Sacred Texts  Hinduism  Index  Previous  Next 
Buy this Book at

The Vedanta Sutras, commentary by Sankaracharya (SBE34), tr. by George Thibaut [1890] at



Adhikarana I (1, 2).--The meditation on the Âtman enjoined by Scripture is not an act to be accomplished once only, but is to be repeated again and again.

Adhik. II (3).--The devotee engaged in meditation on Brahman is to view it as constituting his own Self.

p. lxvii

Adhik. III (4).--To the rule laid down in the preceding adhikarana the so-called pratîkopâsanas, i.e. those meditations in which Brahman is viewed under a symbol or outward manifestation (as, for instance, mano brahmety upâsîta) constitute an exception, i.e. the devotee is not to consider the pratîka as constituting his own Self.

Adhik. IV (5).--In the pratîkopâsanas the pratîka is to be meditatively viewed as being one with Brahman, not Brahman as being one with the pratîka.--Râmânuga takes Sûtra 5 as simply giving a reason for the decision arrived at under Sûtra 4, and therefore as not constituting a new adhikarana.

Adhik. V (6).--In meditations connected with constitutives of sacrificial works (as, for instance, ya evâsau tapati tam udgîtham upâsîta) the idea of the divinity, &c. is to be transferred to the sacrificial item, not vice versâ. In the example quoted, for instance, the udgîtha is to be viewed as Âditya, not Âditya as the udgîtha.

Adhik. VI (7-10).--The devotee is to carry on his meditations in a sitting posture.--Sankara maintains that this rule does not apply to those meditations whose result is samyagdarsana; but the Sûtra gives no hint to that effect.

Adhik. VII (11).--The meditations maybe carried on at any time, and in any place, favourable to concentration of mind.

Adhik. VIII (12).--The meditations are to be continued until death.--Sankara again maintains that those meditations which lead to samyagdarsana are excepted.

Adhik. IX (13).--When through those meditations the knowledge of Brahman has been reached, the vidvân is no longer affected by the consequences of either past or future evil deeds.

Adhik. X (14).--Good deeds likewise lose their efficiency.--The literal translation of the Sûtra is, 'There is likewise non-attachment (to the vidvân) of the other (i. e. of the deeds other than the evil ones, i. e. of good deeds), but on the fall (of the body, i.e. when death takes place).' The last words of the Sûtra, 'but on the fall,' are separated by Sankara from the preceding part of the Sûtra and interpreted to mean, 'when death takes place (there results mukti of

p. lxxviii

the vidvân, who through his knowledge has freed himself from the bonds of works).'--According to Râmânuga the whole Sûtra simply means, 'There is likewise non-attachment of good deeds (not at once when knowledge is reached), but on the death of the vidvân 1.'

Adhik. XI (15).--The non-operation of works stated in the two preceding adhikaranas holds good only in the case of anârabdhakârya works, i.e. those works which have not yet begun to produce their effects, while it does not extend to the ârabdhakârya works on which the present existence of the devotee depends.

Adhik. XII (16, 17).--From the rule enunciated in Adhik. X are excepted such sacrificial performances as are enjoined permanently (nitya): so, for instance, the agnihotra, for they promote the origination of knowledge.

Adhik. XIII (18).--The origination of knowledge is promoted also by such sacrificial works as are not accompanied with the knowledge of the upâsanas referring to the different members of those works.

Adhik. XIV (19).--The ârabdhakârya works have to be worked out fully by the fruition of their effects; whereupon the vidvân becomes united with Brahman.--The 'bhoga' of the Sûtra is, according to Sankara, restricted to the present existence of the devotee, since the complete knowledge obtained by him destroys the nescience which otherwise would lead to future embodiments. According to Râmânuga a number of embodied existences may have to be gone through before the effects of the ârabdhakârya works are exhausted.


This and the two remaining pâdas of the fourth adhyâya describe the fate of the vidvân after death. According to Sankara we have to distinguish the vidvân who possesses the highest knowledge, viz. that he is one with the highest

p. xxix

[paragraph continues] Brahman, and the vidvân who knows only the lower Brahman, and have to refer certain Sûtras to the former and others to the latter. According to Râmânuga the vidvân is one only.

Adhik. I, II, III (1-6).--On the death of the vidvân (i.e. of him who possesses the lower knowledge, according to Sankara) his senses are merged in the manas, the manas in the chief vital air (prâna), the vital air in the individual soul (gîva), the soul in the subtle elements.--According to Râmânuga the combination (sampatti) of the senses with the manas, &c. is a mere conjunction (samyoga), not a merging (laya).

Adhik. IV (7).--The vidvân (i.e. according to Sankara, he who possesses the lower knowledge) and the avidvân, i.e. he who does not possess any knowledge of Brahman, pass through the same stages (i.e. those described hitherto) up to the entrance of the soul, together with the subtle elements, and so on into the nâdîs.--The vidvân also remains connected with the subtle elements because he has not yet completely destroyed avidyâ, so that the immortality which Scripture ascribes to him (amritatvam hi vidvân abhyasnute) is only a relative one.--Râmânuga quotes the following text regarding the immortality of the vidvân:

'Yadâ sarve pramukyante kâmâ ye # sya hridi sthitâh
atha martyo # mrito bhavaty atra brahma samasnute,'

and explains that the immortality which is here ascribed to the vidvân as soon as he abandons all desires can only mean the destruction--mentioned in the preceding pâda--of all the effects of good and evil works, while the 'reaching of Brahman' can only refer to the intuition of Brahman vouchsafed to the meditating devotee.

Adhik. V (8-11) raises; according to Sankara, the question whether the subtle elements of which Scripture says that they are combined with the highest deity (tegah parasyâm devatâyâm) are completely merged in the latter or not. The answer is that a complete absorption of the elements takes place only when final emancipation is reached; that, on the other hand, as long as the samsâra state lasts, the elements, although somehow combined with

p. lxxx

[paragraph continues] Brahman, remain distinct so as to be able to form new bodies for the soul.

According to Râmânuga the Sûtras 8-11 do not constitute a new adhikarana, but continue the discussion of the point mooted in 7. The immortality there spoken of does not imply the separation of the soul from the body, 'because Scripture declares samsâra, i. e. embodiedness up to the reaching of Brahman' (tasya tâvad eva kiram yâvan na vimokshye atha sampatsye) (8).--That the soul after having departed from the gross body-is not disconnected from the subtle elements, is also proved hereby, that the subtle body accompanies it, as is observed from authority 1 (9).--Hence the immortality referred to in the scriptural passage quoted is not effected by means of the total destruction of the body (10).

Adhik. VI (12-14) is of special importance.--According to Sankara the Sûtras now turn from the discussion of the departure of him who possesses the lower knowledge only to the consideration of what becomes of him who has reached the higher knowledge. So far it has been taught that in the case of relative immortality (ensuing on the apara vidyâ) the subtle elements, together with the senses and so on, depart from the body of the dying devotee; this implies at the same time that they do not depart from the body of the dying sage who knows himself to be one with Brahman.--Against this latter implied doctrine Sûtra 12 is supposed to formulate an objection. 'If it be said that the departure of the prânas from the body of the dying sage is denied (viz. in Bri. Up. IV, 4, 5, na tasya prâna utkrâmanti, of him the prânas do not pass out); we reply that in that passage the genitive "tasya" has the sense of the ablative "tasmât," so that the sense of the passage is, "from him, i.e. from the gîva of the dying sage, the prânas do not depart, but remain with it."--This objection Sankara supposes to be disposed of in Sûtra 13. 'By some there is given a clear denial of the departure of the prânas in the case of the

p. lxxxi

dying sage,' viz. in the passage Bri. Up. III, 2, 11, where Yâavalkya instructs Ârtabhâga that, when this man dies, the prânas do not depart from it (asmât; the context showing that asmât means 'from it,' viz. from the body, and not 'from him,' viz. the gîva).--The same view is, moreover, confirmed by Smriti passages.

According to Râmânuga the three Sûtras forming Sankara's sixth adhikarana do not constitute a new adhikarana at all, and, moreover, have to be combined into two Sûtras. The topic continuing to be discussed is the utkrânti of the vidvân. If, Sûtra 12 says, the utkrânti of the prânas is not admitted, on the ground of the denial supposed to be contained in Bri. Up. IV, 4, 5; the reply is that the sense of the tasya there is 'sârîrât' (so that the passage means, 'from him, i.e. the gîva, the prânas do not depart'); for this is clearly shown by the reading of some, viz. the Mâdhyandinas, who, in their text of the passage, do not read 'tasya' but 'tasmât.'--With reference to the instruction given by Yâavalkya to Ârtabhâga, it is to be remarked that nothing there shows the 'ayam purusha' to be the sage who knows Brahman.--And, finally, there are Smriti passages declaring that the sage also when dying departs from the body.

Adhik. VII and VIII (15, 16) teach, according to Sankara, that, on the death of him who possesses the higher knowledge, his prânas, elements, &c. are merged in Brahman, so as to be no longer distinct from it in any way.

According to Râmânuga the two Sûtras continue the teaching about the prânas, bhûtas, &c. of the vidvân in general, and declare that they are finally merged in Brahman, not merely in the way of conjunction (samyoga), but completely 1.

Adhik. IX (17).--Sankara here returns to the owner of the aparâ vidyâ, while Râmânuga continues the description of the utkrânti of his vidvân.--The gîva of the dying man

p. lxxxii

passes into the heart, and thence departs out of the body by means of the nâdis; the vidvân by means of the nâdi called sushumnâ, the avidvân by means of some other nâdî.

Adhik. X (18, 19).--The departing soul passes up to the sun by means of a ray of light which exists at night as well as during day.

Adhik. XI (20, 21).--Also that vidvân who dies during the dakshinâyana reaches Brahman.


Adhik. I, II, III (1-3) reconcile the different accounts given in the Upanishads as to the stations of the way which leads the vidvân up to Brahman.

Adhik. IV (4-6)-By the 'stations' we have, however, to understand not only the subdivisions of the way but also the divine beings which lead the soul on.

The remaining part of the pâda is by Sankara divided into two adhikaranas. Of these the former one (7-14) teaches that the Brahman to which the departed soul is led by the guardians of the path of the gods is not the highest Brahman, but the effected (kârya) or qualified (saguna) Brahman. This is the opinion propounded in Sûtras 7-11 by Bâdari, and, finally, accepted by Sankara in his commentary on Sûtra 14. In Sûtras 12-14 Gaimini defends the opposite view, according to which the soul of the vidvân goes to the highest Brahman, not to the kâryam brahma. But Gaimini's view, although set forth in the latter part of the adhikarana, is, according to Sankara, a mere pûrvapaksha, while Bâdari's opinion represents the siddhânta.--The latter of the two adhikaranas (VI of the whole pâda; 15,16) records the opinion of Bâdarâyana on a collateral question, viz. whether, or not, all those who worship the effected Brahman are led to it. The decision is that those only are guided to Brahman who have not worshipped it under a pratîka form.

According to Râmânuga, Sûtras 7-16 form one adhikarana only, in which the views of Bâdari and of Gaimini represent two pûrvapakshas, while Bâdarâyana's opinion is adopted

p. lxxxiii

as the siddhânta. The question is whether the guardians of the path lead to Brahman only those who worship the effected Brahman, i.e. Hiranyagarbha, or those who worship the highest Brahman, or those who worship the individual soul as free from Prakriti, and having Brahman for its Self (ye pratyagâtmânam prakritiviyuktam brahmâtmakam upâsate).--The first view is maintained by Bâdari in Sûtra 7, 'The guardians lead to Brahman those who worship the effected Brahman, because going is possible towards the latter only;' for no movement can take place towards the highest and as such omnipresent Brahman.--The explanation of Sûtra 9 is similar to that of Sankara; but more clearly replies to the objection (that, if Hiranyagarbha were meant in the passage, 'purusho#mânavah sa etân brahma gamayati,' the text would read 'sa etân brahmânam gamayati') that Hiranyagarbha is called Brahman on account of his nearness to Brahman, i.e. on account of his prathamagatva.---The explanation of 10, 11 is essentially the same as in Sankara; so also of 12-14.--The siddhânta view is established in Sûtra 13, 'It is the opinion of Bâdarâyana that it, i.e. the gana of the guardians, leads to Brahman those who do not take their stand on what is pratîka, i.e. those who worship the highest Brahman, and those who meditate on the individual Self as dissociated from prakriti, and having Brahman for its Self, but not those who worship Brahman under pratîkas. For both views--that of Gaimini as well as that of Bâdari--are faulty.' The kârya view contradicts such passages as 'asmâk kharîrât samutthâya param gyotir upasampadya,' &c.; the para view, such passages as that in the pañkâgni-vidyâ, which declares that ya ittham viduh, i.e. those who know the pañkâgni-vidyâ, are also led up to Brahman.


Adhik. I (1-3) returns, according to--Sankara, to the owner of the parâ vidyâ, and teaches that, when on his death his soul obtains final release, it does not acquire any new characteristics, but merely manifests itself in its true nature.--The explanation given by Râmânuga is essentially

p. lxxxiv

the same, but of course refers to that vidvân whose going to Brahman had been described in the preceding pâda.

Adhik. II (4) determines that the relation in which the released soul stands to Brahman is that of avibhâga, non-separation. This, on Sankara's view, means absolute non-separation, identity.--According to Râmânuga the question to be considered is whether the released soul views itself as separate (prithagbhûta) from Brahman, or as non-separate because being a mode of Brahman. The former view is favoured by those Sruti and Smriti passages which speak of the soul as being with, or equal to, Brahman; the latter by, such passages as tat tvam asi and the like. 1

Adhik. Ill (5-7) discusses the characteristics of the released soul (i.e. of the truly released soul, according to Sankara). According to Gaimini the released soul, when manifesting itself in its true nature, possesses all those qualities which in Kh. Up. VIII, 7, I and other places are ascribed to Brahman, such as apahatapâpmatva, satyasamkalpatva, &c., aisvarya.--According to Audulomi the only characteristic of the released soul is kaitanya.--According to Bâdarâyana the two views can be combined (Sankara remarking that satyasamkalpatva, &c. are ascribed to the released soul vyavahârâpekshayâ).

Adhik. IV (8-9) returns, according to Sankara, to the aparâ vidyâ, and discusses the question whether the soul of

p. lxxxv

the pious effects its desires by its mere determination, or uses some other means. The former alternative is accepted--According to Râmânuga the adhikarana simply continues the consideration of the state of the released, begun in the preceding adhikarana. Of the released soul it is said in Kh. Up. VIII, 12, 3 that after it has manifested itself in its true nature it moves about playing and rejoicing with women, carriages, and so on. The question then arises whether it effects all this by its mere samkalpa (it having been shown in the preceding adhikarana that the released soul is, like the Lord, satyasamkalpa), or not. The answer is in favour of the former alternative, on account of the explicit declaration made in Kh. Up. VIII, 2, 'By his mere will the fathers come to receive him.'

Adhik. V (10-14) decides that the released are embodied or disembodied according to their wish and will.

Adhik. VI (11, 12) explains how the soul of the released can animate several bodies at the same time.--Sûtra 12 gives, according to Sankara, the additional explanation that those passages which declare the absence of all specific cognition on the part of the released soul do not refer to the partly released soul of the devotee, but either to the soul in the state of deep sleep (svâpyaya = sushupti), or to the fully released soul of the sage (sampatti = kaivalya).--Râmânuga explains that the passages speaking of absence of consciousness refer either to the state of deep sleep, or to the time of dying (sampatti = maranam according to 'vân manasi sampadyate,' &c.).

Adhik. VII (17-21).--The released gîvas participate in all the perfections and powers of the Lord, with the exception of the power of creating and sustaining the world. They do not return to new forms of embodied existence.


After having, in this way, rendered ourselves acquainted with the contents of the Brahma-sûtras according to the views of Sankara as well as Râmânuga, we have now to consider the question which of the two modes of interpretation represents--or at any rate more closely approximates to--the true meaning of the Sûtras. That

p. lxxxvi

few of the Sûtras are intelligible if taken by themselves, we have already remarked above; but this does not exclude the possibility of our deciding with a fair degree of certainty which of the two interpretations proposed agrees better with the text, at least in a certain number of cases.

We have to note in the first place that, in spite of very numerous discrepancies,--of which only the more important ones have been singled out in the conspectus of contents,--the two commentators are at one as to the general drift of the Sûtras and the arrangement of topics. As a rule, the adhikaranas discuss one or several Vedic passages bearing upon a certain point of the system, and in the vast majority of cases the two commentators agree as to which are the special texts referred to. And, moreover, in a very large number of cases the agreement extends to the interpretation to be put on those passages and on the Sûtras. This far-reaching agreement certainly tends to inspire us with a certain confidence as to the existence of an old tradition concerning the meaning of the Sûtras on which the bulk of the interpretations of Sankara as well as of Râmânuga are based.

But at the same time we have seen that, in a not inconsiderable number of cases, the interpretations of Sankara and Râmânuga diverge more or less widely, and that the Sûtras affected thereby are, most of them, especially important because bearing on fundamental points of the Vedânta system. The question then remains which of the two interpretations is entitled to preference.

Regarding a small number of Sûtras I have already (in the conspectus of contents) given it as my opinion that Râmânuga's explanation appears to be more worthy of consideration. We meet, in the first place, with a number of cases in which the two commentators agree as to the literal meaning of a Sûtra, but where Sankara sees himself reduced to the necessity of supplementing his interpretation by certain additions and reservations of his own for which the text gives no occasion, while Râmânuga is able to take the Sûtra as it stands. To exemplify this remark, I again direct attention to all those Sûtras which in

p. lxxxvii

clear terms represent the individual soul as something different from the highest soul, and concerning which Sankara is each time obliged to have recourse to the plea of the Sûtra referring, not to what is true in the strict sense of the word, but only to what is conventionally looked upon as true. It is, I admit, not altogether impossible that Sankara's interpretation should represent the real meaning of the Sûtras; that the latter, indeed, to use the terms employed by Dr. Deussen, should for the nonce set forth an exoteric doctrine adapted to the common notions of mankind, which, however, can be rightly understood by him only to whose mind the esoteric doctrine is all the while present. This is not impossible, I say; but it is a point which requires convincing proofs before it can be allowed.--We have had, in the second place, to note a certain number of adhikaranas and Sûtras concerning whose interpretation Sankara and Râmânuga disagree altogether; and we have seen that not unfrequently the explanations given by the latter commentator appear to be preferable because falling in more easily with the words of the text. The most striking instance of this is afforded by the 13th adhikarana of II, 3, which treats of the size of the gîva, and where Râmânuga's explanation seems to be decidedly superior to Sankara's, both if we look to the arrangement of the whole adhikarana and to the wording of the single Sûtras. The adhikarana is, moreover, a specially important one, because the nature of the view held as to the size of the individual soul goes far to settle the question what kind of Vedânta is embodied in Bâdarâyana's work.

But it will be requisite not only to dwell on the interpretations of a few detached Sûtras, but to make the attempt at least of forming some opinion as to the relation of the Vedânta-sûtras as a whole to the chief distinguishing doctrines of Sankara as well as Râmânuga. Such an attempt may possibly lead to very slender positive results; but in the present state of the enquiry even a merely negative result, viz. the conclusion that the Sûtras do not teach particular doctrines found in them by certain commentators, will not be without its value.

p. lxxxviii

The first question we wish to consider in some detail is whether the Sûtras in any way favour Sankara's doctrine that we have to distinguish a twofold knowledge of Brahman, a higher knowledge which leads to the immediate absorption, on death, of the individual soul in Brahman, and a lower knowledge which raises its owner merely to an exalted form of individual existence. The adhyâya first to be considered in this connexion is the fourth one. According to Sankara the three latter pâdas of that adhyâya are chiefly engaged in describing the fate of him who dies in the possession of the lower knowledge, while two sections (IV, 2, 12-14; IV, 4, 1-7) tell us what happens to him who, before his death, had risen to the knowledge of the highest Brahman. According to Râmânuga, on the other hand, the three pâdas, referring throughout to one subject only, give an uninterrupted account of the successive steps by which the soul of him who knows the Lord through the Upanishads passes, at the time of death, out of the gross body which it had tenanted, ascends to the world of Brahman, and lives there for ever without returning into the samsâra.

On an à priori view of the matter it certainly appears somewhat strange that the concluding section of the Sûtras should be almost entirely taken up with describing the fate of him who has after all acquired an altogether inferior knowledge only, and has remained shut out from the true sanctuary of Vedântic knowledge, while the fate of the fully initiated is disposed of in a few occasional Sûtras. It is, I think, not too much to say that no unbiassed student of the Sûtras would--before having allowed himself to be influenced by Sankara's interpretations--imagine for a moment that the solemn words, 'From thence is no return, from thence is no return,' with which the Sûtras conclude, are meant to describe, not the lasting condition of him who has reached final release, the highest aim of man, but merely a stage on the way of that soul which is engaged in the slow progress of gradual release, a stage which is indeed greatly superior to any earthly form of existence, but yet itself belongs to the essentially fictitious samsâra.

p. lxxxix

and as such remains infinitely below the bliss of true mukti. And this à priori impression--which, although no doubt significant, could hardly be appealed to as decisive--is confirmed by a detailed consideration of the two sets of Sûtras which Sankara connects with the knowledge of the higher Brahman. How these Sûtras are interpreted by Sankara and Râmânuga has been stated above in the conspectus of contents; the points which render the interpretation given by Râmânuga more probable are as follows. With regard to IV, 2, 12-14, we have to note, in the first place, the circumstance--relevant although not decisive in itself--that Sûtra 12 does not contain any indication of a new topic being introduced. In the second place, it can hardly be doubted that the text of Sûtra 13, 'spashto hy ekeshâm,' is more appropriately understood, with Râmânuga, as furnishing a reason for the opinion advanced in the preceding Sûtra, than--with Sankara--as embodying the refutation of a previous statement (in which latter case we should expect not 'hi' but 'tu'). And, in the third place, the 'eke,' i.e. 'some,' referred to in Sûtra 13 would, on Sankara's interpretation, denote the very same persons to whom the preceding Sûtra had referred, viz. the followers of the Kânva-sâkhâ (the two Vedic passages referred to in 12 and 13 being Bri. Up. IV, 4, 5, and III, 2, 11, according to the Kânva recension); while it is the standing practice of the Sûtras to introduce, by means of the designation 'eke,' members of Vedic sâkhâs, teachers, &c. other than those alluded to in the preceding Sûtras. With this practice Râmânuga's interpretation, on the other hand, fully agrees; for, according to him, the 'eke' are the Mâdhyandinas, whose reading in Bri. Up. IV, 4, 5, viz. 'tasmât,' clearly indicates that the 'tasya' in the corresponding passage of the Kânvas denotes the sârira, i.e. the gîva. I think it is not saying too much that Sankara's explanation, according to which the 'eke' would denote the very same Kânvas to whom the preceding Sûtra had referred--so that the Kânvas would be distinguished from themselves as it were--is altogether impossible.

The result of this closer consideration of the first set of

p. xc

[paragraph continues] Sûtras, alleged by Sankara to concern the owner of the higher knowledge of Brahman, entitles us to view with some distrust--Sankara's assertion that another set also--IV, 4, l-7--has to be detached from the general topic of the fourth adhyâya, and to be understood as depicting the condition of those who have obtained final absolute release. And the Sûtras themselves do not tend to weaken this preliminary want of confidence. In the first place their wording also gives no indication whatever of their having to be separated from what precedes as well as what follows. And, in the second place, the last Sûtra of the set (7) obliges Sankara to ascribe to his truly released souls qualities which clearly cannot belong to them; so that he finally is obliged to make the extraordinary statement that those qualities belong to them 'vyavahârâpekshayâ,' while yet the purport of the whole adhikarana is said to be the description of the truly released soul for which no vyavahâra exists! Very truly Sankara's commentator here remarks, 'atra kekin muhyanti akhandakinmâtraânân muktasyâânâbhâvât kuta âânikadharmayogah,' and the way in which thereupon he himself attempts to get over the difficulty certainly does not improve matters.

In connexion with the two passages discussed, we meet in the fourth adhyâya with another passage, which indeed has no direct bearing on the distinction of aparâ and parâ vidyâ, but may yet be shortly referred to in this place as another and altogether undoubted instance of Sankara's interpretations not always agreeing with the text of the Sûtras. The Sûtras 7-16 of the third pâda state the opinions of three different teachers on the question to which Brahman the soul of the vidvân repairs on death, or--according to Râmânuga--the worshippers of which Brahman repair to (the highest) Brahman. Râmânuga treats the views of Bâdari and Gaimini as two pûrvapakshas, and the opinion of Bâdarâyana--which is stated last--as the siddhânta. Sankara, on the other hand, detaching the Sûtras in which Bâdarâyana's view is set forth from the preceding part of the adhikarana (a proceeding which, although not

p. xci

plausible, yet cannot be said to be altogether illegitimate), maintains that Bâdari's view, which is expounded first, represents the siddhânta, while Gaimini's view, set forth subsequently, is to be considered a mere pûrvapaksha. This, of course, is altogether inadmissible, it being the invariable practice of the Vedânta-sûtras as well as the Pûrva Mîmâmsâ-sûtras to conclude the discussion of contested points with the statement of that view which is to be accepted as the authoritative one. This is so patent that Sankara feels himself called upon to defend his deviation from the general rule (Commentary on IV, 4, 13), without, however, bringing forward any arguments but such as are valid only if Sankara's system itself is already accepted.

The previous considerations leave us, I am inclined to think, no choice but to side with Râmânuga as to the general subject-matter of the fourth adhyâya of the Sûtras. We need not accept him as our guide in all particular interpretations, but we must acknowledge with him that the Sûtras of the fourth adhyâya describe the ultimate fate of one and the same vidvân, and do not afford any basis for the distinction of a higher and lower knowledge of Brahman in Sankara's sense.

If we have not to discriminate between a lower and a higher knowledge of Brahman, it follows that the distinction of a lower and a higher Brahman is likewise not valid. But this is not a point to be decided at once on the negative evidence of the fourth adhyâya, but regarding which the entire body of the Vedânta-sûtras has to be consulted. And intimately connected with this investigation--in fact, one with it from a certain point of view--is the question whether the Sûtras afford any evidence of their author having held the doctrine of Mâyâ, the principle of illusion, by the association with which the highest Brahman, in itself transcending all qualities, appears as the lower Brahman or Îsvara. That Râmânuga denies the distinction of the two Brahmans and the doctrine of Mâyâ we have seen above; we shall, however, in the subsequent investigation, pay less attention to his views and interpretations

p. xcii

than to the indications furnished by the Sûtras themselves.

Placing myself at the point of view of a Sankara, I am startled at the outset by the second Sûtra of the first adhyâya, which undertakes to give a definition of Brahman. 'Brahman is that whence the origination and so on (i.e. the sustentation and reabsorption) of this world proceed.' What, we must ask, is this Sûtra meant to define?--That Brahman, we are inclined to answer, whose cognition the first Sûtra declares to constitute the task of the entire Vedânta; that Brahman whose cognition is the only road to final release; that Brahman in fact which Sankara calls the highest.--But, here we must object to ourselves, the highest Brahman is not properly defined as that from which the world originates. In later Vedântic writings, whose authors were clearly conscious of the distinction of the higher absolute Brahman and the lower Brahman related to Mâyâ or the world, we meet with definitions of Brahman of an altogether different type. I need only remind the reader of the current definition of Brahman as sak-kid-ânanda, or, to mention one individual instance, refer to the introductory slokas of the Pañkadasî dilating on the samvid svayamprabhâ, the self-luminous principle of thought which in all time, past or future, neither starts into being nor perishes (P. D. I, 7). 'That from which the world proceeds' can by a Sankara be accepted only as a definition of Îsvara, of Brahman which by its association with Mâyâ is enabled to project the false appearance of this world, and it certainly is as improbable that the Sûtras should open with a definition of that inferior principle, from whose cognition there can accrue no permanent benefit, as, according to a remark made above, it is unlikely that they should conclude with a description of the state of those who know the lower Brahman only, and thus are debarred from obtaining true release. As soon, on the other hand, as we discard the idea of a twofold Brahman and conceive Brahman as one only, as the all-enfolding being which sometimes emits the world from its own substance and sometimes again retracts it into itself, ever remaining one in all its

p. xciii

various manifestations--a conception which need not by any means be modelled in all its details on the views of the Râmânugas--the definition of Brahman given in the second Sûtra becomes altogether unobjectionable.

We next enquire whether the impression left on the mind by the manner in which Bâdarâyana defines Brahman, viz. that he does not distinguish between an absolute Brahman and a Brahman associated with Mâyâ, is confirmed or weakened by any other parts of his work. The Sûtras being throughout far from direct in their enunciations, we shall have to look less to particular terms and turns of expression than to general lines of reasoning. What in this connexion seems specially worthy of being taken into account, is the style of argumentation employed by the Sûtrakâra against the Sânkhya doctrine, which maintains that the world has originated, not from an intelligent being, but from the non-intelligent pradhâna. The most important Sûtras relative to this point are to be met with in the first pâda of the second adhyâya. Those Sûtras are indeed almost unintelligible if taken by themselves, but the unanimity of the commentators as to their meaning enables us to use them as steps in our investigation. The sixth Sûtra of the pâda mentioned replies to the Sânkhya objection that the non-intelligent world cannot spring from an intelligent principle, by the remark that 'it is thus seen,' i.e. it is a matter of common observation that non-intelligent things are produced from beings endowed with intelligence; hair and nails, for instance, springing from animals, and certain insects from dung.--Now, an argumentation of this kind is altogether out of place from the point of view of the true Sânkara. According to the latter the non-intelligent world does not spring from Brahman in so far as the latter is intelligence, but in so far as it is associated with Mâyâ. Mâyâ is the upâdâna of the material world, and Mâyâ itself is of a non-intelligent nature, owing to which it is by so many Vedântic writers identified with the prakriti of the Sânkhyas. Similarly the illustrative instances, adduced under Sûtra 9 for the purpose of showing that effects when being reabsorbed into their causal substances

p. xciv

do not impart to the latter their own qualities, and that hence the material world also, when being refunded into Brahman, does not impart to it its own imperfections, are singularly inappropriate if viewed in connexion with the doctrine of Mâyâ, according to which the material world is no more in Brahman at the time of a pralaya than during the period of its subsistence. According to Sânkara the world is not merged in Brahman, but the special forms into which the upâdâna of the world, i.e. Mâyâ, had modified itself are merged in non-distinct Mâyâ, whose relation to Brahman is not changed thereby.--The illustration, again, given in Sûtra 24 of the mode in which Brahman, by means of its inherent power, transforms itself into the world without employing any extraneous instruments of action, 'kshîravad dhi,' 'as milk (of its own accord turns into curds),' would be strangely chosen indeed if meant to bring nearer to our understanding the mode in which Brahman projects the illusive appearance of the world; and also the analogous instance given in the Sûtra next following, 'as Gods and the like (create palaces, chariots, &c. by the mere power of their will)'--which refers to the real creation of real things--would hardly be in its place if meant to illustrate a theory which considers unreality to be the true character of the world. The mere cumulation of the two essentially heterogeneous illustrative instances (kshîravad dhi; devâdivat), moreover, seems to show that the writer who had recourse to them held no very definite theory as to the particular mode in which the world springs from Brahman, but was merely concerned to render plausible in some way or other that an intelligent being can give rise to what is non-intelligent without having recourse to any extraneous means 1.

That the Mâyâ doctrine was not present to the mind of the Sûtrakâra, further appears from the latter part of the fourth pâda of the first adhyâya, where it is shown that Brahman is not only the operative but also the material cause of the world. If anywhere, there would have been

p. xcv

the place to indicate, had such been the author's view, that Brahman is the material cause of the world through Mâyâ only, and that the world is unreal; but the Sûtras do not contain a single word to that effect. Sûtra 26, on the other hand, exhibits the significant term 'parinâmât;' Brahman produces the world by means of a modification of itself. It is well known that later on, when the terminology of the Vedânta became definitely settled, the term 'parinâmavâda' was used to denote that very theory to which the followers of Sankara are most violently opposed, viz. the doctrine according to which the world is not a mere vivarta, i.e. an illusory manifestation of Brahman, but the effect of Brahman undergoing a real change, may that change be conceived to take place in the way taught by Râmânuga or in some other manner.--With regard to the last-quoted Sûtra, as well as to those touched upon above, the commentators indeed maintain that whatever terms and modes of expression are apparently opposed to the vivartavâda are in reality reconcilable with it; to Sûtra 26, for instance, Govindânanda remarks that the term 'parinâma' only denotes an effect in general (kâryamâtra), without implying that the effect is real. But in cases of this nature we are fully entitled to use our own judgment, even if we were not compelled to do so by the fact that other commentators, such as Râmânuga, are satisfied to take 'parinâma' and similar terms in their generally received sense.

A further section treating of the nature of Brahman is met with in III, 2, 11 ff. It is, according to Sankara's view, of special importance, as it is alleged to set forth that Brahman is in itself destitute of all qualities, and is affected with qualities only through its limiting adjuncts (upâdhis), the offspring of Mâyâ. I have above (in the conspectus of contents) given a somewhat detailed abstract of the whole section as interpreted by Sankara on the one hand, and Râmânuga on the other hand, from which it appears that the latter's opinion as to the purport of the group of Sûtras widely diverges from that of Sankara. The wording of the Sûtras is so eminently concise and vague that I find it impossible to decide which of the two commentators--if

p. xcvi

indeed either--is to be accepted as a trustworthy guide; regarding the sense of some Sûtras Sankara's explanation seems to deserve preference, in the case of others Râmânuga seems to keep closer to the text. I decidedly prefer, for instance, Râmânuga's interpretation of Sûtra 22, as far as the sense of the entire Sûtra is concerned, and more especially with regard to the term 'prakritaitâvattvam,' whose proper force is brought out by Râmânuga's explanation only. So much is certain that none of the Sûtras decidedly favours the interpretation proposed by Sankara. Whichever commentator we follow, we greatly miss coherence and strictness of reasoning, and it is thus by no means improbable that the section is one of those--perhaps not few in number--in which both interpreters had less regard to the literal sense of the words and to tradition than to their desire of forcing Bâdarâyana's Sûtras to bear testimony to the truth of their own philosophic theories.

With special reference to the Mâyâ doctrine one important Sûtra has yet to be considered, the only one in which the term 'mâyâ' itself occurs, viz. III, 2, 3. According to Sankara the Sûtra signifies that the environments of the dreaming soul are not real but mere Mâyâ, i.e. unsubstantial illusion, because they do not fully manifest the character of real objects. Râmânuga (as we have seen in the conspectus) gives a different explanation of the term 'mâyâ,' but in judging of Sankara's views we may for the time accept Sankara's own interpretation. Now, from the latter it clearly follows that if the objects seen in dreams are to be called Mâyâ, i.e. illusion, because not evincing the characteristics of reality, the objective world surrounding the waking soul must not be called Mâyâ. But that the world perceived by waking men is Mâyâ, even in a higher sense than the world presented to the dreaming consciousness, is an undoubted tenet of the Sânkara Vedânta; and the Sûtra therefore proves either that Bâdarâyana did not hold the doctrine of the illusory character of the world, or else that, if after all he did hold that doctrine, he used the term 'mâyâ' in a sense altogether different from that

p. xcvii

in which Sankara employs it.--If, on the other hand, we, with Râmânuga, understand the word 'mâyâ' to denote a wonderful thing, the Sûtra of course has no bearing whatever on the doctrine of Mâyâ in its later technical sense.

We now turn to the question as to the relation of the individual soul to Brahman. Do the Sûtras indicate anywhere that their author held Sankara's doctrine, according to which the gîva is in reality identical with Brahman, and separated from it, as it were, only by a false surmise due to avidyâ, or do they rather favour the view that the souls, although they have sprung from Brahman, and constitute elements of its nature, yet enjoy a kind of individual existence apart from it? This question is in fact only another aspect of the Mâyâ question, but yet requires a short separate treatment.

In the conspectus I have given it as my opinion that the Sûtras in which the size of the individual soul is discussed can hardly be understood in Sankara's sense, and rather seem to favour the opinion, held among others by Râmânuga, that the soul is of minute size. We have further seen that Sûtra 18 of the third pâda of the second adhyâya, which describes the soul as 'a,' is more appropriately understood in the sense assigned to it by Râmânuga; and, again, that the Sûtras which treat of the soul being an agent, can be reconciled with Sankara's views only if supplemented in a way which their text does not appear to authorise.--We next have the important Sûtra II, 3, 43 in which the soul is distinctly said to be a part (amsa) of Brahman, and which, as we have already noticed, can be made to fall in with Sankara's views only if amsa is explained, altogether arbitrarily, by 'amsa iva,' while Râmânuga is able to take the Sûtra as it stands.--We also have already referred to Sûtra 50, 'âbhâsa eva ka,' which Sankara interprets as setting forth the so-called pratibimbavâda according to which the individual Self is merely a reflection of the highest Self. But almost every Sûtra--and Sûtra 50 forms no exception-being so obscurely expressed, that viewed by itself it admits of various, often totally opposed, interpretations, the only safe method is to keep in view, in the case of each ambiguous

p. xcviii

aphorism, the general drift and spirit of the whole work, and that, as we have seen hitherto, is by no means favourable to the pratibimba doctrine. How indeed could Sûtra 50, if setting forth that latter doctrine, be reconciled with Sûtra 43, which says distinctly that the soul is a part of Brahman? For that 43 contains, as Sankara and his commentators aver, a statement of the avakkhedavâda, can itself be accepted only if we interpret amsa by amsa iva, and to do so there is really no valid reason whatever. I confess that Râmânuga's interpretation of the Sûtra (which however is accepted by several other commentators also) does not appear to me particularly convincing; and the Sûtras unfortunately offer us no other passages on the ground of which we might settle the meaning to be ascribed to the term âbhâsa, which may mean 'reflection,' but may mean hetvâbhâsa, i. e. fallacious argument, as well. But as things stand, this one Sûtra cannot, at any rate, be appealed to as proving that the pratibimbavâda which, in its turn, presupposes the mâyâvâda, is the teaching of the Sûtras.

To the conclusion that the Sûtrakâra did not hold the doctrine of the absolute identity of the highest and the individual soul in the sense of Sankara, we are further led by some other indications to be met with here and there in the Sûtras. In the conspectus of contents we have had occasion to direct attention to the important Sûtra II, 1, 22, which distinctly enunciates that the Lord is adhika, i. e. additional to, or different from, the individual soul, since Scripture declares the two to be different. Analogously I, 2, 20 lays stress on the fact that the sârîra is not the antaryâmin, because the Mâdhyandinas, as well as the Kânvas, speak of him in their texts as different (bhedena enam adhîyate), and in 22 the sârîra and the pradhâna are referred to as the two 'others' (itarau) of whom the text predicates distinctive attributes separating them from the highest Lord. The word 'itara' (the other one) appears in several other passages (I, 1, 16; I, 3, 16; II, 1, 21) as a kind of technical term denoting the individual soul in contradistinction from the Lord. The Sânkaras indeed maintain that all those passages refer to an unreal distinction

p. xciv

due to avidyâ. But this is just what we should like to see proved, and the proof offered in no case amounts to more than a reference to the system which demands that the Sûtras should be thus understood. If we accept the interpretations of the school of Sankara, it remains altogether unintelligible why the Sûtrakâra should never hint even at what Sankara is anxious again and again to point out at length, viz. that the greater part of the work contains a kind of exoteric doctrine only, ever tending to mislead the student who does not keep in view what its nature is. If other reasons should make it probable that the Sûtrakâra was anxious to hide the true doctrine of the Upanishads as a sort of esoteric teaching, we might be more ready to accept Sankara's mode of interpretation. But no such reasons are forthcoming; nowhere among the avowed followers of the Sankara system is there any tendency to treat the kernel of their philosophy as something to be jealously guarded and hidden. On the contrary, they all, from Gaudapâda down to the most modern writer, consider it their most important, nay, only task to inculcate again and again in the clearest and most unambiguous language that all appearance of multiplicity is a vain illusion, that the Lord and the individual souls are in reality one, and that all knowledge but this one knowledge is without true value.

There remains one more important passage concerning the relation of the individual soul to the highest Self, a passage which attracted our attention above, when we were reviewing the evidence for early divergence of opinion among the teachers of the Vedânta. I mean I, 4, 20-22, which three Sûtras state the views of Âsmarathya, Audulomi, and Kâsakrritsna as to the reason why, in a certain passage of the Brihadâranyaka, characteristics of the individual soul are ascribed to the highest Self. The siddhânta view is enounced in Sûtra 22, 'avasthiter iti Kâsakritsnah' i. e. Kâsakritsna (accounts for the circumstance mentioned) on the ground of the 'permanent abiding or abode.' By this 'permanent abiding' Sankara understands the Lord's abiding as, i. e. existing as--or in the condition of--the individual soul, and thus sees in the Sûtra an enunciation

p. c

of his own view that the individual soul is nothing but the highest Self, 'avikritah paramesvaro gîvo nânyah.' Râmânuga on the other hand, likewise accepting Kâsakritsna's opinion as the siddhânta view, explains 'avasthiti' as the Lord's permanent abiding within the individual soul, as described in the antaryâmin-brâhmana.--We can hardly maintain that the term 'avasthiti' cannot have the meaning ascribed to it by Sankara, viz. special state or condition, but so much must be urged in favour of Râmânuga's interpretation that in the five other places where avasthiti (or anavasthiti) is met with in the Sûtras (I, 2, 17; II, 2, 4; II, 2, 13; II, 3, 24; III, 3, 32) it regularly means permanent abiding or permanent abode within something.

If, now, I am shortly to sum up the results of the preceding enquiry as to the teaching of the Sûtras, I must, give it as my opinion that they do not set forth the distinction of a higher and lower knowledge of Brahman; that they do not acknowledge the distinction of Brahman and Îsvara in Sankara's sense; that they do not hold the doctrine of the unreality of the world; and that they do not, with Sankara, proclaim the absolute identity of the individual and the highest Self. I do not wish to advance for the present beyond these negative results. Upon Râmânuga's mode of interpretation--although I accept it without reserve in some important details--I look on the whole as more useful in providing us with a powerful means of criticising Sankara's explanations than in guiding us throughout to the right understanding of the text. The author of the Sûtras may have held views about the nature of Brahman, the world, and the soul differing from those of Sankara, and yet not agreeing in all points with those of Râmânuga. If, however, the negative conclusions stated above should be well founded, it would follow even from them that the system of Bâdarâyana had greater affinities with that of the Bhâgavatas and Râmânuga than with the one of which the Sankara-bhâshya is the classical exponent.

It appears from the above review of the teaching of the Sûtras that only a comparatively very small proportion of them contribute matter enabling us to form a judgment

p. ci

as to the nature of the philosophical doctrine advocated by Bâdarâyana. The reason of this is that the greater part of the work is taken up with matters which, according to Sankara's terminology, form part of the so-called lower knowledge, and throw no light upon philosophical questions in the stricter sense of the word. This circumstance is not without significance. In later works belonging to Sankara's school in which the distinction of a higher and lower vidyâ is clearly recognised, the topics constituting the latter are treated with great shortness; and rightly so, for they are unable to accomplish the highest aim of man, i.e. final release. When we therefore, on the other hand, find that the subjects of the so-called lower vidyâ are treated very fully in the Vedânta-sûtras, when we observe, for instance, the almost tedious length to which the investigation of the unity of vidyâs (most of which are so-called saguna, i.e. lower vidyâs) is carried in the third adhyâya, or the fact of almost the whole fourth adhyâya being devoted to the ultimate fate of the possessor of the lower vidyâ; we certainly feel ourselves confirmed in our conclusion that what Sankara looked upon as comparatively unimportant formed in Bâdarâyana's opinion part of that knowledge higher than which there is none, and which therefore is entitled to the fullest and most detailed exposition.

The question as to what kind of system is represented by the Vedânta-sûtras may be approached in another way also. While hitherto we have attempted to penetrate to the meaning of the Sûtras by means of the different commentaries, we might try the opposite road, and, in the first place, attempt to ascertain independently of the Sûtras what doctrine is set forth in the Upanishads, whose teaching the Sûtras doubtless aim at systematising. If, it might be urged, the Upanishads can be convincingly shown to embody a certain settled doctrine, we must consider it at the least highly probable that that very same doctrine--of whatever special nature it may be--is hidden in the enigmatical aphorisms of Bâdarâyan1.

I do not, however, consider this line of argumentation

p. cii

a safe one. Even if it could be shown that the teaching of all the chief Upanishads agrees in all essential points (a subject to which some attention will be paid later on), we should not on that account be entitled unhesitatingly to assume that the Sûtras set forth the same doctrine. Whatever the true philosophy of the Upanishads may be, there remains the undeniable fact that there exist and have existed since very ancient times not one but several essentially differing systems, all of which lay claim to the distinction of being the true representatives of the teaching of the Upanishads as well as of the Sûtras. Let us suppose, for argument's sake, that, for instance, the doctrine of Mâyâ is distinctly enunciated in the Upanishads; nevertheless Râmânuga and, for all we know to the contrary, the whole series of more ancient commentators on whom he looked as authorities in the interpretation of the Sûtras, denied that the Upanishads teach Mâyâ, and it is hence by no means impossible that Bâdarâyana should have done the same. The à priori style of reasoning as to the teaching of the Sûtras is therefore without much force.

But apart from any intention of arriving thereby at the meaning of the Sûtras there, of course, remains for us the all-important question as to the true teaching of the Upanishads, a question which a translator of the Sûtras and Sankara cannot afford to pass over in silence, especially after reason has been shown for the conclusion that the Sûtras and the Sankara-bhâshya do not agree concerning most important points of Vedântic doctrine. The Sûtras as well as the later commentaries claim, in the first place, to be nothing more than systematisations of the Upanishads, and for us a considerable part at least of their value and interest lies in this their nature. Hence the further question presents itself by whom the teaching of the Upanishads has been most adequately systematised, whether by Bâdarâyana, or Sankara, or Râmânuga, or some other commentator. This question requires to be kept altogether separate from the enquiry as to which commentator most faithfully renders the contents of the Sûtras, and it is by no means impossible that Sankara, for instance, should in

p. ciii

the end have to be declared a more trustworthy guide with regard to the teaching of the Upanishads than concerning the meaning of the Sûtras.

We must remark here at once that, whatever commentator may be found to deserve preference on the whole, it appears fairly certain already at the outset that none of the systems which Indian ingenuity has succeeded in erecting on the basis of the Upanishads can be accepted in its entirety. The reason for this lies in the nature of the Upanishads themselves. To the Hindu commentator and philosopher the Upanishads came down as a body of revealed truth whose teaching had, somehow or other, to be shown to be thoroughly consistent and free from contradictions; a system had to be devised in which a suitable place could be allotted to every one of the multitudinous statements which they make on the various points of Vedântic doctrine. But to the European scholar, or in fact to any one whose mind is not bound by the doctrine of Sruti, it will certainly appear that all such attempts stand self-condemned. If anything is evident even on a cursory review of the Upanishads--and the impression so created is only strengthened by a more careful investigation--it is that they do not constitute a systematic whole. They themselves, especially the older ones, give the most unmistakable indications on that point. Not only are the doctrines expounded in the different Upanishads ascribed to different teachers, but even the separate sections of one and the same Upanishad are assigned to different authorities. It would be superfluous to quote examples of what a mere look at the Khândogya Upanishad, for instance, suffices to prove. It is of course not impossible that even a multitude of teachers should agree in imparting precisely the same doctrine; but in the case of the Upanishads that is certainly not antecedently probable. For, in the first place, the teachers who are credited with the doctrines of the Upanishads manifestly belonged to different sections of Brahminical society, to different Vedic sâkhâs; nay, some of them the tradition makes out to have been kshattriyas. And, in the second place, the period, whose

p. civ

mental activity is represented in the Upanishads, was a creative one, and as such cannot be judged according to the analogy of later periods of Indian philosophic development. The later philosophic schools as, for instance, the one of which Sankara is the great representative, were no longer free in their speculations, but strictly bound by a traditional body of texts considered sacred, which could not be changed or added to, but merely systematised and commented upon. Hence the rigorous uniformity of doctrine characteristic of those schools. But there had been a time when, what later writers received as a sacred legacy, determining and confining the whole course of their speculations, first sprang from the minds of creative thinkers not fettered by the tradition of any school, but freely following the promptings of their own heads and hearts. By the absence of school traditions, I do not indeed mean that the great teachers who appear in the Upanishads were free to make an entirely new start, and to assign to their speculations any direction they chose; for nothing can be more certain than that, at the period as the outcome of whose philosophical activity the Upanishads have to be considered, there were in circulation certain broad speculative ideas overshadowing the mind of every member of Brahminical society. But those ideas were neither very definite nor worked out in detail, and hence allowed themselves to be handled and fashioned in different ways by different individuals. With whom the few leading conceptions traceable in the teaching of all Upanishads first originated, is a point on which those writings themselves do not enlighten us, and which we have no other means for settling; most probably they are to be viewed not as the creation of any individual mind, but as the gradual outcome of speculations carried on by generations of Vedic theologians. In the Upanishads themselves, at any rate, they appear as floating mental possessions which may be seized and moulded into new forms by any one who feels within himself the required inspiration. A certain vague knowledge of Brahman, the great hidden being in which all this manifold world is one, seems to be

p. cv

spread everywhere, and often issues from the most unexpected sources. Svetaketu receives instruction from his father Uddâlaka; the proud Gârgya has to become the pupil of Agâtasatru, the king of Kâsî; Bhugyu Sâhyâyani receives answers to his questions from a Gandharva possessing a maiden; Satyakâma learns what Brahman is from the bull of the herd he is tending, from Agni and from a flamingo; and Upakosala is taught by the sacred fires in his teacher's house. All this is of course legend, not history; but the fact that the philosophic and theological doctrines of the Upanishads are clothed in this legendary garb certainly does not strengthen the expectation of finding in them a rigidly systematic doctrine.

And a closer investigation of the contents of the Upanishads amply confirms this preliminary impression. If we avail ourselves, for instance, of M. Paul Régnaud's Matériaux pour servir à l'Histoire de la Philosophie de l'Inde, in which the philosophical lucubrations of the different Upanishads are arranged systematically according to topics, we can see with ease how, together with a certain uniformity of general leading conceptions, there runs throughout divergence in details, and very often not unimportant details. A look, for instance, at the collection of passages relative to the origination of the world from the primitive being, suffices to show that the task of demonstrating that whatever the Upanishads teach on that point can be made to fit into a homogeneous system is an altogether hopeless one. The accounts there given of the creation belong, beyond all doubt to different stages of philosophic and theological development or else to different sections of priestly society. None but an Indian commentator would, I suppose, be inclined and sufficiently courageous to attempt the proof that, for instance, the legend of the âtman purushavidha, the Self in the shape of a person which is as large as man and woman together, and then splits itself into two halves from which cows, horses, asses, goats, &c. are produced in succession (Bri. Up. I, 1, 4), can be reconciled with the account given of the creation in the Khândogya Upanishad, where it is said that in the beginning there existed nothing but the sat,

p. cvi

'that which is,' and that feeling a desire of being many it emitted out of itself ether, and then all the other elements in due succession. The former is a primitive cosmogonic myth, which in its details shows striking analogies with the cosmogonic myths of other nations; the latter account is fairly developed Vedânta (although not Vedânta implying the Mâyâ doctrine). We may admit that both accounts show a certain fundamental similarity in so far as they derive the manifold world from one original being; but to go beyond this and to maintain, as Sankara does, that the âtman purushavidha of the Brihadâranyaka is the so-called Virâg of the latter Vedânta--implying thereby that that section consciously aims at describing only the activity of one special form of Îsvara, and not simply the whole process of creation--is the ingenious shift of an orthodox commentator in difficulties, but nothing more.

How all those more or less conflicting texts came to be preserved and handed down to posterity, is not difficult to understand. As mentioned above, each of the great sections of Brahminical priesthood had its own sacred texts, and again in each of those sections there existed more ancient texts which it was impossible to discard when deeper and more advanced speculations began in their turn to be embodied in literary compositions, which in the course of time likewise came to be looked upon as sacred. When the creative period had reached its termination, and the task of collecting and arranging was taken in hand, older and newer pieces were combined into wholes, and thus there arose collections of such heterogeneous character as the Khândogya and Brihadâranyaka Upanishads. On later generations, to which the whole body of texts came down as revealed truth, there consequently devolved the inevitable task of establishing systems on which no exception could be taken to any of the texts; but that the task was, strictly speaking, an impossible one, i.e. one which it was impossible to accomplish fairly and honestly, there really is no reason to deny.

For a comprehensive criticism of the methods which the different commentators employ in systematizing the contents

p. cvii

of the Upanishads there is no room in this place. In order, however, to illustrate what is meant by the 'impossibility,' above alluded to, of combining the various doctrines of the Upanishads into a whole without doing violence to a certain number of texts, it will be as well to analyse in detail some few at least of Sankara's interpretations, and to render clear the considerations by which he is guided.

We begin with a case which has already engaged our attention when discussing the meaning of the Sûtras, viz. the question concerning the ultimate fate of those who have attained the knowledge of Brahman. As we have seen, Sankara teaches that the soul of him who has risen to an insight into the nature of the higher Brahman does not, at the moment of death, pass out of the body, but is directly merged in Brahman by a process from which all departing and moving, in fact all considerations of space, are altogether excluded. The soul of him, on the other hand, who has not risen above the knowledge of the lower qualified Brahman departs from the body by means of the artery called sushumnâ, and following the so-called devayâna, the path of the gods, mounts up to the world of Brahman. A review of the chief Upanishad texts on which Sankara founds this distinction will show how far it is justified.

In a considerable number of passages the Upanishads contrast the fate of two classes of men, viz. of those who perform sacrifices and meritorious works only, and of those who in addition possess a certain kind of knowledge. Men of the former kind ascend after death to the moon, where they live for a certain time, and then return to the earth into new forms of embodiment; persons of the latter kind proceed on the path of the gods--on which the sun forms one stage--up to the world of Brahman, from which there is no return. The chief passages to that effect are Kh. Up. V, 10; Kaush. Up. I, 2 ff.; Mund. Up. I, 2, 9 ff.; Bri. Up. VI, 2, 15 ff.; Prasna Up. I, 9 ff.--In other passages only the latter of the two paths is referred to, cp. Kh. Up. IV, 15; VIII 6, 5; Taitt. Up. I, 6; Bri. Up. IV, 4, 8, 9; V, 10; Maitr. Up. VI, 30, to mention only the more important ones.

Now an impartial consideration of those passages shows

p. cviii

[paragraph continues] I think, beyond any doubt, that what is meant there by the knowledge which leads through the sun to the world of Brahman is the highest knowledge of which the devotee is capable, and that the world of Brahman to which his knowledge enables him to proceed denotes the highest state which he can ever reach, the state of final release, if we choose to call it by that name.--Kh. Up. V, 10 says, 'Those who know this (viz. the doctrine of the five fires), and those who in the forest follow faith and austerities go to light,' &c.--Kh. Up. IV, 15 is manifestly intended to convey the true knowledge of Brahman; Upakosala's teacher himself represents the instruction given by him as superior to the teaching of the sacred fires.--Kh. Up. VIII, 6, 5 quotes the old sloka which says that the man moving upwards by the artery penetrating the crown of the head reaches the Immortal.--Kaush. Up. I, 2--which gives the most detailed account of the ascent of the soul--contains no intimation whatever of the knowledge of Brahman, which leads up to the Brahman world, being of an inferior nature.--Mund. Up. I, 2, 9 agrees with the Khândogya in saying that 'Those who practise penance and faith in the forest, tranquil, wise, and living on alms, depart free from passion, through the sun, to where that immortal Person dwells whose nature is imperishable,' and nothing whatever in the context countenances the assumption that not the highest knowledge and the highest Person are there referred to.--Bri. Up. IV, 4, 8 quotes old slokas clearly referring to the road of the gods ('the small old path'), on which 'sages who know Brahman move on to the svargaloka and thence higher on as entirely free.--That path was found by Brahman, and on it goes whoever knows Brahman.'--Bri. Up. VI, 2, 15 is another version of the Pañkâgnividyâ, with the variation, 'Those who know this, and those who in the forest worship faith and the True, go to light,' &c.--Prasna Up. 1, 10 says, 'Those who have sought the Self by penance, abstinence, faith, and knowledge gain by the northern path Âditya, the sun. There is the home of the spirits, the immortal free from danger, the highest. From thence they do not return, for it is the end.'--Maitr. Up. VI, 30 quotes

p. cix

slokas, 'One of them (the arteries) leads upwards, piercing the solar orb: by it, having stepped beyond the world of Brahman, they go to the highest path.'

All these passages are as clear as can be desired. The soul of the sage who knows Brahman passes out by the sushumnâ, and ascends by the path of the gods to the world of Brahman, there to remain for ever in some blissful state. But, according to Sankara, all these texts are meant to set forth the result of a certain inferior knowledge only, of the knowledge of the conditioned Brahman. Even in a passage apparently so entirely incapable of more than one interpretation as Bri. Up. VI, 2, 15, the 'True,' which the holy hermits in the forest are said to worship, is not to be the highest Brahman, but only Hiranyagarbha!--And why?--Only because the system so demands it, the system which teaches that those who know the highest Brahman become on their death one with it, without having to resort to any other place. The passage on which this latter tenet is chiefly based is Bri. Up. IV, 4, 6, 7, where, with the fate of him who at his death has desires, and whose soul therefore enters a new body after having departed from the old one, accompanied by all the prânas, there is contrasted the fate of the sage free from all desires. 'But as to the man who does not desire, who not desiring, freed from desires is satisfied in his desires, or desires the Self only, the vital spirits of him (tasya) do not depart--being Brahman he goes to Brahman.'

We have seen above (p. lxxx) that this passage is referred to in the important Sûtras on whose right interpretation it, in the first place, depends whether or not we must admit the Sûtrakâra to have acknowledged the distinction of a parâ and an aparâ vidyâ. Here the passage interests us as throwing light on the way in which Sankara systematises. He looks on the preceding part of the chapter as describing what happens to the souls of all those who do not know the highest Brahman, inclusive of those who know the lower Brahman only. They pass out of the old bodies followed by all prânas and enter new bodies. He, on the other hand, section 6 continues, who knows the true Brahman, does not pass out of the body, but becomes one with Brahman then

p. cx

and there. This interpretation of the purport of the entire chapter is not impossibly right, although I am rather inclined to think that the chapter aims at setting forth in its earlier part the future of him who does not know Brahman at all, while the latter part of section 6 passes on to him who does know Brahman (i.e. Brahman pure and simple, the text knowing of no distinction of the so-called lower and higher Brahman). In explaining section 6 Sankara lays stress upon the clause 'na tasya prâna utkrâmanti,' 'his vital spirits do not pass out,' taking this to signify that the soul with the vital spirits does not move at all, and thus does not ascend to the world of Brahman; while the purport of the clause may simply be that the soul and vital spirits do not go anywhere else, i.e. do not enter a new body, but are united, somehow or other, with Brahman. On Sankara's interpretation there immediately arises a new difficulty. In the slokas, quoted under sections 8 and 9, the description of the small old path which leads to the svargaloka and higher on clearly refers--as noticed already above--to the path through the veins, primarily the sushumnâ, on which, according to so many other passages, the soul of the wise mounts upwards. But that path is, according to Sankara, followed by him only who has not risen above the lower knowledge, and yet the slokas have manifestly to be connected with what is said in the latter half of 6 about the owner of the parâ vidyâ. Hence Sankara sees himself driven to explain the slokas in 8 and 9 (of which a faithful translation is given in Professor Max Müller's version) as follows:

8. 'The subtle old path (i.e. the path of knowledge on which final release is reached; which path is subtle, i.e. difficult to know, and old, i.e. to be known from the eternal Veda) has been obtained and fully reached by me. On it the sages who know Brahman reach final release (svargalokasabdah samnihitaprakaranât mokshâbhidhâyakah).

9. 'On that path they say that there is white or blue or yellow or green or red (i.e. others maintain that the path to final release is, in accordance with the colour of the arteries, either white or blue, &c.; but that is false, for the

p. cxi

paths through the arteries lead at the best to the world of Brahman, which itself forms part of the samsâra); that path (i. e. the only path to release, viz. the path of true knowledge) is found by Brahman, i. e. by such Brâhmanas as through true knowledge have become like Brahman,' &c.

A significant instance in truth of the straits to which thorough-going systematisers of the Upanishads see themselves reduced occasionally!

But we return to the point which just now chiefly interests us. Whether Sankara's interpretation of the chapter, and especially of section 6, be right or wrong, so much is certain that we are not entitled to view all those texts which speak of the soul going to the world of Brahman as belonging to the so-called lower knowledge, because a few other passages declare that the sage does not go to Brahman. The text which declares the sage free from desires to become one with Brahman could not, without due discrimination, be used to define and limit the meaning of other passages met with in the same Upanishad even--for as we have remarked above the Brihadâranyaka contains pieces manifestly belonging to different stages of development;--much less does it entitle us to put arbitrary constructions on passages forming part of other Upanishads. Historically the disagreement of the various accounts is easy to understand. The older notion was that the soul of the wise man proceeds along the path of the gods to Brahman's abode. A later--and, if we like, more philosophic--conception is that, as Brahman already is a man's Self, there is no need of any motion on man's part to reach Brahman. We may even apply to those two views the terms aparâ and parâ--lower and higher--knowledge. But we must not allow any commentator to induce us to believe that what he from his advanced standpoint looks upon as an inferior kind of cognition, was viewed in the same light by the authors of the Upanishads.

We turn to another Upanishad text likewise touching upon the point considered in what precedes, viz. the second Brâhmana of the third adhyâya of the Brihadâranyaka. The discussion there first turns upon the grahas and atigrahas,

p. cxii

i. e. the senses and organs and their objects, and Yâgñavalkya thereupon explains that death, by which everything is overcome, is itself overcome by water; for death is fire. The colloquy then turns to what we must consider an altogether new topic, Ârtabhâga asking, 'When this man (ayam purusha) dies, do the vital spirits depart from him or not?' and Yâgñavalkya answering, 'No, they are gathered up in him; he swells, he is inflated; inflated the dead (body) is lying.'--Now this is for Sankara an important passage, as we have already seen above (p. lxxxi); for he employs it, in his comment on Ved.-sûtra IV, 2, 13, for the purpose of proving that the passage Bri. Up. IV, 4, 6 really means that the vital spirits do not, at the moment of death, depart from the true sage. Hence the present passage also must refer to him who possesses the highest knowledge; hence the 'ayam purusha' must be 'that man,' i. e. the man who possesses the highest knowledge, and the highest knowledge then must be found in the preceding clause which says that death itself may be conquered by water. But, as Râmânuga also remarks, neither does the context favour the assumption that the highest knowledge is referred to, nor do the words of section 11 contain any indication that what is meant is the merging of the Self of the true Sage in Brahman. With the interpretation given by Râmânuga himself, viz. that the prânas do not depart from the gîva of the dying man, but accompany it into a new body, I can agree as little (although he no doubt rightly explains the 'ayam purusha' by 'man' in general), and am unable to see in the passage anything more than a crude attempt to account for the fact that a dead body appears swollen and inflated.--A little further on (section 13) Ârtabhâga asks what becomes of this man (ayam purusha) when his speech has entered into the fire, his breath into the air, his eye into the sun, &c. So much here is clear that we have no right to understand by the 'ayam purusha' of section 13 anybody different from the 'ayam purusha' of the two preceding sections; in spite of this Sankara--according to whose system the organs of the true sage do not enter into the elements, but are directly

p. cxiii

merged in Brahman--explains the 'ayam purusha' of section 13 to be the 'asamyagdarsin,' i.e. the person who has not risen to the cognition of the highest Brahman. And still a further limiting interpretation is required by the system. The asamyagdarsin also who as such has to remain in the samsâra-cannot do without the organs, since his gîva when passing out of the old body into a new one is invested with the subtle body; hence section 13 cannot be taken as saying what it clearly does say, viz. that at death the different organs pass into the different elements, but as merely indicating that the organs are abandoned by the divinities which, during lifetime, presided over them!

The whole third adhyâya indeed of the Brihadâranyaka affords ample proof of the artificial character of Sankara's attempts to show that the teaching of the Upanishads follows a definite system. The eighth brâhmana, for instance, is said to convey the doctrine of the highest non-related Brahman, while the preceding brâhmanas had treated only of Îsvara in his various aspects. But, as a matter of fact, brâhmana 8, after having, in section 8, represented Brahman as destitute of all qualities, proceeds, in the next section, to describe that very same Brahman as the ruler of the world, 'By the command of that Imperishable sun and moon stand apart,' &c.; a clear indication that the author of the Upanishad does not distinguish a higher and lower Brahman in--Sankara's sense.--The preceding brâhmana (7) treats of the antaryâmin, i.e. Brahman viewed as the internal ruler of everything. This, according to Sankara, is the lower form of Brahman called Îsvara; but we observe that the antaryâmin as well as the so-called highest Brahman described in section 8 is, at the termination of the two sections, characterised by means of the very same terms (7, 23: Unseen but seeing, unheard but hearing, &c. There is no other seer but he, there is no other hearer but he, &c.; and 8, 11: That Brahman is unseen but seeing, unheard but hearing, &c. There is nothing that sees but it, nothing that hears but it, &c.).--Nothing can be clearer than that all these sections aim at describing one and the same being, and know nothing of the distinctions made by the developed

p. cxiv

[paragraph continues] Vedânta, however valid the latter may be from a purely philosophic point of view.

We may refer to one more similar instance from the Khândogya Upanishad. We there meet in III, 14 with one of the most famous vidyâs describing the nature of Brahman, called after its reputed author the Sândilya-vidyâ. This small vidyâ is decidedly one of the finest and most characteristic texts; it would be difficult to point out another passage setting forth with greater force and eloquence and in an equally short compass the central doctrine of the Upanishads. Yet this text, which, beyond doubt, gives utterance to the highest conception of Brahman's nature that Sândilya's thought was able to reach, is by Sankara and his school again declared to form part of the lower vidyâ only, because it represents Brahman as possessing qualities. It is, according to their terminology, not âna, i. e. knowledge, but the injunction of a mere upâsanâ, a devout meditation on Brahman in so far as possessing certain definite attributes such as having light for its form, having true thoughts, and so on. The Râmânugas, on the other hand, quote this text with preference as clearly describing the nature of their highest, i. e. their one Brahman. We again allow that Sankara is free to deny that any text which ascribes qualities to Brahman embodies absolute truth; but we also again remark that there is no reason whatever for supposing that Sândilya, or whoever may have been the author of that vidyâ, looked upon it as anything else but a statement of the highest truth accessible to man.

We return to the question as to the true philosophy of the Upanishads, apart from the systems of the commentators.--From what precedes it will appear with sufficient distinctness that, if we understand by philosophy a philosophical system coherent in all its parts, free from all contradictions and allowing room for all the different statements made in all the chief Upanishads, a philosophy of the Upanishads cannot even be spoken of. The various lucubrations on Brahman, the world, and the human soul of which the Upanishads consist do not allow themselves to be systematised simply because they were never meant to

p. cxv

form a system. Sândilya's views as to the nature of Brahman did not in all details agree with those of Yâavalkya, and Uddâlaka differed from both. In this there is nothing to wonder at, and the burden of proof rests altogether with those who maintain that a large number of detached philosophic and theological dissertations, ascribed to different authors, doubtless belonging to different periods, and not seldom manifestly contradicting each other, admit of being combined into a perfectly consistent whole.

The question, however, assumes a different aspect, if we take the terms 'philosophy' and 'philosophical system,' not in the strict sense in which Sankara and other commentators are not afraid of taking them, but as implying merely an agreement in certain fundamental features. In this latter sense we may indeed undertake to indicate the outlines of a philosophy of the Upanishads, only keeping in view that precision in details is not to be aimed at. And here we finally see ourselves driven back altogether on the texts themselves, and have to acknowledge that the help we receive from commentators, to whatever school they may belong, is very inconsiderable. Fortunately it cannot be asserted that the texts on the whole oppose very serious difficulties to a right understanding, however obscure the details often are. Concerning the latter we occasionally depend entirely on the explanations vouchsafed by the scholiasts, but as far as the general drift and spirit of the texts are concerned, we are quite able to judge by ourselves, and are even specially qualified to do so by having no particular system to advocate.

The point we will first touch upon is the same from which we started when examining the doctrine of the Sûtras, viz. the question whether the Upanishads acknowledge a higher and lower knowledge in Sankara's sense, i. e. a knowledge of a higher and a lower Brahman. Now this we find not to be the case. Knowledge is in the Upanishads frequently opposed to avidyâ, by which latter term we have to understand ignorance as to Brahman, absence of philosophic knowledge; and, again, in several places we find the knowledge of the sacrificial part of the Veda with its supplementary

p. cxvi

disciplines contrasted as inferior with the knowledge of the Self; to which latter distinction the Mundaka Up. (I, 4) applies the terms aparâ and parâ vidyâ. But a formal recognition of the essential difference of Brahman being viewed, on the one hand, as possessing distinctive attributes, and, on the other hand, as devoid of all such attributes is not to be met with anywhere. Brahman is indeed sometimes described as saguna and sometimes as nirguna (to use later terms); but it is nowhere said that thereon rests a distinction of two different kinds of knowledge leading to altogether different results. The knowledge of Brahman is one, under whatever aspects it is viewed; hence the circumstance (already exemplified above) that in the same vidyâs it is spoken of as saguna as well as nirguna. When the mind of the writer dwells on the fact that Brahman is that from which all this world originates, and in which it rests, he naturally applies to it distinctive attributes pointing at its relation to the world; Brahman, then, is called the Self and life of all, the inward ruler, the omniscient Lord, and so on. When, on the other hand, the author follows out the idea that Brahman may be viewed in itself as the mysterious reality of which the whole expanse of the world is only an outward manifestation, then it strikes him that no idea or term derived from sensible experience can rightly be applied to it, that nothing more may be predicated of it but that it is neither this nor that. But these are only two aspects of the cognition of one and the same entity.

Closely connected with the question as to the double nature of the Brahman of the Upanishads is the question as to their teaching Mâyâ.--From Colebrooke downwards the majority of European writers have inclined towards the opinion that the doctrine of Mâyâ, i. e. of the unreal illusory character of the sensible world, does not constitute a feature of the primitive philosophy of the Upanishads, but was introduced into the system at some later period, whether by Bâdarâyana or Sankara or somebody else. The opposite view, viz. that the doctrine of Mâyâ forms an integral element of the teaching of the Upanishads, is implied in them everywhere, and enunciated more or less distinctly in

p. cxvii

more than one place, has in recent times been advocated with much force by Mr. Gough in the ninth chapter of his Philosophy of the Upanishads.

In his Matériaux, &c. M. Paul Régnaud remarks that 'the doctrine of Mâyâ, although implied in the teaching of the Upanishads, could hardly become clear and explicit before the system had reached a stage of development necessitating a choice between admitting two co-existent eternal principles (which became the basis of the Sânkhya philosophy), and accepting the predominance of the intellectual principle, which in the end necessarily led to the negation of the opposite principle.'--To the two alternatives here referred to as possible we, however, have to add a third one, viz. that form of the Vedânta of which the theory of the Bhâgavatas or Râmânugas is the most eminent type, and according to which Brahman carries within its own nature an element from which the material universe originates; an element which indeed is not an independent entity like the pradhâna of the Sânkhyas, but which at the same time is not an unreal Mâyâ but quite as real as any other part of Brahman's nature. That a doctrine of this character actually developed itself on the basis of the Upanishads, is a circumstance which we clearly must not lose sight of, when attempting to determine what the Upanishads themselves are teaching concerning the character of the world.

In enquiring whether the Upanishads maintain the Mâyâ doctrine or not, we must proceed with the same caution as regards other parts of the system, i.e. we must refrain from using unhesitatingly, and without careful consideration of the merits of each individual case, the teaching--direct or inferred--of any one passage to the end of determining the drift of the teaching of other passages. We may admit that some passages, notably of the Brihadâranyaka, contain at any rate the germ of the later developed Mâyâ doctrine 1, and thus render it quite intelligible that a system like Sankara's

p. cxviii

should evolve itself, among others, out of the Upanishads; but that affords no valid reason for interpreting Mâyâ into other texts which give a very satisfactory sense without that doctrine, or are even clearly repugnant to it. This remark applies in the very first place to all the accounts of the creation of the physical universe. There, if anywhere, the illusional character of the world should have been hinted at, at least, had that theory been held by the authors of those accounts; but not a word to that effect is met with anywhere. The most important of those accounts--the one given in the sixth chapter of the Khândogya Upanishad--forms no exception. There is absolutely no reason to assume that the 'sending forth' of the elements from the primitive Sat, which is there described at length, was by the writer of that passage meant to represent a vivarta rather than a parinâma that the process of the origination of the physical universe has to be conceived as anything else but a real manifestation of real powers hidden in the primeval Self. The introductory words, addressed to Svetaketu by Uddâlaka, which are generally appealed to as intimating the unreal character of the evolution about to be described, do not, if viewed impartially, intimate any such thing 1. For what is capable of being proved, and manifestly meant to be proved, by the illustrative instances of the lump of clay and the nugget of gold, through which there are known all things made of clay and gold? Merely that this whole world has Brahman for its causal substance, just as clay is the causal matter of every earthen pot, and gold of every golden ornament, but not that the process through which any causal substance becomes an effect is an unreal one. We--including Uddâlaka--may surely say that all earthen pots are in reality nothing but earth--the earthen pot being merely a special modification (vikâra) of clay which has a name of its own--without thereby committing ourselves to the doctrine that the change of form, which a lump of clay undergoes when being fashioned into a pot, is not real but a mere baseless illusion.

In the same light we have to view numerous other passages

p. cxix

which set forth the successive emanations proceeding from the first principle. When, for instance, we meet in the Katha Up. I, 3, 10, in the serial enumeration of the forms of existence intervening between the gross material world and the highest Self (the Person), with the 'avyâkrita,' the Undeveloped, immediately below the purusha; and when again the Mundaka Up. II, 1, 2, speaks of the 'high Imperishable' higher than which is the heavenly Person; there is no reason whatever to see in that 'Undeveloped' and that 'high Imperishable' anything but that real element in Brahman from which, as in the Râmânuga system, the material universe springs by a process of real development. We must of course render it quite clear to ourselves in what sense the terms 'real' and 'unreal' have to be understood. The Upanishads no doubt teach emphatically that the material world does not owe its existence to any principle independent from the Lord like the pradhâna of the Sânkhyas; the world is nothing but a manifestation of the Lord's wonderful power, and hence is unsubstantial, if we take the term 'substance' in its strict sense. And, again, everything material is immeasurably inferior in nature to the highest spiritual principle from which it has emanated, and which it now hides from the individual soul. But neither unsubstantiality nor inferiority of the kind mentioned constitutes unreality in the sense in which the Mâyâ of Sankara is unreal. According to the latter the whole world is nothing but an erroneous appearance, as unreal as the snake, for which a piece of rope is mistaken by the belated traveller, and disappearing just as the imagined snake does as soon as the light of true knowledge has risen. But this is certainly not the impression left on the mind by a comprehensive review of the Upanishads which dwells on their general scope, and does not confine itself to the undue urging of what may be implied in some detached passages. The Upanishads do not call upon us to look upon the whole world as a baseless illusion to be destroyed by knowledge; the great error which they admonish us to relinquish is rather that things have a separate individual existence, and are not tied together by the bond of being all of them effects

p. cxx

of Brahman, or Brahman itself. They do not say that true knowledge sublates this false world, as Sankara says, but that it enables the sage to extricate himself from the world--the inferior mûrta rûpa of Brahman, to use an expression of the Brihadâranyaka--and to become one with Brahman in its highest form. 'We are to see everything in Brahman, and Brahman in everything;' the natural meaning of this is, 'we are to look upon this whole world as a true manifestation of Brahman, as sprung from it and animated by it.' The mâyâvâdin has indeed appropriated the above saying also, and interpreted it so as to fall in with his theory; but he is able to do so only by perverting its manifest sense. For him it would be appropriate to say, not that everything we see is in Brahman, but rather that everything we see is out of Brahman, viz. as a false appearance spread over it and hiding it from us.

Stress has been laid 1 upon certain passages of the Brihadâranyaka which seem to hint at the unreality of this world by qualifying terms, indicative of duality or plurality of existence, by means of an added 'iva,' i.e. 'as it were' (yatrânyad iva syât; yatra dvaitam iva bhavati; âtmâ dhyâyatîva lelâyatîva). Those passages no doubt readily lend themselves to Mâyâ interpretations, and it is by no means impossible that in their author's mind there was something like an undeveloped Mâyâ doctrine. I must, however, remark that they, on the other hand, also admit of easy interpretations not in any way presupposing the theory of the unreality of the world. If Yâavalkya refers to the latter as that 'where there is something else as it were, where there is duality as it were,' he may simply mean to indicate that the ordinary opinion, according to which the individual forms of existence of the world are opposed to each other as altogether separate, is a mistaken one, all things being one in so far as they spring from--and are parts of--Brahman. This would in no way involve duality or plurality being unreal in Sankara's sense, not any more than, for instance, the modes of Spinoza are unreal because, according to that philosopher, there is only one universal

p. cxxi

substance. And with regard to the clause 'the Self thinks as it were' it has to be noted that according to the commentators the 'as it were' is meant to indicate that truly not the Self is thinking, but the upâdhis, i.e. especially the manas with which the Self is connected. But whether these upâdhis are the mere offspring of Mâyâ, as Sankara thinks, or real forms of existence, as Râmânuga teaches, is an altogether different question.

I do not wish, however, to urge these last observations, and am ready to admit that not impossibly those iva's indicate that the thought of the writer who employed them was darkly labouring with a conception akin to--although much less explicit than--the Mâyâ of Sankara. But what I object to is, that conclusions drawn from a few passages of, after all, doubtful import should be employed for introducing the Mâyâ doctrine into other passages which do not even hint at it, and are fully intelligible without it 1.

The last important point in the teaching of the Upanishads we have to touch upon is the relation of the gîvas, the individual souls to the highest Self. The special views regarding that point held by Sankara and Râmânuga, as have been stated before. Confronting their theories with the texts of the Upanishads we must, I think, admit without hesitation, that Sankara's doctrine faithfully represents the prevailing teaching of the Upanishads in one important point at least, viz. therein that the soul or Self of the sage--whatever its original relation to Brahman may be--is in the end completely merged and indistinguishably lost in the universal Self. A distinction, repeatedly alluded to before, has indeed to be kept in view here also. Certain texts of the Upanishads describe the soul's going upwards, on the path of the gods, to the world of Brahman, where it dwells for unnumbered years, i.e. for ever. Those texts, as a type of which we may take, the passage Kaushît. Up. I--the fundamental text of the Râmânuga's concerning the soul's

p. cxxii

fate after death--belong to an earlier stage of philosophic development; they manifestly ascribe to the soul a continued individual existence. But mixed with texts of this class there are others in which the final absolute identification of the individual Self with the universal Self is indicated in terms of unmistakable plainness. 'He who knows Brahman and becomes Brahman;' 'he who knows Brahman becomes all this;' 'as the flowing rivers disappear in the sea losing their name and form, thus a wise man goes to the divine person.' And if we look to the whole, to the prevailing spirit of the Upanishads, we may call the doctrine embodied in passages of the latter nature the doctrine of the Upanishads. It is, moreover, supported by the frequently and clearly stated theory of the individual souls being merged in Brahman in the state of deep dreamless sleep.

It is much more difficult to indicate the precise teaching of the Upanishads concerning the original relation of the individual soul to the highest Self, although there can be no doubt that it has to be viewed as proceeding from the latter, and somehow forming a part of it. Negatively we are entitled to say that the doctrine, according to which the soul is merely brahma bhrântam or brahma mâyopâdhikam, is in no way countenanced by the majority of the passages bearing on the question. If the emission of the elements, described in the Khândogya and referred to above, is a real process--of which we saw no reason to doubt--the gîva âtman with which the highest Self enters into the emitted elements is equally real, a true part or emanation of Brahman itself.

After having in this way shortly reviewed the chief elements of Vedântic doctrine according to the Upanishads, we may briefly consider Sankara's system and mode of interpretation--with whose details we had frequent opportunities of finding fault--as a whole. It has been said before that the task of reducing the teaching of the whole of the Upanishads to a system consistent and free from contradictions is an intrinsically impossible one. But the task once being given, we are quite ready to admit that Sankara's system is most probably the best which can be devised.

p. cxxiii

While unable to allow that the Upanishads recognise a lower and higher knowledge of Brahman, in fact the distinction of a lower and higher Brahman, we yet acknowledge that the adoption of that distinction furnishes the interpreter with an instrument of extraordinary power for reducing to an orderly whole the heterogeneous material presented by the old theosophic treatises. This becomes very manifest as soon as we compare Sankara's system with that of Râmânuga. The latter recognises only one Brahman which is, as we should say, a personal God, and he therefore lays stress on all those passages of the Upanishads which ascribe to Brahman the attributes of a personal God, such as omniscience and omnipotence. Those passages, on the other hand, whose decided tendency it is to represent Brahman as transcending all qualities, as one undifferenced mass of impersonal intelligence, Râmânuga is unable to accept frankly and fairly, and has to misinterpret them more or less to make them fall in with his system. The same remark holds good with regard to those texts which represent the individual soul as finally identifying itself with Brahman; Râmânuga cannot allow a complete identification but merely an assimilation carried as far as possible. Sankara, on the other hand, by skilfully ringing the changes on a higher and a lower doctrine, somehow manages to find room for whatever the Upanishads have to say. Where the text speaks of Brahman as transcending all attributes, the highest doctrine is set forth. Where Brahman is called the All-knowing ruler of the world, the author means to propound the lower knowledge of the Lord only. And where the legends about the primary being and its way of creating the world become somewhat crude and gross, Hiranyagarbha and Virâg are summoned forth and charged with the responsibility. Of Virâg Mr. Gough remarks (p. 55) that in him a place is provided by the poets of the Upanishads for the purusha of the ancient rishis, the divine being out of whom the visible and tangible world proceeded. This is quite true if only we substitute for the 'poets of the Upanishads' the framers of the orthodox Vedânta system--for the Upanishads give no indication whatever

p. cxxiv

that by their purusha they understand not the simple old purusha but the Virâg occupying a definite position in a highly elaborate system;--but the mere phrase, 'providing a place' intimates with sufficient clearness the nature of the work in which systematisers of the Vedântic doctrine are engaged.

[paragraph continues] Sankara's method thus enables him in a certain way to do justice to different stages of historical development, to recognise clearly existing differences which other systematisers are intent on obliterating. And there has yet to be made a further and even more important admission in favour of his system. It is not only more pliable, more capable of amalgamating heterogeneous material than other systems, but its fundamental doctrines are manifestly in greater harmony with the essential teaching of the Upanishads than those of other Vedântic systems. Above we were unable to allow that the distinction made by Sankara between Brahman and Îsvara is known to the Upanishads; but we must now admit that if, for the purpose of determining the nature of the highest being, a choice has to be made between those texts which represent Brahman as nirguna, and those which ascribe to it personal attributes, Sankara is right in giving preference to texts of the former kind. The Brahman of the old Upanishads, from which the souls spring to enjoy individual consciousness in their waking state, and into which they sink back temporarily in the state of deep dreamless sleep and permanently in death, is certainly not represented adequately by the strictly personal Îsvara of Râmânuga, who rules the world in wisdom and mercy. The older Upanishads, at any rate, lay very little stress upon personal attributes of their highest being, and hence Sankara is right in so far as he assigns to his hypostatised personal Îsvara 1 a lower place than to his absolute Brahman. That he also faithfully represents the prevailing spirit of the Upanishads in his theory of the ultimate fate

p. cxxv

of the soul, we have already remarked above. And although the Mâyâ doctrine cannot, in my opinion, be said to form part of the teaching of the Upanishads, it cannot yet be asserted to contradict it openly, because the very point which it is meant to elucidate, viz. the mode in which the physical universe and the multiplicity of individual souls originate, is left by the Upanishads very much in the dark. The later growth of the Mâyâ doctrine on the basis of the Upanishads is therefore quite intelligible, and I fully agree with Mr. Gough when he says regarding it that there has been no addition to the system from without but only a development from within, no graft but only growth. The lines of thought which finally led to the elaboration of the full-blown Mâyâ theory may be traced with considerable certainty. In the first place, deepening speculation on Brahman tended to the notion of advaita being taken in a more and more strict sense, as implying not only the exclusion of any second principle external to Brahman, but also the absence of any elements of duality or plurality in the nature of the one universal being itself; a tendency agreeing with the spirit of a certain set of texts from the Upanishads. And as the fact of the appearance of a manifold world cannot be denied, the only way open to thoroughly consistent speculation was to deny at any rate its reality, and to call it a mere illusion due to an unreal principle, with which Brahman is indeed associated, but which is unable to break the unity of Brahman's nature just on account of its own unreality. And, in the second place, a more thorough following out of the conception that the union with Brahman is to be reached through true knowledge only, not unnaturally led to the conclusion that what separates us in our unenlightened state from Brahman is such as to allow itself to be completely sublated by an act of knowledge; is, in other words, nothing else but an erroneous notion, an illusion.--A further circumstance which may not impossibly have co-operated to further the development of the theory of the world's unreality will be referred to later on. 1

p. cxxvi

We have above been obliged to leave it an open question what kind of Vedânta is represented by the Vedânta-sûtras, although reason was shown for the supposition that in some important points their teaching is more closely related to the system of Râmânuga than to that of Sankara. If so, the philosophy of Sankara would on the whole stand nearer to the teaching of the Upanishads than the Sûtras of Bâdarâyana. This would indeed be a somewhat unexpected conclusion--for, judging a priori, we should be more inclined to assume a direct propagation of the true doctrine of the Upanishads through Bâdarâyana to Sankara--but a priori considerations have of course no weight against positive evidence to the contrary. There are, moreover, other facts in the history of Indian philosophy and theology which help us better to appreciate the possibility of Bâdarâyana's Sûtras already setting forth a doctrine that lays greater stress on the personal character of the highest being than is in agreement with the prevailing tendency of the Upanishads. That the pure doctrine of those ancient Brahminical treatises underwent at a rather early period amalgamations with beliefs which most probably had sprung up in altogether different--priestly or non-priestly--communities is a well-known circumstance; it suffices for our purposes to refer to the most eminent of the early literary monuments in which an amalgamation of the kind mentioned is observable, viz. the Bhagavadgîtâ. The doctrine of the Bhagavadgîtâ represents a fusion of the Brahman theory of the Upanishads with the belief in a personal highest being--Krishna or Vishnu--which in many respects approximates very closely to the system of the Bhâgavatas; the attempts of a certain set of Indian commentators to explain it as setting forth pure Vedânta, i.e. the pure doctrine of the Upanishads, may simply be set aside. But this same Bhagavadgîtâ is quoted in Bâdarâyana's Sûtras (at least according to the unanimous explanations of the most eminent scholiasts of different schools) as inferior to Sruti only in authority. The Sûtras,

p. cxxvii

moreover, refer in different places to certain Vedântic portions of the Mahâbhârata, especially the twelfth book, several of which represent forms of Vedânta distinctly differing from Sankara's teaching, and closely related to the system of the Bhâgavatas.

Facts of this nature--from entering into the details of which we are prevented by want of space--tend to mitigate the primâ facie strangeness of the assumption that the Vedânta-sûtras, which occupy an intermediate position between the Upanishads and Sankara, should yet diverge in their teaching from both. The Vedânta of Gaudapâda and Sankara would in that case mark a strictly orthodox reaction against all combinations of non-Vedic elements of belief and doctrine with the teaching of the Upanishads. But although this form of doctrine has ever since Sankara's time been the one most generally accepted by Brahminic students of philosophy, it has never had any wide-reaching influence on the masses of India. It is too little in sympathy with the wants of the human heart, which, after all, are not so very different in India from what they are elsewhere. Comparatively few, even in India, are those who rejoice in the idea of a universal non-personal essence in which their own individuality is to be merged and lost for ever, who think it sweet 'to be wrecked on the ocean of the Infinite 1.' The only forms of Vedântic philosophy which are--and can at any time have been--really popular, are those in which the Brahman of the Upanishads has somehow transformed itself into a being, between which and the devotee there can exist a personal relation, love and faith on the part of man, justice tempered by mercy on the part of the divinity. The only religious books of widespread influence are such as the Râmâyan of Tulsidâs, which lay no stress on the distinction between an absolute Brahman inaccessible to all human wants and sympathies, and a shadowy Lord whose very conception depends on the illusory principle of Mâyâ, but love to dwell on the delights of devotion

p. cxxviii

to one all-wise and merciful ruler, who is able and willing to lend a gracious ear to the supplication of the worshipper.


The present translation of the Vedânta-sûtras does not aim at rendering that sense which their author may have aimed at conveying, but strictly follows Sankara's interpretation. The question as to how far the latter agrees with the views held by Bâdarâyana has been discussed above, with the result that for the present it must, on the whole, be left an open one. In any case it would not be feasible to combine a translation of Sankara's commentary with an independent version of the Sûtras which it explains. Similar considerations have determined the method followed in rendering the passages of the Upanishads referred to in the Sûtras and discussed at length by Sankara. There also the views of the commentator have to be followed closely; otherwise much of the comment would appear devoid of meaning. Hence, while of course following on the whole the critical translation published by Professor Max Müller in the earlier volumes of this Series, I had, in a not inconsiderable number of cases, to modify it so as to render intelligible Sankara's explanations and reasonings. I hope to find space in the introduction to the second volume of this translation for making some general remarks on the method to be followed in translating the Upanishads.

I regret that want of space has prevented me from extracting fuller notes from later scholiasts. The notes given are based, most of them, on the tîkâs composed by Ânandagiri and Govindânanda (the former of which is unpublished as yet, so far as I know), and on the Bhâmatî.

My best thanks are due to Pandits Râma Misra Sâstrin and Gangâdhara Sâstrin of the Benares Sanskrit College, whom I have consulted on several difficult passages. Greater still are my obligations to Pandit Kesava Sâstrin, of the same institution, who most kindly undertook to read a proof of the whole of the present volume, and whose advice has enabled me to render my version of more than one passage more definite or correct.


lxxviii:1 Nanu vidusho#pi setikartavyatâkopâsananirvrittaye vrishtyannâdiphalânîshtâny eva katham teshâm virodhâd vinâsa ukyate. Tatrâha pâte tv iti. Sarîrapâte tu teshâm vinâsah sarîrapâtâd ûrdhvm tu vidyânugunadrishtaphalâni sukritâni nasyantîty arthah.

lxxx:1 Upalabhyate hi devayânena panthâ gakkhato vidushas tam pratibrûyât satyam brûyâd iti kandramasâ samvâdavakanena sarîrasadbhâvah, atah sûkshmasarîram anuvartate.

lxxxi:1 When the gîva has passed out of the body and ascends to the world of Brahman, it remains enveloped by the subtle body until it reaches the river Vigarâ. There it divests itself of the subtle body, and the latter is merged in Brahman.

lxxxiv:1 Kim ayam param gyotir upasampannah sarvabandhavinirmuktah pratyagâtmâ svâtmânam paramâtmanah prithagbhûtam anubhavati uta tatprakâratayâ tadavibhaktam iti visaye so # snute sarvân kâmân saha brahmanâ vipaskitâ pasyah pasyate rukmavarnam kartâram îsam purusham brahmayonim tadâ vidvân punyapâpe vidhûya nirañganah paramam sâmyam upaiti idam ânam upâsritya mama sâdharmyam âgatâh sarve # pi nopagâyante pralayena vyathanti ketyadisruytismritbhyo muktasya parena sâhityasâmyasâdharmyâvagamât prithagbhûtam anubhavatîti prâpte ukyate. Avibhâgeneti. Parasmâd brahmanah svâtmânam avibhâgenânubhavati muktah. Kutah. Drishtatvât. Param brahmopasampadya nivrittâvidyâtirodhânasya yathâtâthyena svâtamano drishtatvât. Svatmanah svarûpam hi tat tvam asy ayam âtmâ brahma aitadâtmyam idam sarvam sarvam khalv idam brahmetyâdisâmânâdhikaranyanirdesaih ya âtmani tishthan âtmano#ntaro yam âtmâ na veda yastâtmâ sarîram ya âtmânam antaro yamayati âtmântaryâmy amritah antah pravishtah sâstâ ganânâm ityâdibhis ka paramâtmâtmakam takkharîtatayâ tatprakâtabhûtam iti pratipâditam avashiter iti kâsakristnety atrâto # vibhâgenaham brahmâsmîty evânubhavati.

xciv:1 Sankara's favourite illustrative instance of the magician producing illusive sights is--significantly enough--not known to the Sûtras.

ci:1 Cp. Gough's Philosophy of the Upanishads, pp. 240 ff.

cxvii:1 It is well known that, with the exception of the Svetâsvatara and Maitrâyanîya, none of the chief Upanishads exhibits the word 'mâyâ.' The term indeed occurs in one place in the Brihadâranyaka; but that passage is a quotation from the Rik Samhitâ in which mâyâ means 'creative power.' Cp. P. Régnaud, La Mâyâ, in the Revue de l'Histoire des Religions, tome xii, No. 3 1885).

cxviii:1 As is demonstrated very satisfactorily by Râmânuga.

cxx:1 Gough, Philosophy of the Upanishads pp. 213 ff.

cxxi:1 I cannot discuss in this place the Maya passages of the--Svetâsvatara and the Maitrâyanîya Upanishads. Reasons which want of space prevents me from setting forth in detail induce me to believe that neither of those two treatises deserves to be considered by us when wishing to ascertain the true unmixed doctrine of the Upanishads.

cxxiv:1 The Îsvara who allots to the individual souls their new forms of embodiment in strict accordance with their merit or demerit cannot be called anything else but a personal God. That this personal conscious being is at the same time identified with the totality of the individual souls in the unconscious state of deep dreamless sleep, is one of those extraordinary contradictions which thorough-going systematisers of Vedântic doctrine are apparently unable to avoid altogether.

cxxv:1 That section of the introduction in which the point referred to in the text p. cxxvi is touched upon will I hope form part of the second volume of the translation. The same remark applies to a point concerning which further information had been promised above on page v.


Così tra questa
Immensità s'annega il pensier mio,
E il naufrago m' è dolce in questo mare.


Next: I, 1, 1