ) and vaacaarambha.na^m vikaaro naamadheyam (according to the interpretation of `Sa^mkara, which has been much contested, this means: "the alteration [of the one into many] depends on speech; it is simply a name"). Whatever facet of this defense we examine, we find it begins with the reality of brahma; the unreality of the external world follows only as a deduction.
Now, the `Sa^mkara and his followers it seemed that the Buddhists completely reversed this process. They began with their theory of nihilism. They found no composites, no wholes, in the world, only constituents, and these were particles existing for only an atom of time. Everything else was divisible; everything was constantly changing. A few Buddhist idealists, `Sa^mkara realized, had posited an unchanging unit within which all this change could take place and they called this unchanging continuum reservoir-consciousness (aalaya-vij~naana) or pure consciousness (vij~naptimaatrataa). But to `Sa^mkara this seemed simply a hypocritical attempt to patch up the picture after it had been torn to shreds. From his comments on Brahma-suutra II.ii.31 and B.rhadaara.nyaka IV.iii.7 it is quite clear that `Sa^mkara refused to take the aalaya-vij~naana seriously.
From a psychological point of view, then, `Sa^mkara's criticism is understandable. From a historical point of view it may be justified. The historical difference goes back to the foundations of both traditions, to the Upani.sads and to the words of the Buddha himself. There may be parallels between these two bodies of teaching, as some recent scholars have tried to show, but it seems to me that are superficial parallels.
The immediate and unmistakable concern of the Upani.sads is with the one principle of life. The approach is sometimes realistic, sometimes idealistic, sometimes enlightened, and sometimes on the basis of ritual and magic. But the one principle is the same. It is omnipresent, it is all-powerful, and we are constantly told that it is bliss, that it is the one without a second, and that it is the existent. Throughout the words of the Buddha, on the other hand, there runs with equal persistence the theme of the mystery of the world.
The world is essentially multiple, and the one hope for unhappy mankind is that if the dharmas (elements of existence) come into being they likewise cease to exist. The Upani.sad tradition is essentially aristocratic and priestly. The Buddhist tradition, on the other hand, is one of revolt.
`Sa^mkara came at a point in history when these two traditions, on an intellectual plane at least, had almost coincided. Out of their original pluralism the Buddhists had evolved a unity, while the Vedaanta had left its early joyous acceptance of the whole of life. It had so concentrated its effort toward the peak of nirgu.na-brahma (qualityless brahma) that the workaday world had become as sorry a place as it was to the early Buddhists. But there remained the memory of a thousand years of mutually antagonistic traditions, and there was still a present and real psychological difference. On these accounts, `Sa^mkara's arguments against the Buddhists, both those which he repeats from the past and those which he originates, seem to me not pointless but deeply significant and worthy of study.