On Quadruple Division Of Yogasastra

Author(s): A Wezler
Publisher: A Wezler
Catalog link: https://jainqq.org/explore/269647/1

Page #1
A. WEZLER ON THE QUADRUPLE DIVISION OF THE YOGASASTRA, THE CATURVYUHATVA OF THE CIKITSASASTRA AND THE << FOUR NOBLE TRUTHS >> OF THE BUDDHA (Studies in the Patanjalayogasastravivarana II) 1. The publication of the Patanjalayogasastravivarana 1 - wrongly entitled << Patanjalayogasutrabhasyavivarana >> by the editors 2 -- has, as the late Paul Hacker aptly remarked 3, << raised a new problem in the study of the history of [Indian] philosophy >> 4. Hacker's attention was drawn to this text, though it is a commentary on the Yogasutra (=YS) and Yogabhasya (=YBh/YBhasya), because it was regarded by the editors to be a work of sankara, the Advaitin. For, Hacker has, as is well known, given much thought to the problem of how to distinguish between the authentic works of Sankara, the Advaitin, and the many spuria ascribed to him, and he has evolved various criteria to prove authenticity. Therefore, it is by no means unexpected that he accepted the challenge posed by the publication of yet another text ascribed to the Advaitin. In his most stimulating article published in 1968 5 he does not, however, really want to prove the identity of the author of the 1. Ed. by Polakam Sri Rama Sastri and S. A. Krishnamurthi Sastri (= MGOS, no. XCIV), Madras, 1952. Most valuable observations on this text and on the authorship problem have been made by W. HALBFASS; cf. the appendix, Notes on the <>, in his recent monograph, <>, Reinbek, 1983, to which I should like to refer the reader also as regards the seconary literature sofar published on the Vivarana. 2. Cf. my article, Philological Observations on the so-called Patanjalayogasutrabhasyavivarana (Studies in the Patanjalayogasastravivarana I), in IIJ, vol. 25 (1983), pp. 17-39. 3. Viz. on pp. 124-218 of his article, Sankara der Yogin und Sankara der Advaitin. Einige Beobachtungen, in << Beitrage zur Geistesgeschichte Indiens. Festschrift fur Erich Frauwallner >>, aus Anlass seines 70. Geburtstags herausgegeben von G. Oberhammer, Wien, 1968 (= WZKSO, 12-13, 1968-69), pp. 119-48 = Kleine Schriften. Herausgegeben von L. Schmithausen (Glasenapp-Stiftung Bd. 15). Wiesbaden. 1978. pp. 213-42. In the following this article is referred to by << (Hacker) 1968 >>. 4. The German original runs as follows: << ... der der Erforschung der Philosophiegeschichte ein neues Problem stellt >>. 5. Cf. fn. 3.

Page #2

290
A. Wezler
Vivarana and Sankara, the Advaitin; instead, what he does is to argue that if we assume this identity we are able to explain peculiarly Yoga features in the authentic works of Sankara, the Advaitin, to some of which Hacker had already drawn attention in an earlier article of his 6; he further argues that if the Vivarana was in fact composed by Sankara, the Advaitin, it cannot but have been the earliest of his works, and that if Sankara was at first an adherent of (Patañjala-) Yoga and became a Vedāntin only later, the relative chronology of at least some of his later works can be established by examining the extent to which they display Yoga influence in terms of philosophical contents and terminology.
2. It is in this connection that Hacker deals with the quadruple division of the Yogaśāstra as it is stated, and most emphatically at that, at the very beginning of the Vivarana. He says ?: « At YBh 2.15, the division of the Yogaśāstra is compared to that of a system of therapeutics. S(ankara) resumes this idea in the introduction to his YV [=Pātañjalayogaśāstravivarana] when explaining the purpose of the Yogaśāstra. This shows that he looked upon Yoga as a therapeutic system. The same idea, except for the division itself (which is proper only to Yoga), is repeated by $ when he in the introduction to his MāBh [=Māņdükya Bhäsya] points out the purpose of the text to be explained [i.e. the Māndükya Up.]; but what he now brings in is the monistic teaching by which man as one sick due to Suffering has to be led to health of the Self. In introductions to other works of his, s recurs to the concept of therapeutics (cikitsita) (viz. USP [=Upadeśasāhasri-Padyaprabandha] 19, 1), and once again to the idea of illness and health, only that the concept of therapeutical treatment is dropped (viz. USG [=Upadeśasāhasrī-Gadyaprabandha] 47). Here it is natural to assume that the
6. Viz., Eigentümlichkeiten der Lehre und Terminologie Sankaras. Avidyā, Namarūpa, Māyā, Iśvara, in ZDMG, 100 (1965), pp. 246-86 = Kleine Schriften..., pp. 69-109.
7. 1968, pp. 125/219. Explanations added by me are placed within square brackets. The German original runs thus: «YBh 2.15 ist die Einteilung des Yogaśāstra mit der eines Systems der Therapeutik verglichen. s nimmt diesen Gedanken auf in der Einleitung zum YV (p. 2), wo er den Zweck des Lehrsystems erklären will. Damit ist der Yoga als eine Therapie aufgefasst. Den gleichen Gedanken, mit Weglassung der (nur für das Yogasystem passenden) Einteilung, wiederholt $, als er in der Einleitung zu MāBh den Zweck des zu erläuternden Textes angeben will; nun ist es die monistische Lehre, die den Leidkranken zur Gesundheit des Selbst führt. In Einleitungen zu anderen Werken Śs kehrt die Idee der Therapie (cikitsita) noch einmal wieder (USP 19.1) und dann noch einmal der Gedanke der Krankheit und Gesundheit, unter Weglassung des Begriffs der Heilbehandlung (USG 47). Die Annahme liegt nahe, dass die drei Werke, wo der Gedanke der Heilbehandlung betont ist, zeitlich zusammengehören und dem Yoga nahestehen. Denn dieser ist methodische Ubung, die leicht mit einer Therapie verglichen werden kann. Śs Vedānta dagegen will keine Methode lehren, die durch Ubungen kontinuierlich zum Ziel der Erlösung führe; darum lehnt er (später) den Yoga ab, vgl. unten. Dementsprechend ist es im USG (einem vielleicht späten Werk) bloss der Schüler, der von seinem Lehrer eine Art Heilbehandlung erwartet; der Lehrer verspricht ihm nichts dergleichen, er will bloss zur Erkenntnis führen »

Page #3

On the Quadruple Division of the Yogaśāstra
291
three works in which the idea of therapeutical treatment is stressed belong together in point of time and are connected with Yoga. For Yoga is a practice that can easily be compared to therapeutics. S's Vedānta, however, does not teach a method by following which one is gradually led to the goal of liberation; that is why he (later) rejects Yoga, cf. below. Accordingly, in the USG (perhaps a late work) it is only the disciple who expects from his teacher a kind of therapeutical treatment; the teacher, however, does not promise any such thing; all he aims at is to lead to liberating insight ».
2.1. The passage of the Vivarana referred to by Hacker is found almost at the very beginning of the text. It is preceded, or rather introduced, by a statement to the effect that the Yogaśāstra will be studied and the method it teaches practised by people only if its sambandha and prayojana are shown. It is to the prayojana, i.e. the « purpose of the Yogaśāstra, that the Vivaranakāra addresses himself first, for he continues 8:
tatra prayojanam tāvat — cikitsāśāstre 9 tac caturvyūhatvapradarśanadvāreņa vyākhyātam / tadyathā — cikitsāśāstram caturvyūham: rogan, rogahetun, ārogyam, bhaisajyam iti / vidhipratişedhaniyamadvāreņa [ca tat 10] caturvyūhavişayavyakhyānaparam / evam ihāpi « pariņāmatāpasamskāraduḥkhair guņavrttivirodhāc ca duḥkham eva sarvam vivekinah » [YS 2.15] ity ārabhya caturvyūhatvam pradarśitam / tadyathā — duḥkhapracuraḥ samsāro heyah / tasyāvidyānimitto drastrdrśyasamyogo hetun / vivekakhyātir aviplavā hānopāyaḥ 11 / vivekakhyātau ca satyām avidyānivsttiḥ, tannivrttāv ātyantiko drastrdrśyasamyogoparamo hānam, tad eva kaivalyam 12 / ārogyasthānīyakaivalyaprayuktatvād asya tad eva kaivalyam prayojanam //.
2.2. In a later part of his article 13 Hacker returns to this passage; what he says there is 14:
8. P. 2, 1. 1-9 of the edition; cf. also p. 168, 1. 15 ff. In addition, he treats of the sambandha, viz. p. 2, 1. 18 ff.
9. Here the Trivandrum MS (cf. the article mentioned in fn. 2) has the reading cikitsäsästravat taco, which most probably originated under the influence of the preceding tăvat or by díttography.
10. ca tat is an emendation of the editors; it is not corroborated by the Triv. Ms. 11. Cf. YS 2.26; see below p. 295. 12. Cf. YS 2.25; see below p. 295. 13. 1968, p. 139 f./233 f.
14. The German original runs thus: «Da wird der Inhalt des Yoga in Analogie zum System der Therapeutik (cikitsa-śāstra) und in Anlehnung an YS 2.15 ff. in vier Kapitel (vyuha) eingeteilt: 1. "Der leidvolle Samsāra ist zu vermeiden " (in der Therapeutik entspricht das der Lehre von der Krankheit, roga). 2. "Seine Ursache ist die Verbindung des Subjekts mit dem Objekt, die im Nichtwissen gründet" (das ist die "Ursache der Krankheit", roga-hetuh). 3. "Mittel des Vermeidens ist die unbeirrte (sic) Klarheit der Unterscheidungserkenntnis" (das "Heilmittel ", bhaisajya).

Page #4

292
A. Wezler
« The contents of Yoga are here divided into four chapters (vyūha) 15 by analogy with the system of therapeutics (cikitsā-śāstra) and in dependence on YS 2.15 ff.: 1) "Samsāra, full of Suffering, is to be avoided ". (In therapeutics this correspondends to the teaching of Illness, roga). 2) "The cause of samsāra is the connection of the subject with the object based on nescience” (this is the "Cause of Illness", roga-hetuh). 3) "The means by which [samsāra] is avoided is the unswerving lucidity of discriminating knowledge " (the "Remedy", bhaişajya). 4) ” When this has been attained then nescience is eliminated, and the connection of subject and object ceases entirely; that is avoidance, absolute independence" (in therapeutics: "Health", ārogya). It may be noted in passing that this division shows a certain similarity also to the Four Noble Truths of the Buddha. Yet the second and third truth — Suffering and its Annihilation — are gathered into one, and in accordance with the more positive outlook of Brahmanism a chapter is added on the final goal — absolute independence, corresponding to Health.
4. "Ist diese erreicht, so ist das Nichtwissen beseitigt, und damit hört die Verbindung von Subjekt und Objekt absolut auf; das ist das Vermeiden, die Absolutheit" (in der Therapeutik: die "Gesundheit", arogya). Nebenbei bemerkt hat diese Einteilung eine gewisse Ähnlichkeit auch mit den Vier Edlen Wahrheiten des Buddha. Doch sind die zweite und die dritte Wahrheit – das Leiden und seine Aufhebung - in eine zusammengezogen, und entsprechend der positiveren Sicht des Brahmanismus ist ein Kapitel über das Ziel - die Absolutheit, entsprechend der Gesundheit - hinzugefügt ».
«Die Yogalehre so einzuteilen, ist gewiss nicht abwegig. Doch ist zu bemerken, dass erstens im System des Pātañjala-Yoga manches enthalten ist, was nicht streng zu diesem Schema passt, z.B. die Lehre von den magischen Vollkommenheiten (siddhi), und dass zweitens dieses System, so wie es in den Sütren vorliegt, nicht so eingeteilt ist. Diese Einteilung ist eine Interpretation Śs. Sie entspricht ganz dem auch in den Advaitaschriften Ss immer wieder zu beobachtenden stark praktischen, didaktischen, ja therapeutischen Interesse ».
«Es ist daher bemerkenswert, dass s, als er als Advaitin den wesentlichen Inhalt seiner Lehre einmal ganz ohne Bezugnahme auf zu erklärende Texte und rein argumentativ darlegte, auf das gleiche Einteilungsschema zurückkam, das er schon als Yogin aufgestellt hatte. Der Gedanke der Therapie musste jetzt zwar zurücktreten (vgl. oben § 2), aber die Disposition des 2. Prosa-Prakarana der US entspricht genau der Vier-Vyüha-Einteilung. USG 45-47 ist vom Leiden des Samsāra gehandelt. 48 beginnt die Belehrung über die Ursache des Leidens: das Nichtwissen. Ohne scharfen Einschnitt geht sie etwa ab 58 allmählich über in die Anleitung zur unterscheidenden Erkenntnis. USG 109 schildert die Erreichung des Zieles ».
15. In rendering the term vyūha by « chapter » Hacker has, I think, made the wrong choice, though he is in accordance with the PW (vol. VI, column 1485). If its meaning is not simply « part, division >> here, one should rather think of its being used (e.g. in the Arthaśāstra) as a military term to denote a « battle-array », i.e. in a more general sense a «purposeful, systematic arrangement ». Cf. also R. Garbe's explanation of this term as used in Vijñānabhikṣu's Samkhyapravacanabhāşya (see below fn. 31), Sâmkhyapravacana-bhâshya ... aus dem Sanskrit übersetzt..., Leipzig, 1889, p. 11, fn. 2: «... bei dem vyûha liegt ein kramena 'ngopârgânám vinyasah vor...».
Note that the author of the Vivarana offers (p. 169, 1. 3 ff.) two alternative explanations of the compound caturvyűha, viz. catvāro vyūhā asya and eşv artheşu (= sansära, samsārahetu etc.) caturdha vyūho 'sya.

Page #5

On the Quadruple Division of the Yogaśästra
293
To divide the teaching of Yoga in this manner is, to be sure, by no means improper. But it has to be noted that, firstly, the system of Patañjala-Yoga contains various elements that do not strictly fit into the [fourfold] scheme, e.g. the teaching of the magical perfections (siddhi), and that, secondly, this system [i.e. that of Yoga], as it is found in the sūtras themselves, is not divided in this manner. The division is an interpretation introduced by $; it wholly corresponds to the decidedly practical, didactic, nay, therapeutical interest that is to be noted again and again in the Advaita works of S's also 16.
Therefore, it is noteworthy that the Advaitin s, at the one place where he expounds the essential content of his teaching without any reference to texts to be explained and in a purely argumentative manner, takes up the very scheme of division he had already established when still a Yogin. The idea of therapeutics had now to recede into the background (cf. above $ 2), but the disposition of the 2. prose prakarana of the US corresponds exactly to the four vyūha division. USG 45-47 deals with samsāra as Suffering. In 48 we have the beginning of the instruction about the Cause of Suffering: nescience. Without sharp cleavage approximately from 58 onwards it gradually yields to initiation into discriminating knowledge. USG 109 describes the attainment of the goal ».
These remarks call for a critical examination in more than one respect. It appears as if Hacker has lost sight of what he had himself correctly stated earlier, viz. that the quadruple division of Yogaśāstra as taught at the beginning of the Vivarana is already met with in the YBhāṣya on YS 2.15 and, moreover, in the YS itself. Indeed, what the Bhāṣyakāra says on YS 2.15 (parināmatāpasamskāraduḥkhair guņavrttinirodhāc ca duhkham eva sarvam vivekinah, quoted also by the Vivarana, cf. above under 2.1.), after first explaining its wording and purport in detail, is this 17
... tad asya mahato duḥkhasamudāyasya prabhavabījam avidyā / tasyāś ca samyagdarśanam abhāvahetuh / yathā cikitsāśāstram caturvyūham — rogo rogahetur ārogyam bhaisajyam iti, evam idam api śāstram caturvyūham eva / tadyathā 18 — samsāraḥ samsārahetur mokso moksopāya 19 iti / tatra duḥkhabahulah samsāro heyah / pradhānapuruşayoḥ 20 samyogo heyahetuh / samyogasyātyantiki nivettir hānam / hānopāyaḥ samyagdarśanam / ...
This is not only by itself an unequivocal statement, but comparison also clearly shows that it was simply reformulated by the author of
16. Note that this sentence stands in contradiction with what H. had said at the end of the passage quoted above (s 2).
17. I quote the text portion as given on p. 168, 1. 1-7 of the edition of the Vivarana. Cf. in addition p. 171, 1. 1: tad etac chăstram caturvyūham ity adhidhiyate.
18. The author of the Vivarana attests yathā only, cf. p. 168, 1. 24. 19. mokşahetuh according to the Vivaranakära.
20. The author of the Vivarana obviously reads pradhānapuruşasamyogo; cf. pp. 168, 1. 25 and 169, 1. 7.

Page #6

294
A. Wezler
the Vivarana; the one substantial difference to be dealt with first is the point at which this quadruple division is taught in either work; apparently, by the time of the Vivaranakära it had become usual for (sub)commentators or authors of scientific works to deal already at the beginning with the « purpose » (prayojana), etc., of their main subject in order to convince the reader or hearer that it was worthwhile to take the trouble of studying it carefully. That is why the Vivaraṇakāra decided to point out the quadruple division of the Yogaśästra already at the very outset of his commentary; and, to be sure, he need not state explicitly that this « example (dṛṣṭanta 2), i.e. the comparison of the Yogaśāstra with the science of medicine, was not his own invention; instead, what was of importance to him was to emphasize that the division is in accordance with explicit statements of Patanjali himself. And this claim of the Vivaraṇakāra cannot be a limine treated with contempt as yet another example of the common practice of commentators to have what is but their own idea or, at least, an idea of later origin sanctioned, as it were, by the accepted authority of the mula text. His assertion that the caturvyuhatva of the Yogaśastra is pointed out (pradarśita) in the YS itself, beginning with sūtra 2.15, obviously does not belong to the category of such wishful, forced interpretations. Yet, it calls, nevertheless, for closer examination.
The Vivaraṇakāra's reference to YS 2.15 may have been provoked in the first place by the fact that it is that sutra in explaining which the Bhāṣyakāra deals with the subject of the caturvyuhatva, i.e. makes the statement (quoted above) about the quadruple division of the Yogaśāstra; nonetheless it cannot be denied that the sutra itself:
pariņāmatāpasamskāraduḥkhair gunavirodhac ca duḥkham eva sar
vam vivekinaḥ
stands out in that in it the term duḥkha is not simply used as in others also, but is central to it; for it declares that << to a vivekin, i.e. the Yogin who possesses discriminating insight into reality as analysed by Yoga, everything is Suffering in view of the various forms of Suffering consisting in change, mental or physical pain and subliminal impressions 22 and because of the fact that the functions of the constituents [of primordial matter] are opposed to each other ». Undoubtedly, this is a central element of Yoga as expounded in the Sūtra. Ascribing to life, as
21. This term is used by the Vivaraṇakāra himself, viz. p. 168, 1. 21.
22. The author of the Vivarana explains (p. 159, 1. 12) that strictly speaking all three of them are but « causes of Suffering » (duḥkhanimittäni). For further elucidation, I should like to refer the reader to the Bhasya and the Vivarana on this sutra. That this sutra clearly exhibits Buddhist influence was shown by L. DE LA VALLÉE POUSSIN, Mélanges Chinois et Bouddhique, V, 1936-37, p. 234 f. The term samskära has, of course, quite another meaning in Buddhist texts. Though one cannot, of course, be absolutely sure, it was probably reinterpreted already by the author of YS 2.15.

Page #7

On the Quadruple Division of the Yogaśāstra
295
it does, the character of being nothing but Suffering, YS 2.15 must indeed be regarded as the pivotal statement about the nature of reality in the YS. And in the immediately following sūtra (2.16: heyam duḥkham anāgatam) it is taught that it is Suffering yet to come that has to be avoided >> 23, whereas in YS 2.17 the « cause of that which has to be avoided » (heyahetu) is defined as drastrdrśyayoḥ samyogah, i.e. « the connection of the subject with the object ». As the three latter terms call for further elucidation, actually given in sūtras 2.18-23, it is only in YS 2.24 (tasya hetur avidyā) that nescience is expressly stated to be the cause of this connection. The term hāna, which the reader now expects, is in fact met with in the next sūtra (2.25) which reads thus: tadabhāvāt samyogābhāvo hānam, tad drśeh kaivalyam: « Avoidance is the nonexistence of the connection on account of its (i.e. the nescience's) nonexistence; this is the seer's (i.e. the spirit's or soul's) being-absolutelyby-himself ». The obvious question as to the means of avoidance, i.e. of attaining the state of liberation, is answered by the immediately following sūtra (2.26): vivekakhyātir aviplavā hānopāyah « the means to avoidance is the steady or firm lucidity-of-discriminating-knowledge », i.e. the complete or perfect realization of the absolute difference between purusa and prakrti.
Thus it is already on the historical level of the YS itself that, in a series of sūtras obviously forming a unitary whole, four clearly demarcated concepts are attested, viz.
-
23. The last word of this sūtra (anāgatam) which is left out in the Bhāşya- and Vivaranakāra's independent expositions of the caturvyūhatva, most probably for practical reasons only, is accounted for by the author of the YBh in the following manner: dunkham atitam (Viv.: vartamānāj janmano 'tikrāntajanmalaksanam dunkham) upabhogenātivāhitam na heyapakse vartate (Viv.: svayam eva hinatvāt) / vartamānanca (Viv.: janmalaksanam dunkham) svakşane bhogārūdham (Viv.: svam vartamānam bhogavisiştam ksanam adhvānam upabhujyamānātmakatvenārūdham) iti na tat kşanāntare heyatām āpadyate (Viv.: svayam eva bhogena hinam na hānāya kşarāntaram apeksate) / tasmād yad evānāgatam (Viv.: prāyanāntarabhāvijanmaduḥkhalaksanam) dunkham tad evākṣipātrakalpam (cf. YBh on YS 2.15) yoginam kliśnāti (= udvejayati, cf. Viv. p. 164, 1. 7 ff.) < netaram pratipattāram > (not attested in the Vivarana and indeed to be suspected as being a later addition; cf. also Bh 164, 1. 3) / tad eva heyatām ápadyate (the Viv. reads tad eva heyam here; the other reading might, in fact have originated only secondarily, viz. due to heyatām āpadyate in the second sentence; the Viv. adds by way of explanation: samyagdarśanena hāniyam ucyate / bhavişyajjanmābhāvāyaiva yatitavyam/ na vartamānadunkhanirodhāya / vartamānajanmayiyatişāyām hi samyagdarśanam asakyaviniyogatvād anarthakam syāt / muktabānavat pravsttaphalatvād vartamānaduḥkhasya / bhavisvati punar apravrttatvād bijanirodha upakalpata iti samyagdarśanārthavattvam / svavişayo hi samyagdarśanasya sa [i.e. bhavisyajjanmā] iti /). That is to say according to the Vivaranakāra at least, « Suffering yet to come » means « Suffering consisting in and experienced in the next birth(s)». In any case, the sūtra is right as it is in fact only anāgatam duḥkham that can, if at all, be avoided; but at the same time the addition of anāgatam may be taken to indicate that the author was more interested in actual Yoga practice and its exact description than in making general abstract statements.

Page #8

296
A. Wezler
heyam defined as 24 duḥkham heyahetuh
drastrdrśyayoh samyogan hānam
samyogābhāvah, i.e. kaivalyam and finally, hānopāyaḥ
vivekakhyātir aviplavā. Evidently this quadruple division is substantially identical with that expounded by the Bhāsyakāra in his commentary on 2.15; and what the Vivaranakāra does is to resume this division. On the other hand, one cannot fail to observe that in the YS itself the division is neither expressly stated to be a divisionary scheme; nor to be a quadruple one; nor to apply to the Yogaśāstra; nor, finally, is it compared to a similar division of the Cikitsāśāstra. Nevertheless, the conclusion apparently arrived at by the Bhāsyakāra, viz. that what the author of these sūtras actually had in mind was really a quadruple division of the system of Yoga, cannot but be accepted, by any unbiased reader who is aware of the peculiarities of a Sūtra text, as being perfectly legitimate and cogent. The only material difference worth noticing here lies in the absence of even an allusion to the science of medicine in the YS: to all appearance, this is a new element that was probably introduced for the first time by the YBhāsya.
Drawing now the final conclusion from what has been stated in the foregoing on the evidence found in the Bhāsya and in the Sūtra for the quadruple division of the Yogaśāstra, I cannot help stating in so many words the impression, created not only by this part of Hacker's article, that he does not in this case endeavour to get the better of his prejudices, but overrides, rather arbitrarily, all facts that might stand in his way and are apt to undermine the basic assumption from which he starts. Thus his assertion that the system of Yoga «as found in the sūtras themselves, is not divided in this manner »> is not merely bold, but demonstrably wrong 25, and it is by no means legitimate to give the author of the Vivarana credit for having « established > this division as Hacker says, or to state that « the division is an interpretation introduced by S.».
2.3. As regards this division, any comparative study of the Vivarana, on the one hand, and of the presumably early or late works of S., the Advaitin, on the other, must hence start from the following observations: 1. The division is clearly attested already in the YS though it is not
explicitly taught as quadruple and applying to the Yogaśāstra.
24. I am, of course, aware of the fact that not all the sütras quoted are « definitions >> in the strict sense of the word.
25. At least as regards the caturvyūhatva as a divisionary, scheme. Obviously Hacker had lost sight of the important difference between a systematic division or divisionary scheme as such and its having actually been made the basis of a corresponding disposition of the material.

Page #9

On the Quadruple Division of the Yogaśāstra
297
2. It is expounded likewise by the author of the YBhāşya, unequi
vocally and in detail, and it is he who apparently has to be given the credit for comparing the quadruply divided Yogaśāstra to the science of medicine; in any case, this comparison seems to have
been added later. 3. The division is but repeated by the Vivaranakāra, viz. for the
obvious reasons stated already at the very outset of his work. In order not to go astray in interpreting the relevant passages, it is further advisable, if not even imperative, not to follow Hacker's argumentary approach. As has also been shown elsewhere 26, it is by far better for considerations of method not to let oneself be impressed too much by the fact that the Vivarana is in its colophons ascribed to a Sankarabhagavant; what we should do instead is to take the authorship of the Vivarana to be a problem which is still unsolved. References to the quadruple division as a whole, or in part, found in a work of S., the Advaitin, should not therefore be studied on the basis of the assumption made by Hacker. Instead of arguing the way he does, viz. that if we assume the identity of the Vivaranakära and S., the Advaitin, we are able to explain peculiarly Yoga features in the authentic works of S., the Advaitin, in the manner outlined by Hacker, we should face up to the real problem in all its seriousness and intricacy, viz. put ourselves, without bias, the pivotal question whether the quadruple division, if met with in a work of $.'s, or even suggestions of such a division there, can in fact best be explained by assuming the Vivarana to be another and then, to be sure, the earliest work of the famous Advaita philosopher.
This assumption could be regarded as necessary if, and only if, the comparative study of the relevant passages in authentic works of $. listed by Hacker himself, on the one hand, and the exposition of the caturvyūhatva of the Yogaśāstra in the Vivarana, on the other, were to reveal so specific a correspondence that the hypothesis that the author is one and the same in both cases would suggest itself as the only solution or, at least, the one most plausible.
In view of the close similarity between the Vivaranakāra's exposition and that of the Bhāṣyakāra, « specific correspondence » could in our case firstly mean « literal agreement ». No such agreement has, however, been pointed out by Hacker, and, to be sure, there is none. But, secondly, what about the condition of a peculiar property being common both to the Vivarana and an authentic work of S., the Advaitin?
Is not the sequence of the last two members of the fourfold division in the case of the USG reversed, and does not this change, on the one hand, stand clearly in contrast to the order of enumeration as found in the YS and -Bhāsya, and strikingly agree, on the other, with the Viva
26. Viz, in the article mentioned in fn. 2.

Page #10

298
A. Wezler
raņakāra's exposition? One cannot but answer this question in the positive, and it is surprising that Hacker is silent on this. Nevertheless, one will hesitate to rest satisfied with simply noting this admittedly interesting point of agreement, at least, if one wants to escape the reproach of superficiality or of drawing premature conclusions. For, the correspondence might equally be coincidental.
Now, as for the USG, the reason for S.'s giving first what Hacker calls the « initiation into discriminating knowledge and describing only thereafter the final goal or its attainment, is, I think, clear enough and need not provoke controversy. Since this portion of the USG is conceived by ś. as a dialogue between a disciple and a teacher approached for instruction, there can hardly be any doubt that the instruction is given with a view to its application; what $. has in mind is the actual succession of stages the disciple has to pass through; and that in reality the final goal is, if at all, attained last need hardly be mentioned.
The crucial point, however, is: are we to assume that the Vivaraņakāra in changing the sequence was led by the same or a similar motive? In his case, the fact that he does not keep to the order of succession as found in the YS and Bhāsya is even more striking, since the four vyūhas of the Cikitsāśästra are enumerated by him exactly in the same sequence as by the Bhāsyakāra, the result being that the two series lack complete parallelism, and that the author is compelled to make clear in his last sentence that kaivalya in the quadruple division of Yoga corresponds to ārogya in that of the Cikitsāśāstra (cf. arogyasthānīyao in the passage quoted above).
It is, however, this last sentence, together with the context in which the issue of the caturvyūhatva as a whole stands in the Vivarana (cf. above p. 291), that provides us with the key for discovering the reason for the transposition in this text. The problem from which the author starts is, as has been noted already above, the exigency of dealing at the outset of his work explicitly with the prayojana of the Yogaśāstra. Therefore, it is entirely understandable that particular stress is laid on that vyūha to which the practice of Yoga is ultimately said to lead, and that this emphasis is achieved also by letting the whole discussion culminate in the exposition of what is taken to be the « purpose », i.e. of kaivalya.
To adduce further circumstantial evidence, if additional arguments are at all necessary: the transition from YS 2.24 to the subsequent one is gained in the Bhāşya by the following remark: heyam duḥkham heyakāranam ca samyogākhyam sanimittam uktam / ataḥ param hānam vaktavyam/. All the Vivaranakāra says in commenting on the latter sentence is (p. 203.13-14): atah param hānam ārogyasthānīyam mokşaśāstraprayojanam vaktavy am / vakşyamāṇasamkārtanam ca śrotsbuddhisamādhānārtham /. It is true that he does not state explicity why « avoidance is to be taught next », but at the same time the absence of any remark on his part to the effect that in reality, i.e. in the practice

Page #11

On the Quadruple Division of the Yogaśāstra
299
of Yoga, what is called hāno pāya comes, of course, first is, I think, quite conspicuous.
These observations will, I trust, suffice to corroborate the assumption made above, viz. that the conspicuous transposition of the last two vyūhas in the introduction of the Vivarana is caused by nothing else than the wish to bring out into proper relief the « śāstrārambhaprayojana » (cf. p. 161.11) 27.
Therefore 28, it can safely be stated that the agreement between the USG and the Vivarana as regards the interchange of the last two vyūhas is but coincidental, i.e. caused in each case by quite different deliberations. It is hence quite possible, that s., if at all under the influence of Yoga texts in this respect, was inspired either by the YBhāsya or perhaps even the YS itself. On the basis of the methodological considerations outlined above (p. 297) and in view of the observations made with regard to the exposition of the theory of the caturvyūhatva in the three Yoga texts (cf. above p. 296), one cannot, therefore, help drawing the final conclusion that the comparative study of the USG and the Vivarana does not by any means yield sufficient evidence for the identity of the authors of the two texts as taken for granted by Hacker.
Although Hacker's basic assumption is thus — once more shown to be highly problematic, nay practically unjustified, the question arising next should not be left unasked, viz. whether an element in authentic works of s., the Advaitin, reminding us of the Yoga theory of caturvyūhatva warrants the assumption of an external influence in general or of an influence exercised on s. by the YBhāsya or the YS in particular. What is called for in this connection is a critical reexamination of the passages referred to by Hacker; such a reexamination yields the following results.
27. I do not, of course, want to keep silent about the fact that there is, on the other hand, a passage in the Vivarana where in striking and awkward contrast to its beginning it is the hänopāya, viz. samyagdarśana, that is spoken of as prayojana, viz: p. 169, 1. 14-15: ārabhyamānasya śāstrasya samyagdarśanam eva prayojanam (cf. also the quite frequent expression samyagdarśanaśästra). If this latter remark is not merely meant to account for the somewhat puzzling fact that the Bhāşyakāra already in commenting on YS 2.15 (and not only on 2.20) touches on the topic of the hātr (cf. p. 169, 1. 10 ff.), other explanations one could think of are [1] that this apparent inconsistency is due to an upacāra: the term which primarily denotes the means is used metonymically to denote its result; or [2] that samyagdarśana insofar as it is the pratipaksa (cf. 168.16) of Nescience, the true cause of Suffering, is here regarded as the main thing to be achieved, ensuing almost automatically as it does hāna; or [3] that samyagdarśana is given preference because hāna is an avastu, i.e. not something positive, but only the not-existing-any-longer of the connection between subject and object (bandhanoparama) (cf. p. 204, 1. 13-16). On the other hand, however, the text might be corrupt here: śāstrasya could have replaced an original sūtrasya (cf. also l. 14 ... Šāstram atanparam ārabhyate) and samyagdarśanam could resume this term as used in a narrower sense in the YBh itself, viz. p. 170, 1. 2.
28. In any case, the Vivaranakāra's motive for the transposition cannot have been s's also! On the order of enumeration of the four vyūhas in Yoga texts and that of the Four Noble Truths cf. below p. 306.

Page #12

300
A. Wezler
On the one hand it has to be admitted that the disciple in USG 47 compares himself to a sick person (rogin) and that S., led by deliberations like those outlined above, might have deliberately changed the sequence of the last two vyūhas against that attested in the YBhāṣya and YS. On the other hand, it cannot be disputed that the quadruple division itself is nowhere in the USG stated explicitly. This fact was, admittedly, noticed by Hacker; but is the explanation he offers, viz. that the division « is proper only to Yoga », not too simple a solution, or rather an attempt to avoid addressing himself to the methodological problem involved? In any case, Hacker's remarks are thought-provoking; for he is clearly of the opinion that the dispositional structure underlying the USG materially corresponds to the four vyūhas as enumerated by the Vivaraṇakāra; this might be true, though there still remains a doubt, viz. that when he « discovered » this internal structure of the USG Hacker was fully convinced of the identity of the two authors. It is, of course, not possible to adduce against Hacker the argument that it would then hardly be understandable that S. should, nevertheless, have refrained from explicitly introducing the quadruple division itself precisely because it «< is proper only to Yoga » 29. For Ś. may have applied the quadruple division to the teaching of Vedanta inspite of its impropriety, i.e. deliberately, yet also so inconspicuously that it passes almost unnoticed.
Therefore, what can be said is at best that the possibility of an influence exercised by the YBhāṣya or perhaps the YS itself cannot be entirely ruled out. And if the internal structure of the USG really reflects the quadruple division, such an influence may even be regarded as comparatively probable in view of the fact that S.'s works show this influence in other respects, too. But, the other possibility cannot be excluded with absolute certainty, viz. that Ś. was, on the contrary, under the influence of some other tradition, or under no specific influence at all. As to this latter alternative, one cannot but ask the following questions: would not anyone who takes life to be nothing but Suffering, or Illusion, and who accordingly looks upon man as in need of liberation incline most naturally to the view that man as he is can be compared to a person suffering from a physical ailment? Is not anyone who takes such a view of life quite easily led to put himself the question as to
29. Apart from the self-contradictory character of Hacker's respective remarks (cf. fn. 16), I should like to stress the following: even if S did not regard the quadruple division as « proper only to Yoga », the fact that it is not stated by him expressly to be the USG's divisionary scheme might be due to some other reason also; after all this is also not done in the YS. In any case, this possibility does not affect my argument.

Page #13

On the Quadruple Division of the Yogaśāstra
301
the reasons of this metaphysical ailment, the means of overcoming it and the state of final liberation? Thus it is, perhaps, even equally possible that S. was here influenced by a general religious tradition or a more specific one to which he himself belonged.
The probative force of the other two passages to which Hacker has drawn attention, viz. USP 19.1. and the introduction of the Māndükyabhāsya, turns out to be weaker still if examined on the basis of the methodological considerations which have been outlined above.
The metaphor contained in the expression jñānavirāgabhesajam and the characterization of this medicine as trsnājvaranāśakāranam cikitsitam in the USP do not by themselves warrant the conclusion that what S. is actually doing here is to resume concepts developed by him when still an adherent of Yoga. This holds equally good for the passage in the Māndükyabhāsya Hacker must have had in mind; there the question as to the prayojana of the teaching expounded in this Upanişad is answered by the following statement:
rogārtasyeva roganivrttau svasthatā / tathā duḥkhātmakasyātmano dvaitaprapañcopaśame 30 svasthatā / advaitabhāvah prayojanam / dvaitaprapancasyāvidyākrtatvād vidyayā tadupaśaman syād iti brah
mavidyāprakāśanāya asyarambhaḥ kriyate 7. Since the problem from which s. starts here is the same as that discussed by the Vivaranakāra, one could, of course, take here, too, the view that $. is but resuming older ideas of his, but again it is by no means self-evident that other possibilities of accounting for this comparison, possibilities which suggest themselves most easily, can be excluded with any degree of certainty. On the contrary, one cannot but wonder if, before the publication of the Vivarana expressly attributing it to Sankara, Hacker himself or anybody working in the same field would have thought of suspecting these passages of showing an (external) influence.
2.4. What remains to be followed up now is Hacker's suggestion of a « certain similarity between the quadruple division of the YogaŠāstra and the Four Noble Truths of the Buddha. Unfortunately, it is not quite clear whether what he had in mind was only a structural similarity or an historical connection: the manner in which he formulates his view, though probably with the aim of revealing yet another instance of Hinduism being indebted to Buddhism, does not by itself preclude the first interpretation. Be that as it may, in order to examine critically his standpoint it is advisable to summarize the results of our foregoing study of Yoga texts, and this is best done by a synoptical chart:
30. The term prapancopaśama is met with also in Nāgārjuna's Mūlamadhyamikakārikā, viz. 25.4.

Page #14

caturvy hatva
of the Cikitsaśāstra according to the YBhāṣya and the Vivarana
1. rogaḥ
2. rogahetuḥ
3. ārogyam
4. bhaiṣajyam
caturvyuhatva of the Yogaśāstra according to the
1. Yogasütra
heyam = duḥkham
heyahetuḥ = drastrdṛśyayoh samyogaḥ
hānam =
samyogābhāvaḥ, viz. kaivalyam
hānopāyaḥ = vivekakhyātir aviplavā
2. Yogabhāṣya
heyam =
duḥkhabahulaḥ samsāraḥ
hanam/mokṣaḥ= samyogasyätyantiki nivṛttiḥ
3. Vivarana 11
hānopāyaḥ/mokṣopāyaḥ = samyagdarśanam
heyam =
heyahetuḥ/samsārahetuḥ = tasya hetuḥ =
pradhanapuruṣayoḥ samyogaḥ
duḥkhapracuraḥ samsāraḥ
avidyānimitto draştrdṛśyasamyogaḥ
hānam =
vivekakhyātau satyam avidyānivṛttiḥ
tannivṛttāv ǎtyantiko drastrdṛśyasamyogoparamaḥ
hänopāyaḥ = vivekakhyātir aviplavā
31. Later occurrences of the theory of the caturvyuhatva are found e.g. in Madhava's Sarvadarśanasamgraha at the end of its exposition of the « Patañjaladarśana» (ed. by T. G. Mainkar, Government OS, Class A, no. 1, Poona, 19783, p. 388), Vijñānabhikṣu's Samkhyapravacanabhäşya at the end of the introduction as well as at the end of the first Adhyaya (ed. by R. Garbe, HOS, II, Cambridge, Mass. 1943, p. 5 and p. 74) and (except for the comparison with the science of medicine) in Nārāyaṇatirtha's commentary, called Candrika, on the Samkhyakārikā (ed. by Dhundhirājā Sāstrī, Haridas SS, 132, Benares, 1941, p. 1, ed. by Bechanarama Tripathi, BSS, 5, Benares, 1883, p. 2; cf. also TH. AUFRECHT, Catalogi codicum manuscriptorum Bibliothecae Bodlianae, pars 7, codices sanscriticos complectens, Oxford, 1864, p. 237b [no. 569]).
302
A. Wezler

Page #15

On the Quadruple Division of the Yogaśästra
303
Hacker's remarks imply that the first vyūha, i.e. the first line of our chart, is taken by him to correspond to the first Noble Truth of the Buddha, i.e. that of dunkha (P. dukkha); and this correspondence can in fact be hardly disputed. Yet, of the second and third truth he says that they « are gathered into one »; this cannot but mean that there is but one element in the quadruple division of the Cikitsā- and Yogaśāstras corresponding to the truths of duhkhasamudaya (P. dukkhasamudaya) and duḥkhanirodha (P. dukkhanirodha) taken together. Hacker does not specify to which of the remaining vyūhas these two are to correspond, and it is indeed quite difficult to reconstruct his argument. In any case, his assertion is far from being convincing. Therefore, it seems advisable to examine the question at issue once more, without reference to Hacker's view.
What is meant by the concept rogahetu of the Cikitsāśāstra and by heyahetu, corresponding to it in Yoga, is clearly nothing but « what constitutes the cause of the disease >> or < the cause of that which has to be avoided ». If dunkhasamudaya, on the other hand, is taken to denote a process, i.e. the rise of Suffering, one would have to take note of a striking conceptional difference, viz. that between cause and process as denoted by these terms respectively. But, can it simply be taken for granted that it is this and nothing else that is meant by the term duḥkhasamudaya 3la? Such an assumption is scarcely satisfactory, and not so much because the formation of the primary noun samudaya- would, at least according to Pāņ. 3.3.56 in connection with 3.3.18 and 19, allow among others of a meaning « that from which or by which something arises », but because there are passages in canonical Buddhist texts which seem to show clearly that the term was in fact used to denote that by which the rising of dunkha is caused; e.g. at Vbh 107 it is expressly stated: tanhā ca avasesā ca kilesā ayam vuccati dukkhasamudayo; or, to give another example, at S III.158 the question katamo ca bhikkhave dukkhasamudayo is answered thus: yāyam tanhā ponabbhavikā nandi rāgasahagatā tatra tatrābhinandini, seyyathidam kāmatanhā bhavatanhā vibhavataṇhā ayam vuccati bhikkhave dukkhasamudayo.
One cannot, of course, be absolutely sure that these explanations are meant to be proper definitions of the term dukkhasamudaya; nor that they meet the original intention of the term; nor that the term was not elsewhere understood to denote a process. But it has, on the other hand,
31a. For a detailed and careful discussion of the grammatical and syntactical problems posed by the different versions in which the Four Noble Truths are stated in Buddhist texts cf. the recent article of K. R. NORMAN, The Four Noble Truths: A Problem of Pāli Syntax, in « Indological and Buddhist Studies, Volume in Honour of Professor J. W. de Jong on his Sixtieth Birthday », ed. by L. A. Hercus et al., Canberra, 1982, pp. 377-91. Norman does not, however, address himself to the semantic problems involved. As for dukkhasamudaya, cf. L. SCHMITHAUSEN, On Some Aspects of Descriptions or Theories of « Liberating Insight >> and < Enlightenment » in Early Buddhism, in «Studien zum Jainismus und Buddhismus, Gedenkschrift für Ludwig Alsdorf >>, Wiesbaden, 1981, p. 203, fn. 14a.

Page #16

304
A. Wezler
to be admitted as another possibility that Buddhist monks themselves have taken it to mean the cause of Suffering, i.e. that which causes duḥkha to arise. And it is this possibility that alone matters in the present context; for, if understood thus, the second Noble Truth would indeed fully correspond to the second vyūha.
As for duhkhanirodha, similar observations can be made. In view of its formation, the term could denote the process of suppressing or destroying, viz. that which causes Suffering; or else it could mean the result this process is ultimately to lead to, viz. the final suppression or rather cessation of Suffering. Again, there are passages indicating that Buddhist authors themselves may well have taken the term to denote the result; e.g. at S III.158 it is defined as tanhāya asesavirāganirodho cāgo paținissaggo mutti anālayo, and explained at ND 1.94 by nibbāņam. There are two possibilities of interpreting the first of these two statements: either it is a definition proper; in this case, the implicit equation of dukkhanirodha with tanhānirodha (which is expressly given at A III.416) precludes the possibility of taking nirodha to denote the result; or it is a quasi-definition, i.e. an inexact explanation where the substitution of the result by that which causes it, i.e. of dukkha by tanhā, is quite understandable, and not only in terms of metonymy; in this latter case, it would have to be regarded as evidence warranting the assumption that the term dukkhanirodha has in fact been used to denote the result. Interpreted in this manner, and, to be sure, this interpretation is more probable, the third Noble Truth could be said to be quite similar to the third vyūha, i.e. ārogya/hāna 32. This assumption is further corroborated by the fact that the nirvāṇa is compared also in the Abhidharmakośabhāşya to the state of arogya 33
However, the fact should not be lost sight of that the latter term virtually means the state of being again free of disease », i.e. that the basic idea is in this case palpably different in that, in accordance with biological facts, a previous state of health is presupposed which is but regained. The Buddhist analysis of existence does not, of cours, know of an analogous previous state of freedom from Suffering; on the contrary, Suffering is recognized as the fundamental constituent element of existence. However, this conceptional difference does not really detract from the conclusion arrived at above, viz. that the third Noble Truth in fact corresponds to the third element in the quadruple division of the Cikitsāśāstra; for, what has been said with regard to the
32. Note that hāna is explained by Yoga authors to consist in the non-existence (abhāva) or in the coming-to-an-end (nivstti, uparama) of that which causes Suffering or in liberation (mokşa).
33. P. 202.8; I thank Dr. Ch. Lindtner for having drawn my attention to this passage. According to the Hôbôgirin p. 229 « le Nirvāņa est déjà comparé à l'absence de maladie ... dans une stance gnomique du Madhyamāgama... [MN 1.510.9-10] ». Cf. also Pancaskandhaprakarana, ed. by Ch. Lindtner, in AO, 40 (1979), p. 122, fn. 28, as well as the references s.v. arogya- in the CPD.

Page #17

On the Quadruple Division of the Yogaśāstra
305
relation in which the term duḥkhanirodha stands to the term ārogya, holds good likewise for the comparison between the science of medicine and Yogaśāstra: for, also according to Yoga there is no such thing like a previous state free from duḥkha, i.e. preceding samsāra 34. Therefore, one cannot but observe that already the correspondence between the four vyūhas of the Cikitsāśāstra, on the one hand, and the four divisions of the Yogaśāstra, on the other, is conceptionally and terminologically not an absolutely complete one. Yet, at the same time one will recall the mediaeval proverb « omne simile claudicat » 35 and, accordingly, refrain from overrating this element of discrepancy in either case, i.e. as regards the correspondence between the Cikitsāśāstra and the Yogaśāstra as well as that between this caturvyūhatva and the Four Noble Truths. In any case, one cannot subscribe to Hacker's views, including that last voiced by him, viz. that « in accordance with the more positive outlook of Brahmanism a chapter is added on the final goal »; for, though the concepts of hāna, etc., on the one hand, and of duhkhanirodha, etc., on the other, are admittedly different in terms of philosophical content, one cannot fail to notice that the descriptions given in Yoga texts of the final goal, i.e. « absolute independence >> (kaivalya), too, do not stand out by being particularly detailed. On the contrary, one is rather struck by the obvious reserve of Yoga authors in this regard 36.
Thus one cannot but arrive at the conclusion that, since the third line of our chart might well correspond to the third Noble Truth, there is hardly any basis for Hacker's assumption that this vyūha forms a peculiar addition on the part of the Brahmanical Sastras, not to speak of the « more positive outlook » of Brahmanism it allegedly demonstrates. The Buddha, too, did not confine himself to analysing existence and to unveiling its basic character of being nothing but Suffering, but has at the same time claimed to have discovered a practical path to liberation.
The final question to be discussed in this connection is whether the last. Noble Truth, i.e. that of the dunkhanirodhagāmini pratipad (P. dukkhanirodhagāmini patipadā), can also be compared to one of the four vyūhas. Though Hacker is silent on this question, it is, I think, by no means illegitimate to assume a correspondence between this āryasatya and the remaining vyūha, i.e. bhaisajya or hānopāya.
Systematically, they are in fact not different: common to all three of them is the status and function of a remedy. But there are also
34. At least not for the individual living beings.
35. Cf. H. WALTHER, Lateinische Sprichwörter und Sentenzen des Mittelalters, Göttingen, 1963-69, Bd. II, 3, p. 590 (nr. 19877b).
36. This reserve may be caused either by the Yoga conception of kaivalya itself (cf. the Vivaranakāra's characterization of hāna as an avastu; see fn. 27) or to the well-known reluctance of mystics to give a detailed positive description of their experience(s), or to the lack of such experience(s) on the part of the authors concerned, or, finally, to a combination of some of these reasons.

Page #18

306
.
A. Wezler
differences which cannot be ignored. Terminologically and conceptionally the fourth Noble Truth is distinguished by the fact that it explicitly states that the « remedy >> is not something lying at hand, but that in order to reach the final goal of duḥkhanirodha, one has to take a specific route leading to it. Though the distance to be covered is, quite understandably, not specified, the image of a path forms a peculiar and central element of the Buddhist concept 37. Of course, there cannot be the least doubt that it is precisely in this regard that the Yogic concept of « means of avoidance »is virtually identical with the fourth Noble Truth; for, the central purport of Yoga is to teach a practice by applying which one is gradually led to vivekakhyāti that ultimately brings about the irreversible cessation of the connection between subject and object, i.e. kaivalya 38. But it remains noteworthy that the character of being a process is not indicated by the corresponding term employed in Yoga texts or by the explanations given for it by the commentators 39. Inspite of this difference, however, the correspondence between the fourth Noble Truth and the vyūha listed in the last line of our chart cannot but be styled as quite close.
The final result achieved by our comparison of the Four Noble Truths as a whole with the quadruple division of the two Brahmanical Šāstras is, therefore, that the correspondence between them can be regarded as even strikingly close, if we make the assumptions noted above with regard to the explanation of Buddhist terms. It must not, however, be forgotten that it is but a systematic correspondence which has thus been discovered. The historical problems evidently involved have still to be taken up for consideration.
Regarding their order of succession the Four Noble Truths perfectly agree with the four vyūhas of the Cikitsāśāstra and those of the Yogaśāstra as enumerated in the YS and YBhāşya 40. In view of the above (p. 297 f.) discussion on the transposition of the last two members common to both, the Vivarana and the USG of Sankara, the Advaitin,
37. Note that the expression dukkhanirodhagamini pafipadā is replaced by maggo in what Norman (cf. fn. 31a) calls the « mnemonic >> sets. That this image is met with in the YBh also, viz. on YS 2.26 (p. 205, 1. 3 of the edition of the Vivarana), is most probably due to Buddhist influence (on which in general cf. L. DE LA VALLÉE POUSSIN, loc. cit., fn. 22).
38. In this connection cf., apart from passages already quoted, Vivarana e.g. p. 2.23 ff. and p. 121, 1. 4 f.: ... kaivalyasya sādhanam samyagdarśanam / yogasādhanäni ca yogadvārena samyagdarśanasādhanány eva / ...
39. Cf. fn. 37.
40. In the Sāmkhyapravacanabhāşya (cf. fn. 31) the order of succession has been changed, viz. to roga (= heya), arogya (= hāna), roganidāna (heyahetu) and bhaisajya (= hänopaya). This is, no doubt, equally a systematical order, only that the rationale is slightly different, viz.: first the contrary states are named and only thereafter, but with reference to them, the cause and the means, which again form a pair of corresponding entities. The term nidāna is also noteworthy in that it is used frequently in Buddhist texts, too, though first of all to denote the cause of metaphysical ailment.

Page #19

On the Quadruple Division of the Yogaśāstra
307
and on the basis of the tentative explanations proposed for the transposition in either text, the conclusion that suggests itself is that the traditional order of succession of the Four Noble Truths, strictly observed, as far as I know, whenever all of them are enumerated in Buddhist texts, is like that of the four vyuhas of the Cikitsāśāstra a systematical one, i.e. reflecting logical stages of analysis, but not stages passed through successively in actual practice.
3.1. A further problem, and to be sure, an important one, not discussed in the foregoing is posed by the comparison drawn in Yoga texts between the science of Yoga and the science of medicine. What I am referring to is the assertion met with first in the YBhāṣya that systematically the Cikitsāśāstra is divided into four parts. This is stated in such a manner that one cannot but gather the impression that the division is a genuine one: it is hardly conceivable that the caturvyuhatva of the Cikitsāśāstra as expounded in the YBhāṣya and the Vivarana is simply a fabrication made for the sole purpose of establishing a parallel to the fourfold division of the Yogasastra as implied already by the Sūtra itself. The whole tenor of the relevant passages has, on the contrary, to be taken to clearly indicate that the caturvyuhatva of the science of medicine was regarded by the authors as an indisputable, if not even a well-known, fact they could draw upon, in accordance with the basic function of a dṛṣṭanta, for the sake of illustration.
But, I think, one can go even a step further and make the assumption that this correspondence is not stated merely for the sake of illustration, but that much more is intended by it, viz. to bring the method of Yoga into close systematic proximity to the Cikitsäsästra and to intimate thereby that it is not merely of equal importance, but that it surpasses the science of medicine in that it does not cure simply an accidental physical disease, but forms the (allegedly) only remedy against the ailment every living being naturally suffers from, an ailment medicine is unable to cope with.
That this interpretation does not go too far is clearly corroborated by a well-known Samkhya text, viz. Isvarakṛṣṇa's Samkhyakārikā where already at the very outset, i.e. in Kārikā 1, it is expressly stated that the perceptible means of removal of the threefold suffering have neither an absolutely sure nor a necessarily final effect (dṛṣṭe sāpārthā cen naikāntātyantato'bhāvāt), And there can hardly be any doubt that the commentaries, above all the Yuktidīpikā 41, do meet the intention of Isvarakṛṣṇa when referring here to the Ayurveda.
3.2. Therefore, it is legitimate to check this assertion of Yoga authors and to ask the elementary question whether this quadruple division is attested in any of the extant medical texts. But before actually looking into any of these works it is necessary to realize a basic problem
41. Cf. p. 12, 1. 9 ff. in R. C. Pandeya's edition, Delhi, 1967.

Page #20

308
A. Wezler
involved. Is the assumption at all justified that the division when attested in such texts forms at the same time the scheme on which the actual exposition of the argument of the texts concerned is based? In other words: Can we expect to find the quadruple division attested in an Ayurvedic text in such a manner that this text itself is explicitly or implicitly divided into exactly four systematic parts or chapters corresponding to the four vyūhas?
As to this, it seems expedient to proceed with caution; for, a first warning against such an assumption is the observation, simple though it is, that, in any case in Yoga texts, the caturvyūhatva is at best stated to be a divisionary scheme, but that it is nowhere actually made the basis of a corresponding disposition of the material: there is no Yoga text in which this systematic division is taken seriously in such a manner that the exposition actually follows this scheme. A second warning is given by another division of the science of medicine referred to in medical texts themselves, viz. the octopartite division of therapy. Yet, it is not these eight parts of Ayurveda as listed e.g. in the Suśrutasamhitā I, 1, 7, or in the Astāngahşdaya I, 1, 5cd-bab, that matter in the present context, but a fact stressed by J. Filliozat in the introduction to his recent edition of the Yogaśataka, a medical text attributed to Nāgārjuna 42; for, referring to an article of C. Vogel's 43 he points out that « dans le titre des Astāngasamhitā et Astāngahrdayasamhitā de Vāgbhata, astānga désigne la science médicale en général et non les parties de l'ouvrage >> 44.
What both these observations come to is that it would indeed be well to distinguish between a division of the science of medicine taught somewhere, on the one hand, and the actual internal organisation of medical texts, on the other. Therefore, one has to reckon with the possibility that all one finds in medical texts is simply a reference to the scheme, perhaps even in an offhand manner. The degree of probability that at least this kind of evidence can actually be found is, however, quite high, for the quadruple division is, according to C. Vogel 45, referred to by yet another witness, and, to be sure, an independant one, viz. Bu-ston (1290-1364) who in his comprehensive « History of Buddhism », coming to speak of medical works, says 46: « As for the works on medicine, they teach four (topics): disease, cause of disease, medicament as antidote to disease, and method of curing thereby ». Though the succession of the last two members is, again, reversed and the Tibetan gso-ba'i spyod-lam does not fully correspond to the Sanskrit term
42. Publications de l'Institut Français d'Indologie, no. 62, Pondichéry, 1979.
43. Viz., On Bu-ston's View of the Eight Parts of Indian Medicine, in IIJ, VI (1962), pp. 290-95.
44. Op. cit., p. IV. In Vogel's article (the reference should be «p. 291, n. 3», not «n. 2 »), however, this is not expressly stated though perhaps implied.
45. The reference is to his article mentioned in fn. 43. 46. Quoted from Vogel, loc. cit., p. 290.

Page #21

On the Quadruple Division of the Yogaśāstra
309
ārogya 47, it is highly probable that what Bu-ston has in mind is the same quadruple division as drawn upon in our Yoga texts.
After consulting well-known works of secondary literature on Indian medicine where, however, the quadruple division is apparently not mentioned at all, I decided to turn for help to some of the specialists in this field of learning instead of setting out myself on possibly a wildgoose chase in the impenetrable jungle of the extant Ayurvedic texts (begging pardon for the all too mixed metaphor). It was my colleague R. E. Emmerick who drew my attention to Carakasamhitā, sūtrasthāna 9.19, i.e. to the following verse:
hetau linge praśamane rogānām apunarbhave / jñānam caturvidham yasya sa rājārho bhișaktamaḥ //.
Cakrapāņidatta does not explain this verse, but there can hardly be any doubt that the « fourfold knowledge » spoken of here is practically identical with our quadruple division of the Cikitsāśāstra; for the expression apunarbhava can be equated with our ārogya, the only difference being that by the former health as something regained is paraphrased as «the not-occurring-again [of a disease]»; further, the term praśamana closely corresponds to our bhaisajya, at least if taken to denote « that by which [a disease] is tranquillized, i.e. cured »; and, finally, as regards the expression linga, here it cannot but mean « symptom », and thus it corresponds to our roga, only that it is not the disease itself that is denoted, but its « characteristic sign(s) », and this is clearly done with the aim of bringing out into proper relief an essential virtue of any physician, viz. to make a proper diagnosis by the symptoms. Particularly noteworthy, however, is that it is not every physician who is expected to have this fourfold knowledge, but only a bhisaktama, « an excellent physician », and that he is in addition stated to be « fit or worthy of [treating] a king »; for, this cannot but imply that the common variety of a bhişaj ordinary people have to depend on need not necessarily be equally well versed in all four parts of the science of medicine. Nevertheless, what is attested in this verse of the Carakasamhitā is clearly a systematic division of the Cikitsāśāstra into four parts and, to be sure, a division which may, inspite of certain terminological and conceptional differences, well be taken to be virtually identical with that referred to by the author of the YBhāşya, etc.; on the other hand, however, it has to be stressed that even if the verse
47. Viz. insofar as for obvious reasons the medical treatment itself is given prominence and not the aim it has to achieve.

Page #22

310
A. Wezler
forms an original part of the Carakasamhitā 48 the quadruple division is, strangely enough, given but an extremely marginal importance.
To an eventual second instance 49 of the quadruple division being attested in a medical text my attention was drawn by K. G. Zysk's review 50 of Filliozat's edition and translation of the Yogaśataka, referred to above. Zysk critically examines Filliozat's rendering of verse 2 which runs thus:
parīksya hetvāmayalaksanāni cikitsitajñena cikitsakena /
nirāmadehasya hi bheşajāni bhavanti yuktāny amstopamāni //; he not only proposes an, indeed, more appropriate translation for hetvāmayalaksaņāni, viz. «the characteristics of the diseases and [their] causes » (which should, however, perhaps be replaced by the cause[s] and the characteristics of diseases »), but also points out that the Tibetan translator, apparently equally at a loss with the expression nirāmadehasya, renders the second line thus: «The body becomes free from disease by the combination of medicines which resemble ambrosia ». Zysk rightly adds the remark that « the Tibetan translation seems more acceptable », but in addition he refers approvingly to Filliozat's note « T, au 3e pada, a remplacé āma par nad, āmaya, et entend nirāma "sans dénutrition " comme nirāmaya " sans maladie" ». However, I am not sure that this assumption is justified; for according to MonierWilliams 51 Indian lexicographers list an āma,« sickness, disease », which is referred to also by Edgerton 52. On the other hand one wonders how the Tibetan translator thought that the line construes and whether his interpretation is at all possible.
Yet I do not want to enter on my part into a detailed discussion of the philological problems involved here; instead, I should like to confine myself to referring to the Tibetan translation; for, it is there that we find the four systematic parts of the science of medicine reflected, to wit, in a manner that reminds us of the verse quoted above from Caraka.
Both these references to the quadruple division — and, of course, first of all the more unequivocal one of the Carakasamhitā — are no
48. Apart from the fact that this verse is not commented upon by Cakrapāņidatta (a fact which by itself is, of course, by no means conclusive), what creates suspicion is that the verse does not really fit into the context so that it might well have been added later since in the immediately preceding verse a special designation, viz. prānābhisära, is taught for a bhișaj who possesses certain (other) qualifications.
49. The « fourfold physicians » (caturvidha vaidyāḥ) mentioned in Mbh (Poona ed.) 12.69.57 are most probably to be taken in accordance with Nilakantha's explanation (vişaśalyarogakrtyäharā) as reference to four different types of medical >> specialists. Cf. also R. P. Das' review of G. N. THITE, Medicine. Its Magico-Religious Aspects according to the Vedic and Later Literature, in IIJ, 27 (1981). p. 235 f.
50. In IIJ, 23 (1981), pp. 309-13. 51. A Sanskrit-English Dictionary, p. 146. 52. Buddhist Hybrid Sanskrit Dictionary s.v. 2 āma.

Page #23

On the Quadruple Division of the Yogaśästra
311
more than bare references, and the result of our examination of Ayurvedic texts is accordingly rather disappointing, for in them the quadruple division is not given the fundamental importance peculiar to it in Yoga texts. One is, therefore, left with two alternatives, viz. either that it was these Yoga thinkers who realized the full importance of the division, or that what these Yoga thinkers refer to is but a later development in Indian medicine 53. However, this much is clear beyond doubt: the fourfold scheme as such, even if not also applied in writing medical works, did actually exist.
This conclusion is additionally corroborated by a clearly independent witness not called into evidence so far. What I am referring to is an hitherto unpublished passage in the śrutamayī bhūmi of the Yogacārabhūmi which was pointed out and given to me by my friend L. Schmithausen 54. It runs thus: cikitsa katamā sā caturākārā veditavya tadyatha äbädhakausalyam abädhasamutthanakausalyam utpannasyābādha[...]sya prahāṇakausalyam prahinasyābādha[...]syāyatyām anutpadakausalyam esam ca kausalyanam vibhango yathāsūtram eva veditavyah // 55. « Which of many things (or: Of what kinds) is medical treatment? It has to be regarded as having four forms, viz. skill [in diagnosing] the disease, skill [in determining] the origination (i.e. cause) of the disease, skill [in applying] the means of getting rid of the disease which has arisen [and finally] skill [in ensuring] that [the disease] which one has got rid of does not arise again in future. Moreover, as for the detailed explanation of these [medical] skills, it [is not given here, but] has to be known precisely according [to what has been taught] in the [respective] Sūtras ».
Though the terminology is palpably different, there can hardly be any doubt that the quadruple division taught is practically identical with that referred to in Caraka, and thus, ultimately, also with that drawn upon in our Yoga texts, the four kausalyas corresponding to roga, rogahetu, bhaiṣajya and arogya, respectively 5%. The last sentence of the passage quoted is of particular importance; for it cannot but be taken as evidence for the fact that this division was expounded already
53. In this latter case one would have to assume that it was not given due attention in later medical texts because the tradition of how a medical treatise should be dispositionally structured had become fixed to such an extent that alterations were not possible (any more).
54. On this occasion I should also like to acknowledge with sincere gratitude the vivid interest he showed in the present study in general and its part 4 in particular, for much of the material referred to there I owe his pigeon-hole.
55. The passage is found in the Patna MS of the Sravakabhūmi at 2 B 6-2 [= 96 a 2 of the Patna MS of the Yogacarabhūmi]. The former adds vi at the very beginning, which, however, seems to have been deleted, and the latter reads draṣṭavyā in the place of veditavya and anu[...]tpăda'. By square brackets deleted akşaras are indicated.
56. Particularly close is obviously the similarity with the Carakasamhita (cf. the term apunarbhava in the latter); see below p. 323.

Page #24

312
A. Wezler
in a Sūtra, to be sure a Buddhist Sūtra. In view of the probable date of the Yogacărabhūmi, this Sūtra, according to Schmithausen, can hardly belong to a period after the beginning of the 4th century, but is probably earlier.
In the light of this additional evidence the following observations and assumptions can be made:
a) that the science of medicine or medical treatment can be systematically divided into four parts was common knowledge of educated people, or at least it was widespread;
b) this knowledge is attested obviously independently in Brahmanical as well as Buddhist 57 sources the most ancient of which date back to the first half of the 1st millennium;
c) it seems natural to assume that the division was first conceived by a medical author and only later referred to and utilized by others, i.e. applied to other Ŝāstras also which, though significantly different, nevertheless exhibited a specific similarity with the science of medicine.
4. In what follows, Buddhism again may serve as a cue for focussing attention on still another problem raised by the theory of the quadruple division of the Cikitsāśāstra, a problem which is perhaps even more important than all those discussed so far in the course of the present study. What I have in view is the assertion found in not a few works on or expositions of (early) Buddhism, viz. that this fourfold division of medicine it was that inspired the Buddha to his << Four Noble Truths ». Thus e.g. E. Frauwallner 58 simply states as though it were a fact established beyond any doubt, and without giving any reference, that « the fourfold division of the truth discovered [by the Buddha]... is borrowed from the medical method ». Equally apodictic is H. Zimmer who in his << Philosophies of India » 59 remarks: << Following the procedure of the physician of his day inspecting a patient, the Buddha makes four statements concerning the case of man. These are the socalled "Four Noble Truths" which constitute the heart and kernel of his doctrine ».
57. Of the two other Buddhist authors referred to in the foregoing, viz. Bu-ston and Nāgārjuna, at least the former might have derived his knowledge from the lost Sūtra and not directly from a medical text. In any case, there is no clear evidence in favour of the assumption that the Buddhist sources depend on the Brahmanical or vice versa as regards the quadruple division of medicine.
58. Geschichte der indischen Philosophie, Bd. I, Salzburg, 1953, p. 184 History of Indian Philosophy, tr. by V. M. BEDEKAR, vol. I, Delhi, 1973, p. 146.
59. New York, 1951, p. 467; the German translation, Zürich, 1961 (= Frankfurt, 1973, Suhrkamp Taschenbuch Wissenschaft 26), p. 417 f. Cf. also H. ZIMMER, Indische Sphären, Zürich-Stuttgart, 1963, pp. 219 and 221.

Page #25

313
On the Quadruple Division of the Yogaśāstra
That this opinion has become a commonplace with many a Buddhologist can also be seen in A. Bareau's monograph 60 where it is said: << The Four Noble Truths are, as regards their classification, obviously 61 taken from the dialectics of an old Indian medical school. Starting from the disease one is led to its origin, from this to its suppression, i.e. the restoration of health, and finally to the medicine which brings it about >>; or, to give just one more example: D. Schlingloff 62, too, is thoroughly convinced that << just as in the Old Indian medicine the theory of disease, of the origin of disease, of the suppression of disease and of the ways leading to the suppression of disease was developed, here [i.e. in the Four Noble Truths], Suffering, its origin, its suppression and the way leading to its suppression are spoken of ».
Other scholars, however, are a bit more cautious in that they confine themselves to merely pointing out the close similarity between the Four Noble Truths and the corresponding division in medicine.
Thus E. Conze states in his « Buddhism. Its Essence and Development » 63: « The holy doctrine is primarily a medicine. The Buddha is like a physician. Just as a doctor must know the diagnosis of the different kinds of illness, must know their causes, the antidotes and remedies, and must be able to apply them, so also the Buddha has taught the Four Holy Truths, which indicate the range of suffering, its origin, its cessation, and the way which leads to its cessation ». A similar statement is found in K. Mizuno's book 64, viz.: « The Four Noble Truths were taught on the basis of a principle of healing spiritual suffering and misery that is similar to the principles that doctors follow in curing illness of the body »>.
Yet it is with the first kind of statements that I am concerned here. It is, I think, not necessary to try to collect more such examples from secondary sources on Buddhism; and it would be rather futile to rack one's brains about the question who copied whom in this case. Instead, it should be stated in summing up that it is evidently a widespread conviction of scholars of Buddhism that it was this systematic division of the science of medicine that served as a model for the conception of
60. Der indische Buddhismus, in: Die Religionen Indiens III (Die Religionen der Menschheit Bd. 13), Stuttgart, 1964, p. 33.
61. If the unknown French original of the German « augenscheinlich »> (which I render by « obviously ») is used to indicate that there obtains here, to some degree at least, an incertitude (the German expression does indicate this), Bareau's position would have to be classified with the more guarded ones quoted below.
62. Die Religion des Buddhismus I. Der Heilsweg des Mönchstums (Sammlung Göschen Bd. 174), Berlin, 1962, p. 70. H. J. GRESCHAT, Die Religion der Buddhisten (Uni-Taschenbücher 1049), München, Basel, 1980, p. 66, expresses himself in such a vague manner that it is not possible to decide if he is to be classed with the foregoing group of scholars or with the subsequent one.
63. Oxford, 1951, p. 17. In the German translation (Urban-Bücher 5), Stuttgart, 19562, the corresponding passage is found on p. 14.
64. The Beginnings of Buddhism..., transl. by RICHARD L. GAGE, Tokyo, 1980, p. 43.

Page #26

314
A. Wezler
the Four Noble Truths. One may find it surprising or even significant that none of the specialists quoted until now thought it, if not necessary then at least, desirable to adduce the reasons, too, that could convince their readers that what they have to do with is not an arbitrary and fanciful assumption, but a well-founded hypothesis. Yet, it is by no means a peculiar feature of Indology that an allegedly true and certain piece of knowledge is handed down from generation to generation, not because of a general lack of the critical faculty, but simply because nobody happened to take a second look at it, and in the present case it is but fair to admit that the assumption is in fact suggestive to such an extent that one is easily lulled into a sense of absolute security. Nevertheless, it bears renewed discussion.
4.1. The first question to be asked, then, is this: Who was the first scholar to make this assumption and what gave him this idea? By a note of Oldenberg's 65 one is referred to the original source, viz. H. Kern's « Geschiedenis van het Buddhisme in Indië » 66 that was translated into German by H. Jacobi in 1882 67. It is there that for the first time in a foot-note it is apodictically stated that the Four Truths are borrowed from the art of healing » 68. In the text itself of Kern's book a statement to the same effect is found at p. 367 (=469 in the German translation), and surprisingly enough followed by a reference to the YBhāsya on YS 2.15 from which the relevant passage is quoted in translation. Though it is not expressly stated one cannot but draw the conclusion that it was precisely this passage that gave Kern the idea of the Four Noble Truths having been borrowed from the art of healing. This conclusion is in fact corroborated by Kern's « Manual of Indian Buddhism » 69 where it is said: « It is not difficult to see that these four Satyas are nothing else but the four cardinal articles of Indian medical science, applied to the spiritual healing of mankind, exactly as in the Yoga doctrine » (in a footnote a portion of the text of the YBh on YS 2.15 is given), and two passages from the Lalitavistara are quoted to vindicate the assertion that « this connection of the Aryasatyas with medical science was apparently not unknown to the Buddhists themselves » 70. The Lalitavistara,
65. Viz. in his: Buddha, Sein Leben, seine Lehre, seine Gemeinde, 1923, p. 236. It should be noted that J. Filliozat in the article mentioned in fn. 99 below arrives at the same conclusion.
66. Haarlem, 1882 (Eerste deel), 1884 (Tweede deel).
67. Der Buddhismus und seine Geschichte in Indien. Eine Darstellung der Lehren und Geschichte der buddhistischen Kirche, 2 Bde., Leipzig, 1882-84.
68. Viz. fn. 4 on p. 207 of the first part (corresponding to p. 265 in the German translation).
69. (Grundriss d. Indo-Arischen Philologie u. Altertumskunde III.8) Strassburg, 1896; the quotation is from p. 46 f.
70. The expression « applied to >> and the reference to the YBhāşya preclude, I think, the possibility of taking Kern's remarks as meant to point out a systematic similarity only.

Page #27

On the Quadruple Division of the Yogaśāstra
315
however, is of so late a date that it could not by any means be regarded as evidence for the fact that the quadruple division of the science of medicine antedates the historical Buddha or was even developed in his youth. Yet in reality the division as such is not mentioned at all, nay not even alluded to in this text; what is said in the two passages 71 pointed out by Kern is merely that the Buddha is avaidyarāja in that he is a pramocakaḥ sarvaduḥkhebhyah or a sarvavyādhipramocakah; that is to say, he is but compared to a physician.
4.2. But before continuing the critical discussion of Kern's remarks, and the arguments brought forward by others, it is advisable to reflect on the implications the assumption examined here cannot but have. This might seem superfluous, but is nevertheless necessary since nowhere in the relevant secondary literature do I find them explicitly stated 72 so that one cannot help suspecting that they were not clearly realized by Kern and those depending on him directly or indirectly. These implications are, of course, that the science of medicine was either already before the Buddha or at least in his times divided in this manner and that the Buddha knew this systematic division.
Our knowledge of the historical person called the Buddha does not permit us to answer the question whether at all or to what extent he was familiar with the contemporary science of medicine. Therefore it is not possible to check whether the second condition can be regarded as fulfilled. As for the first and basic implication, one would have to look for pre-Buddhist medical literature. As even the most ancient of the so-called Ayurvedic texts are of later origin, it could only be one of the Vedic texts 73 where one could strike on a relevant statement. Nobody, however, has so far come up with any such reference. This holds good for the other possibility also, viz. that the quadruple division be attested in a non-medical text of pre-Buddhist origin. Therefore, it has to be stressed that all we actually have are at best references in
71. Viz. Lal. V, p. 448: ut panno vaidyarājah pramocakah sarvadunkhebhyah, pratişthāpako nirvänasukhe, nişannas Tathāgatagarbhe Tathāgatamahädharmarājāsane, and p. 458: cirāture jivaloke klesavyādhiprapidite / vaidyarāt tvam samutpannah sarvavyādhipramocakah 17.
72. Laudable exceptions to the rule, however, are H. Oldenberg and J. Filliozat. Oldenberg (in the foot-note referred to above in fn. 65) clearly dissociates himself from Kern's opinion with the remark that « it will not be possible to ascertain whether, as regards the fourfold division, Buddhism is the borrowing side », a remark which does not simply show scepticism, but is most probably due to 0.'s being well aware of the fact that there is no pre-Buddhist medical text to attest the division; Filliozat in his article (cf. below fn. 99) says more clearly: « Il eut fallu prouver que cette médecine les [= les quatre vérités] possédait avant le Bouddhisme; Kern ne l'a point fait ».
73. From the description of Indian medical literature as given e.g. by J. Jolly (in Medicin, Grundriss der Indo-Arischen Philologie u. Altertumskunde III.10, Strassburg, 1901) it follows that « we have no medical texts of the intervening period [between the Vedic literature and the texts of Caraka and Suśruta] » as stated by A. L. Basham (The Wonder that was India, London, 1954, p. 499).

Page #28

316
A. Wezler
post-Buddhist Indian literature. To say that the references discussed in the preceding part of this article, including the YBhāşya, so important to Kern, do furnish the evidence needed, would be nothing but an arbitrary assertion, merely meant to warrant an assumption that cannot be supported otherwise.
4.3. But what about the « evidence » found in Buddhist texts themselves? As regards the Lalitavistara passages, all that can be learned from them is that the Buddha was compared to a physician by his adherents. And this can hardly be regarded as a new and important piece of information. For, firstly, the idea that God 74 or a particular god or the propounder of a doctrine of salvation helps men by healing physical and/or spiritual ailments is so widely spread that there is nothing strange in that such a comparison may have been drawn by Buddhist authors, too, without their possessing any tradition that the Buddha himself had actually learnt from medicine, and shrewdly applied its division to his own teaching; and, secondly, it is already in much earlier Buddhist texts that the Buddha is characterized or sometimes even made to say of himself that he is an anuttaro bhisakko sallakatto, « an unsurpassed physician, (the best) surgeon » 75. This comparison (echoed as it were in statements like those of Conze and Mizuno quoted above) is not only met with rather frequently already in early canonical texts, but has obviously also proved a quite fruitful idea, variously developed in later times, as was shown recently by R. Birnbaum 76 who aptly remarks 77 that « the Buddha frequently made analogies to disease and healing to explain various facets of his teaching ». Nevertheless, one cannot but observe that in the passages inspected or referred to until now the Four Noble Truths are not mentioned at all, not to speak of explicitly paralleling them with corresponding systematic parts of the science of medicine. And, to be sure, passages like Mil 247.11 where a
74. Cf. also the recent study of G. BUDDRUSS, Khowar-Texte in arabischer Schrift (Akademie der Wissenschaften und der Literatur, Mainz), Wiesbaden, 1982, p. 12. For an interesting comparison between «God the Lord >> and a physician in the context of a theodicy-like discussion see BHĀSARVAJÑA'S Nyāyabhūsana, p. 458.16 ff. As noted by W. Halbfass (Studies in Kumārila and Sankara, Reinbek, 1983, p. 15, cf. also fn. 81), «the samsāramocaka [too] presents himself as a benevolent physician in the wider context of samsāra », and « reference is made to the expertise of the good doctor who knows that sometimes he has to apply harsh means to bring about a change for the better ». God is called a medicament (ausadha) by Madhva in his Bhāşya on Taitt. Up. 2.2.
75. It 101.15-16; cf. (without bhisakko) Sn 560 and (sallakatto only) Sn 562. The first of these passages is referred to Mil. 215.11; as for the well-known parable of the man wounded by an arrow cf. M 1.429 as well as M II 216, Mil 169.9 ff., 247.10 ff.; for a kusalo bhisakko in a simile cf. A III.238.5 ff., Mil 229.5 ff.; bhisakko is called an adhivacana of the Buddha A IV.340.5 ff.; the Buddha is compared to a kusalo vejjo Pj 1.21.19 f.
76. The Healing Buddha, Boulder (Co.), 1979. 77. Viz, on p. 15; cf. also the subsequent sections.

Page #29

On the Quadruple Division of the Yogaśästra
317
physician (bhisakka) is characterized as roguppattikusala, well-acquainted with / competent in discovering the origin of disease »>, is by no means conclusive.
But what about the passage in Buddhaghosa's Vism 512.7-9 referred to by Birnbaum 78 which runs thus: rogo viya ca dukkhasaccam, roganidanam iva samudayasaccam, rogavūpasamo viya nirodhasaccam, bhesajjam iva maggasaccam? It is true, the Four Noble Truths are here clearly compared to corresponding parts of medical science; but, on the other hand, one must not forget that the famous commentator is separated approximately by a millenium from the Buddha himself and one cannot overlook that this is but one in a series of different analogies given by Buddhaghosa without (explicit) reference to canonical texts and that it does not at all imply that the Truths were « borrowed from the medical method ».
There is, however, a further reference by Oldenberg which has still to be followed up, viz. that to an article of L. de la Vallée Poussin 79 who after having quoted the relevant lines from Kern's « Manual of Indian Buddhism » 80 draws on his part attention to some more material, viz. two passages in the Bodhicaryāvatāra and an explanation found in Yasomitra's Abhidharmakośavyākhyā. Among these quotations the two former ones are but reformulations of ideas attested already in canonical Pāli texts in the context of the comparisons mentioned above (p. 316). What the author is concerned with is to lay stress (1) on the extreme foolishness of him who though suffering from a disease refuses the help of a person capable of healing it and (2) the (correct) observation that << there is no physician but cures disease with some pain in the treatment »> 81. The latter reference, however, is indeed of such a kind
78. Viz. in fn. 37 on p. 22.
79. JRAS, 1903, pp. 578-80. The two passages quoted from the Bodhicaryāvatāra are II.55 ff.:
itvaravyadhibhito 'pi vaidyavakyam na langhayet /
tatra sarvajñavaidyasya sarvasalyapahāriṇaḥ / vākyam ullanghayāmīti dhig mam atyantamohitam //, and VIII.22 ff.:
sarve 'pi vaidyaḥ kurvanti kriyāduḥkhair arogatām / tasmäd bahuni duḥkhani hantum şodhavyam alpakam // kriyām imām apy ucitām varavaidyo na dattavan / madhurenopacāreņa cikitsati mahāturan.//
adau sākādidane 'pi niyojayati nayakaḥ /
tat karoti kramat paścãd yat svamämsäny api tyajet // 80. Practically all of them are quoted also above, p. 315 and fn. 71.
80a. The same idea is also met with e.g. in Samadhirajasūtra 9.43-46.
81. This is de la Vallée Poussin's rendering of the first line of his second quotation from the Bodhicaryavatāra (cf. above fn. 79), the gist of which, however, is that the Buddha differs from a physician precisely in this respect. The idea expressed in the first line was appealed to also by the so-called Samsaramocakas; cf. fn. 74 above.

Page #30

318
A. Wezler
that it fully warrants Oldenberg's remark 82 that it « compares quite expressly the four truths to the quadruple medical teaching of disease, its origin, its healing [and] the prevention of its resurgence ». For what Yaśomitra 83 says is:
vyādhim drșțveti. vyādhir duḥkha-satyasyopamānam. tannidānam samudaya-satyasya. tat-kşayo nirodha-satyasya. tadbhesajam mārgasatyasyopamānam 84. sütre 'py esa satyānām dřstāmga iti. Vyādhyādisūtre 85. kathani. caturbhir amgaiḥ samanvāgato bhișak salyāpahartā 86 rājārhaś ca bhavati rāja-yogyaś ca rājāmgatve ca samkhyām gacchati 87 katamais caturbhih. ābādha-kušalo bhavati. ābādha-samutthāna-kuśalah. ābādha-prahāna-kuśalah. prahinasya cābādhasyāyatyām anut pāda-kusalah. evam eva caturbhir amgaiḥ samanvāgatas Tathāgato 'rhan samyak-sambuddho 'nuttaro bhişak 88 salyāpahartety ucyate. katamaiś caturbhih. iha bhiksavas Tathāgato 'rhan samyaksambuddha idam dunkham ārya-satyam iti yathābhūtam prajānāti. ayam duḥkha-samudayaḥ. ayam duḥkha-nirodhaḥ. iyam duḥkhanirodha-gāmini pratipad ārya-satyam iti yathābhūtam prajānātīti.
As stated by de la Vallée Poussin 89 the Sūtra referred to by Vasubandhu and quoted by Yaśomitra is « d'après la version de Hiuan-Tsang,
82. Loc. cit., cf. fn. 65.
83. I quote from Sphuţartha Abhidharmakośavyākhyā by Yaśomitra, ed. by U. Wogihara, Tokyo, 1971, p. 514 f.; see, however, fn. 85. Mr. Yoshifumi Honjo whom I met at the XXXI CISHAAN in Japan kindly drew my attention to the fact that a more complete quotation from that very Sūtra is found in Samathadeva's Abhidharmakośapäyikā, viz. Peking Tanjur, Thu 32 b 6 ff.
84. From sūtre 'py esa onwards this passage is also quoted by DE LA VALLÉE POUSSIN, in note 4 on p. 121, of his translation, L'Abhidharmakośa de Vasubandhu traduit et annoté..., cinquième et sixième chapitres, Paris, 1925, the final part, however, being abbreviated, viz. iha bhikṣavas tathāgato... idam duḥkham āryasatyam iti yathābhūtam prajānāti...
85. In his article (cf. fn. 79) de la Vallée Poussin leads vyādhisütre; the reading as given by Wogihara is, however, also found in the footnote in de la Vallée Poussin's translation of the Abhidharmakośa, and has to be accepted as correct; interpret: « in the Sūtra (dealing with) disease (cause of disease], etc. ».
86. The MS has here and infra) kalpāpah"; the emendation, proposed also by de la Vallée Poussin in his article and accounted for by the remark «I suppose that the old Nepalese ligature has been misunderstood by a modern copyist>>, is most convincing, not only because of the Tibetan equivalent, viz. zug-rnu 'byin-pa, but also and above all in view of the Pāli parallels (see above p. 316 and fn. 75).
87. The following sentences up till ābādha-prahāna-kuśalah are in Wogihara's edition printed in italics and thus characterized as a quotation. This cannot but be a mistake, for evidently the quotation from the « Vyādhyādisātra >>, the beginning of which is found already in the Kośa itself, extends as far as yathābhūtam prajānātiti, the iti indicating its completion, because the reference in the Kośa is introduced by the remark sūtre 'py eşa eva satyānām drstānto darsitah «this very analogy of the [Four Noble] Truths is shown also in a Sütra».
88. In his article de la Vallée Poussin reads anuttarabhisak.
89. Viz, in the foot-note to his translation of the Abhidharmakośa referred to in fn. 84.

Page #31

On the Quadruple Division of the Yogaśāstra
319
le Sūtra du bon médecin; d'après Paramārtha, le Sūtra de la comparaison du médecin (Samyukta 15.19; 17.14) 90 ». Unfortunately, de la Vallée Poussin does not give any further explanation on the identity of the « Sūtra du bon médecin ». His second reference, however, can be traced, namely with the help of the Hôbôgirin 91 where the relevant passage from the Samyuktāgama as preserved in Chinese translation is rendered into French. It clearly shows a close similarity to the quotation by Yaśomitra, but does not, on the other hand, seem to agree fully with it 92. On closer examination, however, it becomes evident that what Demiéville — or one of his collaborators — actually did was to give but a free paraphrase of the Chinese translation, intermingling later portions of the « detailed explanation » (vibhanga) with earlier parts of the Sutra. When the Chinese translation itself (T 99, no. 389) is compared with the Sanskrit text quoted by Yaśomitra, one cannot but arrive at the conclusion that it is precisely this Sanskrit text that was translated by Gunabhadra 92a, i.e. that what Yaśomitra quotes is in fact nothing else but this Sūtra of the Samyuktāgama. The Chinese translator, however, seems to have read rājabhogyas instead of rājayogyas and the Chinese text obviously contains a corruption, not recognized by Demiéville: the character standing between that denoting « King » and that equivalent
90, I do not know which edition is referred to by de la Vallée Poussin; by the former reference, however, T 99, no. 344, might be meant and by the latter T 99, no. 389. In the latter passage three « ailments are referred to, viz. kāma-, bhavaand avidyā- (roga, or a similar expression) (evidently under the influence of the three asravas); however, it is not a comparison that is drawn here between the « Four Noble Truths » and medicine, but the concept of ailment is only used metaphorically in a context modelled after the Truths.
91. Viz. s.v. byô, p. 224 ff.
92. For according to Demiéville (loc. cit., p. 228) it runs thus: « On appelle un grand roi des médecins celui qui réalise les quatre Essences suivantes: (1) Bien connaitre la maladie, i.e. les différentes sortes de maladies...; (2) bien connaître l'origine de la maladie: due au Vent, ou au Flegme..., ou aux sécrétions salivaires..., aux différentes espèces de froid, à un fait actuel, à la saison...; (3) bien connaître le Contrecarrant de la maladie: onguents, expectorants ou vomitifs..., évacuations inférieures (purgatifs), instillations nasales..., fumigations..., sudorifiques...; (4) bien savoir traiter la maladie de telle façon qu'ayant été traitée il n'y ait aucun danger de récidive... de même. le Tg.-Arhat-So so buddha, ce grand roi de médecins, réalise quatre vertus grâce auxquelles il soigne les maladies des Etres; en effet, il connaît les quatre Saintes Vérités: (1) de la Douleur...; (2) de la Formation de la Douleur...; (3) du Barrage de la Douleur...; (4) du Chemin de ce Barrage ... Le médecin Mondain ne connaît pas vraiment, tels qu'ils sont..., les Contrecarrant radicaux (fondamentaux), à savoir le Contrecarrant radical de la Naissance..., et celui de la Vieillesse, de la Maladie. de la Mort, du chagrin, des plaintes, des douleurs, des Tribulations...». The differences, marked here by italicizing the corresponding elements in the translations, are, on the other hand, of such a kind that (except perhaps for the final portion which may simply have been left out of his quotation by Yasomitra) they can easily be accounted for by the assumption that they represent but later additions of an explanatory character to a text which may well have been identical with that attested in the Abhidharmakośavyākhyā. See, however, also below p. 320.
92a. Viz. 436-443 A.D. according to E. LAMOTTE, Histoire du Bouddhisme Indien, Louvain, 1958, p. 169.

Page #32

320
A. Wezler
to "arha, viz. *, seems rather to be an error for Ź; if the latter is accepted, there is full agreement between the Sanskrit original and the Chinese translation also as regards the expression rājārha! The comparison further shows that the vibhanga portion of the original Sūtra, which preceded the sentence beginning with evam eva caturbhir amgaih, was simply left out by Yaśomitra, and it was left out in the Yogācārabhūmi (s. above p. 311) as well, where the reader is, however, referred to the original source precisely as regards this « detailed explanation !
But there is a very similar passage in a second, probably older, yet rather inaccurate translation of the Samyuktāgama, viz. T 100, no. 254, where also the vibhanga is missing. Therefore, one cannot be sure which of the two « versions » of the SĀ are actually referred to by the Yogācārabhūmi and Vasubandhu or quoted by Yaśomitra, respectively. In any case, the passage as a whole is missing in the Pali SN, and there is hence a high degree of probality that it forms but a later addition; for it is quite impossible to discover a motive for a transmitter dropping it. Yet, whatever the exact historical relation may be in which the different sources stand to each other, this much has to be accepted as established: Hīnayānistic Buddhist texts knew of a quadruple division of the science of medicine and compared the teaching of the « Four Noble Truths » to it.
The most important source is, of course, the Vyādhisūtra passage as quoted in the Abhidharmakośavyākhyā: it is preserved in Sanskrit and it is here that the Four Noble Truths are expressly compared to the corresponding systematic parts of the science of medicine. The parallel drawn is, however, not completely convincing, as already noted by de la Vallée Poussin who adds the remark: «You 93 will observe that, according to this sūtra, the third satya is the way (mārga, upāya) to the destruction of the disease, and the fourth the way to its notappearing-again (a-punarbhava). The scholastic point of view, so far as I know, is different ».
One might feel some reluctance to follow de la Vallée Poussin as regards his interpretation of the expression prahīņasyābādhasyāyatyām anut pādakuśalah; for it is highly improbable that it refers to a « way »! However, the impression one cannot but gather is that there is no full correspondence between the last two « medical skills » and the last two of the Four Noble Truths. And this becomes particularly evident in the case of the fourth « Noble Truth » which is clearly different from what is meant by prahinasyābādhasyāyatyām anut padakuśalah. One further wonders whether de la Vallée Poussin's adding the term apunarbhava by way of explanation is merely due to a free association or else evoked by a particular passage in an Indian medical text. For, it is exactly the term which is used in the verse from the Carakasamhitä drawn upon
93. De la Vallée Poussin's short article (cf. fn. 79) has actually the form of a letter addressed to Professor Rhys Davids.

Page #33

On the Quadruple Division of the Yogaśästra
321
above (S 3.2.). In any case, the fourth systematic part of the science of medicine as taught in the Vyādhisūtra, conceptionally agrees fully with what is called rogānām apunarbhavaḥ in Caraka, and hence with ārogya. In passing it may be noted that the slight unevenness the comparison in the Vyādhisūtra thus shows was apparently felt among others (cf. Hôbôgirin p. 230r) by Aśvaghoşa also; for verse 41 of Canto XVI of his Saundarānanda reads as though it were a deliberate improvement on it, viz.
tad vyādhisamjñām kuru duḥkhasatye doșeșv api vyādhinidānasamjñām / ārogyasamjñām ca nirodhasatye bhaişajyasamjñām api mārgasatye // 93a.
But there are more peculiarities to be observed in the passage quoted by Yaśomitra. It should be emphasized that Yasomitra, or rather, to be sure, already Vasubandhu, speaks of a drstānta only. That is to say, both of them took the relation between a physician and the Tathāgata as pointed out in the Sūtra to be but a comparison meant to exemplify the particular and extraordinary qualities of the Buddha as one healing not a physical illness, but the ailment of existence itself. Evidently they did not know of any tradition - nor did it occur to them independently — that it was this quadruple division of medicine that inspired the Buddha! In fact, this it is what all relevant passages in Buddhist texts amount to the Buddha is merely compared to a physician or the doctrine to a medicine, etc.; nowhere can be found any traces of an awareness that the Buddha in conceiving the Four Noble Truths could have drawn on a similar systematic division of the Cikitsāśāstra! Of course, one might object here that it is hardly to be expected that a Buddhist monk even if he were fully aware of it, should have overtly admitted that the kernel of the Buddha's teaching is but an adaptation of principles of medical science. Nevertheless, it is worth taking note of that for the Buddhist tradition itself this similarity was obviously never anything more than a comparison. On the other hand I cannot but add the following remark here: the fact that the equation of the Four Noble Truths with the four parts of medical science is in fact attested in Buddhist literature, though with varying degrees of agreement in conception and terminology, fully corroborates what has been said above (p. 303) in my critical discussion of Hacker's view on the relation between the caturvyūhatva and the Four Noble Truths: Hacker's view does not only not stand an independent critical test, but
93a. Full parallelism can be observed also in Ratnagotravibhāga, 4.52; this kārikā is translated by J. TAKASAKI, A Study on the Ratnagotravibhāga (Uttaratantra). Being a Treatise on the Tathāgatagarbha Theory of Mahāyāna Buddhism (SOR, vol. XXXIII), Rome, 1966, p. 367, as follows: « Illness is to be cognized, its cause removed, Health should be attained, and a remedy used; Like that, Suffering. its Cause, Extinction and the Path, Are to be cognized, removed, touched and observert».

Page #34

322
A. Wezler
it also stands clearly in contradiction to unequivocal passages in Buddhist texts themselves!
There is, however, still another striking feature in the Vyādhisātra passage one should not silently pass over as done by de la Vallée Poussin. What I am referring to is the fact which cannot but be styled as strange that the physician conversant with the four argas is characterized as « worthy of [treating] a king » (rājārha), « befitting a king » (rājayogya), and is, finally, said « to class with the king's property » (rājāngatve ca samkhyām gacchati) 94. It would seem hardly necessary to state that these predicates can by no means have to do with the frequent designation of the Buddha as vaidyarāja 95.
Yet one might feel induced to refer to the well-known fact that Buddhist monks off and on succumbed to the temptation of acting as medical practitioners %, and that for many reasons they might have even tried to qualify as royal physicians. But though it has to be admitted that historical social reality is reflected in (Buddhist texts in this regard, too, in the passage under discussion there is no indication whatsoever that we have to do with such a case.
I think I should not any longer avoid coming to the point which, of course, is the following: the characterization of the physician as rājārha, etc., in the Vyādhisūtra cannot consistently and convincingly be accounted for but by assuming that it was retained when the exposition of the quadruple division of the science of medicine was taken over from another source. Another particular feature that was simply retained was the order of enumeration of the four skills of a physician which, significantly enough, does not perfectly correspond to the traditional order of succession of the Four Noble Truths. And this source
94. In view of the parallels in the Abhidharmakośa, viz. pp. 136.15.3, 114.23, 115.9, and in Pali texts (e.g. AN 1.244.8-10 where the stock phrase rājāraha rājabhogga (sic!] rañño angan t'eva sarkham gacchati, said of a thoroughbred horse, is met with, or AN 1.284.11-12, etc.) one would expect here rather räjāngam iti ca... Yet, the text as it is transmitted can, I think, be accepted if -tva is taken to denote a collective idea and is construed in accordance with J. S. SPEIJER, Sanskrit Syntax, repr. Kyoto, 1968, § 238. It should, however, be noted that "tve ca in Sanskrit might go back to Pali tveva which latter, quite often met with, is according to an information kindly given me by my friend o. von Hinüber a wrongly sanskritized form for teva (< ity eva). In rendering... samkhyām gacchati I follow the proposal of the PD pp. 570 and 664; as for the expression rājānga, cf. CPD s.v. anga.
95. bhisakka is styled an adhivacana of the Tathāgata at AN IV.340.5-6; as for vaidyarāja itself cf. the book of R. Birnbaum (cf. fn. 76) as well as the Hôbôgirin s.v. byô, p. 230. ff.
96. Cf. e.g. A. L. BASHAM, The Wonder that was India, London, 1954, p. 499, and R. BIRNBAUM, op. cit., p. 6 ff., p. 20 f. (with further references); as noted by the latter, the Suvarnaprabhāsottamasūtra even describes «how Säkyamuni [himself] in a past life studied all the principles of medicine in order to aid his contemporaries, a « legend >> which not only has grown out of particular Mahāyānistic ideas, but is also most probably due to the wish to make the Buddha himself the prototype of what a Buddhist monk had become in the course of time.

Page #35

On the Quadruple Division of the Yogaśāstra
323
cannot have been a Buddhist text; it was in all probability a medical text, or, at least, the science of medicine. This conclusion, though cogent in itself, is further and most clearly corroborated by the fact that the only Sanskrit medical text in which the four systematic parts are at all mentioned, viz. the Carakasamhitā, surprisingly enough comes to speak of them only in connection with the description of the qualifications of a rājārho bhisaktamaḥ 97 and shows the same order of enumerating them. To regard this striking agreement as quite an accident would, no doubt, be nonsensical. Of course, one cannot be absolutely sure that it was this verse of the Carakasamhitā which was the source drawn upon by the author of the Vyādhisūtra, for the alternative possibility, viz. that both go back to a common, though unknown source, cannot be ruled out with certainty. But this much seems clear: the fact that the science of medicine falls into four systematic parts, viz. those mentioned in quite a few different texts, was discovered first by medical men.
To return now to the central question 98 as formulated at the outset of this paragraph. It has to be stated that Buddhist sources do not by any means support the assumption that «les "Quatre Vérités” du Bouddhisme sont empruntées à la médecine »; on the contrary, they give clear evidence that the science of medicine with its four systematic parts was drawn upon only for the sake of illustration; it is but the following hypothesis these sources warrant to frame: the obviously old comparison of the Buddha to a physician, of his teaching to a medicine and similar ideas were in the course of time elaborated, but it was not until the quadruple division of the science of medicine originating in medical circles became known to Buddhist authors that the Four Noble Truths as such were by way of comparison paralleled to the corresponding parts of the Cikitsāśāstra, though a full parallelism, i.e. regarding the order of succession also, was achieved only by a second step. However, it seems that the full importance of this systematic division was not realized by medical men, but significantly enough by quite different people, viz. those Yoga and Buddhist authors who applied it for the first time to their respective doctrines of salvation.
Therefore, one cannot but draw the final conclusion that there is not the slightest evidence for the assumption that this fourfold division of the science of medicine it was that inspired the Buddha to his Four
97. As there were already similar idioms in Buddhist tradition (cf. p. 316 and fn. 94), it was quite easy for Buddhist authors to seize on this suggestion of medical texts.
98. In contradistinction to the authors of the Hôbôgirin who (loc. cit.) voice the opinion that « la question de savoir si le bouddhisme a "emprunté" ses quatre Vérités à la médecine n'a guère d'intérêt », I find this question extremely interesting, and not only because many Buddhologists do believe in this « borrowing >>; but I differ also from Oldenberg in that I am less pessimistic as to the possibility of finding an answer to it, though, of course, only a hypothetical one.

Page #36

324
A. Wezler
Noble Truths. This was virtually already shown by J. Filliozat as early as in 1934, but his article 99 does not seem to have been given the attention it, undoubtedly, deserved. The conclusion arrived at in the present more comprehensive study does not, of course, « prove » that such an influence of the contemporary medicine on the Buddha is by no means impossible; all I want to say, yet this emphatically, is that this assumption is without any basis whatsoever, i.e. that as to its philological foundation it stands in a vacuum as it were. For, the similarity between the Four Noble Truths and the four parts of medicine, highly suggestive though it is, cannot by itself be regarded as warranting any such assumption: this similarity can consistently be accounted for by assuming a material coincidence arising quite naturally out of the essential element both, physical illness and existence viewed as suffering, do have in common; «for > — to recapitulate what has been stated already above (p. 300 f.), yet this time in the words of Vasubandhu 100 — « that to which one is attached and by which one is given pain and from which one seeks to free oneself, it is that is examined first in the phase of consideration, i.e. the Truth of Suffering; thereafter [one puts oneself the question] "What is its cause (i.e. the cause of Suffering)?” (and thus examines] the Truth of that which causes [Suffering] to arise 101; [then one puts oneself the question] ” What does its suppression (i.e. the suppression of suffering) consist in?” [and thus examines] the Truth of Suppression; [and finally one puts oneself the question] "Which is the way [leading to] it (i.e. to its suppression)?” (and thus examines ] the Truth of the Way » 102
5. The idea of a quadruple division of the doctrine of salvation is, however, not confined to the traditions of Pātañjala-Yoga and of Buddhism alone. It is equally attested in yet another school of thought, viz. that of Nyāya. The historical problems the philologist is faced with are hence even more complicated than was assumed until now in the course of the present study.
-
99. JA (1934), pp. 301-7, referred to in HOBOGIRIN, loc. cit.
100. Viz. Abhidharmakośa, ed. by P. Pradhan, p. 328.6-8: yatra hi sakto yena ca badhyate yataś ca mokşam prarthayate tad evādau vyavacāranāvasthāyām duḥkhasatyam parikşyate / paścāt ko 'sya hetur iti samudayasatyam ko 'sya nirodha iti nirodhasatyam ko 'sya märga iti märgasatyam /.
101. Incidentally, this passage in the Kośa corroborates my interpretation of the term samudaya as given above p. 303.
102. Cf. also the verse of the Saundarānanda quoted above (p. 321) as well as fn. 93a. For the « gradual apprehension » (anu pūrvābhisamaya) of the Four Noble Truths cf. also a fragmentary Sanskrit MS from the Turfan discoveries described and edited by E. WALDSCHMIDT, Sanskrithandschriften aus den Turfanfunden, Teil I (Verzeichnis der orientalischen Handschriften in Deutschland, hrsg. von W. Voigt, Bd. X.1), Wiesbaden, 1965. p. 294; this passage is of particular interest in another respect also in that it shows close similarity to the passage from the Samyuktāgama (cf. fn. 92) by equally pointing out different causes of disease and different methods of treatment.

Page #37

On the Quadruple Division of the Yogaśāstra
325
5.1. Commenting on NS 1.1.1 Paksilasvāmin Vātsyāyana states almost right at the beginning: ātmādeh khalu prameyasya tattvajñānān nihśreyasādhigamah / tac caitad uttarasütrenānūdyata iti / heyam, tasya nirvartakam, hānam ātyantikam, tasyopāyo 'dhigantavya ity etāni catvāry arthapadāni samyag buddhvā niņśreyasam adhigacchati /, « Out of proper knowledge of what such objects of valid cognition like soul, etc., in reality are, the Highest Good is attained; and this is repeated in the following sūtra (i.e. NS 1.1.2), thus [has NS 1.1.1 to be interpreted] 103. One has to understand that which is to be avoided, that which brings it forth, [its] absolute avoidance [and] the means [leading to] it; having properly understood these four right statements, one attains the Highest Good ».
In justification of my interpretation which is at variance with the explanation given by Uddyotakara, some brief remarks should be made. According to Uddyotakara the four arthapadas are heya, hāna, upāya and adhigantavya equated by him to moksa. That is to say, he started from the assumption that adhi-gam must necessarily be used here in the same sense it has in the syntagma niņśreyasam adhigacchati, i.e. the predicate of the sentence as a whole. This conclusion is, however, by no means cogent, and the parallel in the introductory portion of the Bhāsya on the 2. Ahnika of the 4. Adhyāya to which one could refer 104, viz, apavargo 'dhigantavyaḥ, does not help much because it is continued by tasyādhigamopayas tattvajñānam, i.e. by an unequivocal statement according to which the means for attaining liberation consists in tattvajñāna 105. On the contrary, in explaining hānam by tattvajñānam, and upāyaḥ by śāstram, Uddyotakara stands clearly in contradiction to this statement of the Bhāsyakāra's. Besides, the construction of the Bhāşya passage obviously assumed by Uddyotakara would be rather odd, there
103. One of the problems the commentators of NS 1.1.1 are faced with lies in that pramāna is mentioned along with prameya among the entities true knowledge of which leads to the Highest Good. Following Paksilasvāmin all of them, therefore, opt for an interpretation according to which it is the tattvajñāna of the different prameyas only which has this function.
104. In G. Jha's edition (POS 58), Poona, 1939, this parallel is found on p. 289.3 f.
105. The fourfold division spoken of in this passage of the NBhāşya (evam catassbhir vidhābhih prameyam vibhaktam āsevamānasya... tattvajñānam utpadyate) does at first sight appear to be not identical with the four arthapadāni mentioned in the commentary on NS 1.1.1; for the four elements seem to be « things, viz. rebirth, result (of acts) and pain, to be known » (jñeya), «things, viz. karman and defects, to be avoided » (praheya), « liberation as that which has to be attained » (apavargo'dhigantavyaḥ) and, finally, « true knowledge as the means for attaining liberation » (tasyādhigamopāyas tattvajñānam). In the light of relevant passages in the preceding part of the Bhāşya, viz. yas tu duḥkham duḥkhāyatanam duḥkhānuşaktam sukham ca sarvam idam dunkham iti paśyati sa duhkham parijānāti / parijñātam ca duhkham prahin am bhavaty anupādänāt savisānnavat / evam doşān karma ca dun k hahetur iti paśyati I, however, it becomes clear that in fact the same quadruple division is ultimately intended. But I find it difficult to decide whether catassbhir vidhābhih has to be construed with vibhaktam or else with āsevamānasya.

Page #38

326
A. Wezler
being no reason why Paksilasvāmin should in a series of nouns have decided in favour of a verbal and imprecise expression 105a as regards the final member. There is hence good reason for taking adhigantavyaḥ as predicate to be construed with all four subjects. In addition it should be noted that it is on the basis of this assumption that a proper understanding of the term arthapada can be achieved: evidently it is to be regarded as a tatpuruşa-compound the prior member of which corresponds to a dative (arthāya padam), and thus fully agrees with Pāli atthapada for which the CPD gives the meaning « a right or profitable word (often referring to the holy texts) », i.e. a word that is to the advantage of another person. The equation of hāna with tattvajñāna is furthermore highly unconvincing because its characterization as ātyantika in the Bhāsya itself, which clearly echoes NS. 1.1.22 (tadatyantavimokso 'pavargah), cannot but be taken to indicate that what is meant by the expression hāna here is not « means of avoidance », but « avoidance » itself as that which is to be attained. Taking into account this necessary correction of Uddyotakara's interpretation, one can unhesitatingly subscribe to his concluding remark 106: etāni catvāry arthapadāni sarvāsv adhyātmavidyāsu sarvācāryair varnyanta iti /; it is noteworthy, nay extremely interesting that Uddyotakara did not regard the teaching of these four systematic parts to be a specific feature of Nyāya only, but expressly states them to be a common possession of all adhyātmavidyās without exception 107. Evidently it did not bother him where this doctrine developed first: historical problems were not of concern to him — in contradistinction to the Western philologist. He will not rest satisfied with observing that the distinction drawn by Paksilasvāmin between heyam, tasya nirvartakam, hānam ātyantikam and tasyopāyaḥ is in striking agreement with that drawn by Yoga texts between heyam, heyahetuh, hānam and hānopāyaḥ, but face up to the historical problem(s) involved.
5.2. Yet, before doing so it is advisable to inspect some more relevant material from Nyāya literature, though a comprehensive study of it cannot be undertaken here, as it would inevitably mean writing a history of the doctrine of liberation in this school of thought. There is, however, one Naiyāyika whose testimony seems to be of such outstanding importance that it should not be omitted, viz. that of Bhāsarvajña
105a. The gender, too, would be most disturbing, for Paksilasvämin would indeed demand a great deal from his readers were they compelled to realize on their own that it is masculine because apavargah has to be supplemented!
106. Nyāyadarśana of Gautama with the Bhāşya of Vātsyāyana, the Vārttika of Uddyotakara, the Tätparyaţikā of Vacaspati and the Parisuddhi of Udayana, vol. I, chapter 1. ed. by A. Thakur (Mithila Institute Series, Ancient Text no. 20), Darbhanga, 1967, p. 14, 1. 4.
107. Already Paksilasvāmin is, however, at pains to show that the Nyāya system is by no means an adhyatmavidyā only like the Upanişads, but represents, on the contrary, an independent fourth vidya; cf. NBhāşya 3.3-8.

Page #39

On the Quadruple Division of the Yogaśāstra
327
(who is anyway an extremely remarkable author). In justifying the question kimlaksanam of the mūla, referring to prameyam 108, he explains that what is meant is a particular « object of valid cognition » (prameyaviseşa) defined by him, thus 109: y a dvişa yam tatt vajñānamanya jñānānu payogitvena i vanih sreyasangam bhavati, mithyājñānam ca samsāram pratanoti, tatpram eyam 1: « Object of valid cognition is that the tattvajñāna of which forms a means for [attaining] liberation without being of any use for [the acquisition of] other knowledge 110, and the mithyājñāna (i.e. false knowledge) [of which] extends Samsāra ». Bhāsarvajña adds the further explanations 111:
tad eva tatt vato jñātavy am sarvadā bhāvitavy am (read: bhāvayitavyam) ca / na tu kita sa mkhyā dy api prameyam teşām jñānasya nih sreya sānu payogitvāt / 112 taccaturvidham: heyam, tasya nirvartakam, hānam ātyanti kam, tasyo pāya iti /. « This is that has to be known according to its true nature and what has always to be kept present. The number of beetles, and similar things, however, does not equally represent an object of valid cognition113, because knowing them is of no use for [attaining] liberation. This [object of valid cognition] is of four kinds, viz. (iti) that which is to be avoided, that which brings it (i.e. the former) about, absolute avoidance [and] the means for [attaining] it (i.e. absolute avoidance) ».
108. N(yāya) Bhūş(ana), ed. by. Svāmī Yogindrānanda, Benares, 1968, p. 436.1 f.
109. NBhūş 436.11-12; passages which literally agree with formulations of the Nyāyasära are printed in spaced type.
110. Similarly Nyāyamañjari (ed. by K. S. Varadacharya, ORI Series no. 46, Mysore, 1969). 22.5-6: ātmāpavargaparyanta- (read: ātmādyapavargaparyanta-) dvādaśavidhaprameyajñānam tāvat anyajñānānaupayikam eva sākşād apavargasādhanam iti vaksyāmah (viz. at the beginning of the prameyabhāga); for anyajñānānau payikam Cakradhara (Nyāyamañjarigranthibhanga, ed. by Nagin J. Shah, L.D. Series 35, Ahmedabad, 1972) gives the following explanation (5.12-14): anyajñānam anaupayikam advāram anupāyo yasya / upāya evau payikam iti svārthe vinayādipäthät thak hrasvas ca (cf. Pān. 5.4.34 and Ganapatha 211.3) / anyajñānasya vānupāyah sad apavargasadhanam na punaḥ pramānādijñānam iva prameyajñānopāyatayety arthan // It is, of course, the second of these explanations which has to be accepted; the expression used in the NBhüş, viz. anyajñānānupayogitvenaiva, is unequivocal in this respect; what is meant in both cases is to point out that the prameyajñana differs from the knowledge acquired with the help of means of valid cognition, etc., in that the former is of direct importance for liberation.
111. NBhūş 436.14-16.
112. The formulations in the Nyāyasāra are slightly different, viz. na ca kitasankhyādi tajjñānasyānu payogitvät.
113. The editor of the NBhūş refers here to Pramāņavārttika 2.33: tasmād anuştheyagatam jñānam atra vicāryatām/ kitasamkhyāparijñānam tasya nah kvopayujyate //.

Page #40

328
A. Wezler
My interpretation of the NBhāsya passage quoted above is thus fully corroborated by Bhāsarvajña 114. In addition it has to be noted that Bhāsarvajña is more explicit than Paksilasvāmin in that he unequivocally states that what falls into these four parts is in fact the « object of valid cognition », whereas in the Nyāyabhāşya it is left open whether the four arthapadāni are at all related to the category « prameya >> and, if so, precisely in which manner 115. According to Bhāsarvajña 116, too, the category « object of valid cognition is in itself ambivalent: if it becomes the object (vişaya) of proper knowledge, i.e. if its true nature is cognized, its knowledge leads to liberation; however, if its true nature is not understood, this false knowledge has the opposite effect, i.e. causes the process of rebirth to continue. It must, however, be borne in mind that this category includes only certain entities that are to be truly cognized, viz. those which are of direct importance for the attainment of the final goal.
Bhāsarvajña's assertion that it is divided into four parts only, viz. heyam etc., is quite understandably objected to by an opponent who rightly draws attention to the fact that according to NS 1.1.9 where the different pramevas are enumerated and according to NS 1.1.10 ff. in which the individual « objects of valid cognition » are defined, there is a total of twelve of them, but not four. This objection 117 is met by Bhāsarvajña by the following arguments 118. satyam, dvādaśavidham uktam, kim tu tad dvādaśavidham api heyādirūpena caturdhā bhāvyamānam nihśreyasanimittam bhavatity ataś caturvidham uktam / sūtrakāras tu nyāyaprāptam cāturvidhyam iti manyamānaḥ sākṣān noktavān / na hi heyādirūpāniscaye śarīrādisy ahamkāranivettivairāgyot pādanadvärena tadatyantavimokso 'pavargaḥ sambhavati caturvargāntarbhūtatvāc ca dvādaśavidhasyāpi /. « It is true, [the "object of valid cognition"] is taught [in the Sütra] as twelvefold; but it functions as the cause [of attaining] liberation [only] when it, though twelvefold, is presented to the mind as fourfold, [i.e.] as having the forms of that which has to be
114. Because he follows the tradition of those Naiyāyikas who wrote commentaries on the NBhäşya, i.e. the so-called Vyākhyātāraḥ (cf. A. Wezler, in WZKSO, XIX, 1975, 135-146)?
115. The term arthapada itself does not help in answering this question, and the other passage in the NBhāşya quoted above (p. 325) can be interpreted in two different ways (cf. fn. 105).
116. Who is, however, in agreement with Paksilasvāmin [cf. Bhāşya on NS 1.1.9, 2378: asty anyad api dravyagunakarmasāmānyaviseşasamaväyäh prameyam (viz. that taught by the Vaiseşikas) tadbhedena căparisankhyeyam: asya tu (i.e. that taught in NS 1.1.9) tattvajñānād apavargo mithyājñānāt samsāra ity ata etad upadistam viseseneti //] and Uddyotakara (cf. his Värttika on the beginning of the 4th Adhyāya; ed. by V. P. Dvivedin and L. S. Dravid, KSS 33, Benares, 1916-17, 500.8-10: yas tattvajñānavişayatvena vyavatişthamānah samsāravyavacchedahetur bhavati yaś ca mohavisayatvāt samsāram pratanoti sa tattvato jñeya iti /).
117. Extending from NBhūş 437.1 to 441.23. 118. NBhūş 441.25 and 442.3.

Page #41

On the Quadruple Division of the Yogaśāstra
329
avoided, etc.; it is for this reason that it is stated [by me] to be of four kinds. The author of the Sūtra, however, did not state this expressly as he was of the opinion that the fact that it falls into four parts follows logically [so that it need not be made explicit]; for, liberation as an absolute freeing oneself from Suffering by means of stopping egotism and by creating [in oneself] the state of indifference with respect to ["objects of valid cognition "] like the body and so forth is not possible if [these objects] are not ascertained as having the form of that which has to be avoided, etc.; and (the fourfoldness of the " object of valid cognition" was not expressly stated by the Sūtrakāra also] because [the " object of valid cognition "] even if twelvefold is subsumed in the four groups (of that which has to be avoided, etc.] ».
That is to say, according to Bhāsarvajña the two divisions are not mutually exclusive; on the contrary, the twelvefold division is contained in the fourfold; the latter one, however, is the decisive one in so far as it, and it only, leads to liberation: evidently the tattvajñāna of the twelve different prameyas is regarded as not sufficient for attaining the « Highest Good »; each and every « object of valid cognition >> is to be looked into as regards whether it represents something that has to be avoided or something that causes that which has to be avoided, etc.; and only then is it that knowing their true nature functions as a means to liberation.
This is clearly a doctrine which goes beyond the wording and intention of the NS: the impact of soteriological ideas which had led to refashioning the old Nyāya tradition so as to make it, too, look like a mokşaśāstra, continued to be felt by Naiyāyikas who met the challenge by elaborating ideas found in the Bhāşya and, of course, by trying their very best to vindicate their conviction that what are but later ideas were at least thought of and realized by the Sūtrakāra himself.
Nevertheless, Bhāsarvajña's view, though historically untenable, is, of course, worthy of notice, and not only as regards the manner in which he handles the received mūla text, viz. the NS 119. For, the explanation he.gives for the Sūtrakāra's not having expressly stated the cāturvidhya of the < object of valid cognition is not accepted by his opponent who raises the objection that Gautama should have taught it 120 precisely in view of its central importance. Bhāsarvajña does not let this objection pass, his answer being 121: na, paramatanişedhārthatvāt / tatra sāmkhyānām matam: heyam duḥkham rajovsttyātmakam, « drastrdrsyayon samyogo heyahetuḥ » (YS 2.17), « tasya hetur avidyā » (YS 2.24), «tadabhāvāt samyogābhāvo hānam tad drśeh kaivalyam » (YS 2.25), « vivekakhyātir aviplavā hānopāyaḥ » (YS 2.26) iti / sākyānām ca: duḥkhasamu
119. It is in this regard that the NBhūs in general is of particular interest and hence calls for a careful and comprehensive study.
120. NBhūş 442.5: yady evam cāturvidhyam eva sūtrakrtā kasmān noktam? 121. NBhūş 442.5-11 and 19-20.

Page #42

330
A. Wezler
dayanirodhamārgākhyacaturāryasatyajñānam muktyartham istam /... 122 ... tad evam paramate 'pi heyādicatustayajñānam muktinimittam istam / tanmatanişedhārtham ātmādibhedopanyāsaḥ krtaḥ 1.
To put it in a nutshell: according to Bhāsarvajña — who, on the one hand, almost grotesquely comes short of historical truth, but, on the other, i.e. if his presuppositions are taken for granted, argues in a quite understandable and remarkable manner - Gautama did not expressly state the prameya to be fourfold, but instead confined' himself to teach and define the twelve different « objects of valid cognition », because he wanted to refute the opinion of others, viz. the Samkhya-Yogins and the Buddhists, who equally considered the knowledge of the four elements of heya etc. to be a means for liberation. His final remark may be taken as a hint at the importance Bhāsarvajña attributes in this connection to the concept of ātman and the Nyāya view of it. This is corroborated by what he says in introducing the next paragraph, viz. 123 tatrātmajñāne sati paralokākāmkşā bhavati nānyatheti vāksyāmah ), «the desire for the other world (i.e. the wish to attain liberation) arises [only] when among these [different kinds of knowledge] there is knowledge of the ātman, not otherwise this I shall teach '[later] >> 124
But before examining the further development of his argument it is necessary to dwell on the passage just quoted. Bhāsarvajña was evidently not the first to realize that the idea of the doctrine of salvation being divided into the four systematic parts does not constitute a peculiarity of Nyāya; but Uddyotakara's relevant remark 125 looks rather colourless in that he still confines himself to claiming for the Nyāyaśāstra the status of being, among other things, an adhyātmavidyā, too, and does not deem it necessary to determine more precisely the relation in which his school stands in this respect to other doctrines of salvation. Viewed against this background Bhāsarvajña's corresponding remarks are not only much more informative, but are also evidence of a considerable progress in analysis. For he not only points out the basic doctrinal correspondence between the school of Nyāya and particular other systems which he does not hesitate to name, but he also enters into a discussion of the specific doctrinal differences. It is, of course, by no means surprising that what he right from the beginning aims at is to prove the claim that, inspite of the quadruple division being common to other doctrines also, it is the Nyāya doctrine of salvation alone that stands a critical test and should hence be universally accepted. This claim is not, however, connected with the fourfold division as such, but with the definitions of the four systematic parts as given by the Sāmkhya-Yogins
122. Read thus against the edition! In the passage left out here, viz. NBhüş 442.11-19, Bhäsarvajña gives a brief explanation of the Four Noble Truths.
123. NBhūş 442.22. 124. Viz. NBhūş 461.15 ff. 125. Cf. above p. 326.

Page #43

On the Quadruple Division of the Yogaśāstra
331
and Buddhists, respectively: it is but the conceptional and doctrinal content that his criticism is directed against. Yet his remarks, systematically comprehensive as they are, arrest the philologist's attention in other respects, too: They furnish a welcome corroboration of the conclusion arrived at earlier in the present study as regards the problems of the quadruple division being attested already in the YS itself and the correspondence between it and the Four Noble Truths of the Buddha. Noteworthy, however, is that Bhāsarvajña does not refer in this connection to the caturvyūhatva of the Cikitsāśāstra; he must have known the relevant passage in the YBhāsya; apparently he was not interested in pointing out this similarity; a further reason may have been the conspicuous absence of this comparison in the older Nyāya tradition.
In continuing his exposition Bhāsarvajña draws attention to the fact that there is still another reason for the prameya having been taught in the NS as being twelvefold. The argument is as follows 126: when the body (sarira) is recognized as being different from the ātman, it is not any longer mistaken for the ātman and henceforth one does not commit acts of violence, etc., for the sake of one's body. When it is realized that the body is the seat of pain and suffering (duḥkha), attachment to it ends. The next step consists in coming to know the nature of the senses (indriya), i.e. that defects (dosa) do not arise when the senses do not function even though there may be causes that could give rise to defects; thereupon one strives to withdraw the senses (pratyāhāra), and in doing so one does not any longer look upon their objects (artha) (which have been recognized as causing suffering) as something one should grasp, and then one frees oneself of all desires (vairāgya). Having realized that false (mithyābuddhi) and true knowledge (tattvabuddhi) are the causes of liberation and samsāra, respectively, one eliminates the former and develops the latter by making consistent efforts for it. Then greatest efforts are made to conquer the mind (manas) which is the root of all functions of the senses 127. When the activities (pravrtti) are recognized to be the root of suffering in so far as they bring about merit (dharma), demerit (adharma), etc., they are avoided. As soon as the true nature of the defects (doșa) is understood, viz. in the light of NS 4.1.60 128, one undertakes the greatest exertions to remove them. Then it has to be realized that the beginningless process of rebirth (pretyabhāva) 128a and, hence, the extreme suffering in the form of birth and death cannot be brought to an end but by liberation (apavarga); and similarly it is to be realized that the result (phala) [of actiyities and defects] 129 consisting
126. Cf. NBhūş 442.22-26 and 443.6-15. 127. Cf. NBhūş 440.10-13 to which Bhäsarvajña refers back.
128. According to the numbering in: Die Nyāyasūtra's. Text, Ubersetzung, Erläuterung und Glossar von W. Ruben (AKM XVIII.2), Leipzig, 1928.
128a. Cf. Paksilasvāmin's definition (NBhāşya 31.7): pretyabhāvo mrtvā punarjanma.
129. Cf. NS 1.1.20.

Page #44

332
A. Wezler
in pain and also in alleged pleasure is ultimately the same in that both, pleasure and pain, are brought about by one's own karman 130. Next, pain and suffering (duhkha) have to be properly understood so as to be able to keep them present to one's mind. When one has come to realize that each and everything, from God Brahmā down to a grass-blade, is permeated by pain and suffering, one does not take interest in it any longer. And when, finally, liberation (apavarga) is recognized to be the unsurpassed, endless and altogether pure means for the ceasing of all pain and suffering, one strives for that goal only.
Bhāsarvajña concludes this passage by stating 131 tad evam dvādaśaprahāresv api tattvajñānabhāvanā copayujyata ity ato 'pi dvādaśadhā prameyam uktam / samksepabhāvanārtham ca tasya cāturvidhyam ucyate /: « Hence, true knowledge is also realized in all these twelve ways; it is for this reason, too, that the " object of valid cognition" is taught [in the NS] as twelvefold; and that it is (on the other hand] of four kinds [only], this is taught for the sake of keeping (true knowledge] present to one's mind in a concise form ».
That is to say, Bhāsarvajña first demonstrates the true knowledge one has to develop with respect to the twelve prameyas, following their order of enumeration as given in NS 1.1.9, and states thereafter by way of summarizing his exposition, that they have been taught by the Sütrakāra not only in order to refute the theory of the quadruple division of the doctrines of salvation as expounded by other schools, but also to intimate that all twelve of them should be reflected upon in the manner outlined by Bhāsarvajña himself. His very last remark is particularly interesting, and that in two respects: it determines more precisely the relation between the two alternative divisions of the prameya, the twelvefold and the quadruple; and the characterization of the latter as being introduced « samkşepabhāvanārtham > makes one wonder whether this is another instance of the well-known tendency, so important in the history of Indian religions down to the present day, to replace complicated and time-consuming religious acts by « abridgements ». But it is by no means clear that what Bhāsarvajña has in mind is such a substitution; he might have equally thought of a real alternative only, or even regarded the choice of the samkşepa as a means to intensifying the bhāvanā of true knowledge which ultimately leads to liberation.
Yet, such theoretical deliberations are of little use here, at least as long as the apparent contradiction is not discussed in which these remarks of Bhāsarvajña's seem to stand to what had been said by him earlier on the relation between the two divisions. Had he not claimed that it is the quadruple division only that leads to liberation and does he not clearly contradict himself in asserting now that « true knowledge is also realized in all these twelve ways »? I think that the two relevant passages appear contradictory only at first sight. For what Bhāsarvajña
130. CE. NBhūs 441.2-7 and 9-15 to which Bhāsarvajña refers back. 131. NBhūş 443.14-15. Read 'bhāvanopayujyate or 'bhāvanaivopa'!

Page #45

On the Quadruple Division of the Yogaśāstra
333
actually had said, and, to wit, also with respect to the particular procedure of the Sūtrakāra, was that « liberation as an absolute freeing oneself from Suffering by means of stopping egotism and by creating (in oneself] the state of indifference with respect to ["objects of valid cognition"] like the body and so forth is not possible if [these objects] are not ascertained as having the form of that which has to be avoided, etc. »; and what he shows in the second passage is precisely the cognitive and spiritual development of a Naiyāyika who reflects upon the twelve « objects of valid cognition on the basis of the more important division of heyam, etc.! Therefore, not only his earlier statement that « [the object of valid cognition] even if twelvefold is subsumed in the four groups [of that which is to be avoided, etc.] », but also what he says here, viz. that « the fourfold division is taught samksepabhāvanārtham », are fully intelligible in that, understood thus, the fourfold division in fact comprises in itself the twelve prameyas and for this very reason forms the quintessence of their bhāvanā. In the light of these deliberations it is hence more probable that what Bhāsarvajña wants to intimate by his final remark is but the superior importance of the cāturvidhya as emphasized by him already at the very outset.
Space being limited it is not possible to deal here with the same extensiveness as hitherto with what Bhāsarvajña says in the subsequent paragraphs of his Bhūsaņa. The gist of his exposition, however, may be briefly outlined, following mainly the Nyāyasāra: in order to define the concept of heyam he quotes YS 2.16 132 and adds that suffering is of twenty-one kinds, i.e. that it comprises the body, the six senses (manas is included), their six objects, the [corresponding] six [types] of cognition, pleasure and pain in that all — except for the body which is the seat of pain and except for pleasure which is nothing but pain because it does not occur without pain 133 and, of course, except for pain itself which is duḥkha par excellence as it consists in bādhā, pīdā and santāpa — of them lead to suffering. That which brings about the heyam (tasya nirvartakam), i.e. its specific cause, is stated to be nescience (avidya) and « thirst » (trsnā), on the one hand, and merit and demerit (dharmādharmau), on the other; nescience is then defined as false knowledge of that which has been shown by true adhyātmavids 134, and it is made clear that it includes the mental impression left by nescience 135,
132. See above p. 295. 133. NBhūş 444.5-19; cf. also 441.9-15 and 445.24 ff.
134. Of the two Slokas quoted by Bhāsarvajña to give an example of what «true knowers of the ātman >> and «those blind with passion >> say (NBhüş 444.27-20), not identified by the editor, the first one is from Mbh. (Poona) 12.316.42 and 43; the second one seems to belong to the vast stock of floating stanzas one finds so often quoted in anthologies and Indian narrative literature; it is included in 0. BÖHTLINGK's, Indische Sprüche, I, repr. Osnabrück, 1966, p. 98 (no. 3565); for further references see L. STERNBACH, Supplement to 0. Böhtlingk's Indische Sprüche (AKM XXXVII.1), Wiesbaden, 1965, p. 9.
135. Cf. NBhūş 444.23-30 and 445.4-7. Note that by duḥkhasūtre (which is met with also NBhūş 443.22, 445.20, 446.14 and 584.16) Bhäsarvajña refers to NS 1.1.2 dealt with in detail NBhūş 72.15 ff.

Page #46

334
A. Wezler
« thirst >> is explained to be the longing for rebirth 136. The reason for dharma and adharma being mentioned separately is that nescience is the main cause of suffering (pradhānam duḥkhanimittam) 137, and merit and demerit are, properly speaking, specific causes of pleasure and pain, respectively, and that both of them are equally to be regarded as duhkhakārana because it is a matter of experience that one is given pain by cold wind, etc., even when nescience and thirst do not persist any longer 138. Hānam is then defined to be the cessation of suffering (duḥkhaviccheda), and the addition of «absolute » (ātyantika) is vindicated by referring to the fact that a temporal elimination of personal (ādhyātmika) pain and suffering can be achieved also by applying the means taught in the Dharma-, Artha-, Kāma- and Vaidya-Sāstras 139; that is to say, the attribute is meant to make clear that the liberated ātman does not ever again and in no manner come in contact with duhkha.
Finally, the fourth element, i.e. the means for attaining complete avoidance (tasyopāyah), is determined to consist in true knowledge of the ātman. That out of the twelve « objects of valid cognition » ātman alone is mentioned is justified by the wish to stress its outstanding importance (prādhānya) which, according to Bhāsarvajña, has also led to its being enumerated first in NS 1.1.9 140. The ātman is said to be of two kinds, viz. para and apara, of which two the former is none else but the Maheśvara 141 and the latter is « he who experiences the result of Samsāra » (samsāraphalopabhoktr), i.e. the individual permanent and omnipresent soul 142.
Regarding the rôle these two « objects of valid cognition » play in attaining the final goal, it is stated that the true knowledge of the aparātman represents a means to it (nihśreyasānga) because it leads to activity aimed at « the other world > in that it helps to realize that the paraloka really exists and because it causes the elimination of demerit 143; the true knowledge of the parātman, on the other hand, is ascribed a similar function, only that it works by being a constitutive element of the upāsanā of the Maheśvara 144. That both these tattvajñānas are not of exactly the same rank is finally stated by emphatically proclaiming that liberation is attained only because of perceiving God Siva (tasmāc
132. See above p. 295. 137. Cf. NBhūş 445.20-21. 138. Cf. NBhūs 445.22 ff.
139. The text as read in the edition of the NBhūş, viz. 446.7 dharmādharmarthakāmavaidyaśāstrair..., is corrupt and has to be emended to dharmärtha.
140. Cf. NBhūş 446.13-16. 141. Hence the subsequent long isvaravāda section NBhûş 447 ff. 142. Cf. NBhūş 487.19 ff. 143. Cf. NBhūş 548.1-2 and 19 ff. 144. Cf. NBhūş 574.1 and 12 ff.

Page #47

On the Quadruple Division of the Yogaśāstra
335
chivadarśanād eva moksa iti) 145. Out of the many points in Bhāsarvajña's extensive « commentary on NS 1.1.9 » which ought to be taken up for discussion the following should not be passed over silently here. On the one hand, Bhāsarvajña quotes a series of relevant sūtras from the YS (above p. 329) and clearly in a pūrvapaksa section; on the other, he uses YS 2.16 in his own definition of heyam (cf. above p. 333) as also YS 2.15 in his explanation of phala 146. This contradiction is, of course, only seemingly one. The solution cannot but be that Bhāsarvajña accepts such elements of Pātañjala-Yoga as are doctrinally so to say neutral in that they refer to a practical method of self-control, etc. 146a. In doing so he is not only clearly in keeping with the NS itself, and he himself quotes the relevant sūtra viz. 4.2.45 147, but he also gives Yoga as such its due in so far as it essentially teaches a method that can be and was applied to different systems of thought, can and was practised by people of different persuasions. Bhāsarvajña does, of course, not show « faith in the doctrines of that school », as V.S. Abhyankar and I. R. Devadhar would have it 148; on the contrary, he severely and repeatedly criticizes them. All one can perhaps say is that « he shows a strong predilection for Yoga », also in a later part of his work 149, yet this is a side-issue which I cannot follow up here 150.
145. NBhūş 590.6 and 7 ff. as well as 514.24; cf. also NBhūş 590.13-14: ātmatväviśeşeti (read: 'višeşe 'pi) kasmān nānyadarśanan mokşa iti cet, na, anyasya (i.e. paramătmanah Sivād anyasya) sarvānimittatvāt: na hy anyo 'pi ātmā sarvot pattimatām nimittam iti /. On the distinction between the muktātman and the parameśvara cf. also NBhūş 449.4 ff.
146. Viz. NBhūş 441.13 f.
146a. According to the editor's « Nyāyabhūşanataţtippaņoddhștavākyāvalih >> (pp. 631 ff.) it is a total of 16 sūtras of the YS which are quoted by Bhāsarvajña. On the occasion of quoting the concluding portion of NS 4.2.45 Bhāsarvajña makes the following important remark (NBhūş 584.11 ff.): ... iti vacanāt / pātañjalādiśāstresv apy upadiştah upāyaḥ svaśāstravirodhena (read, of course, Pāvirodhena) pratipattavyah / tatra yuktisāmarthyāc ca [/] « paramatam apratişiddham anumatam » iti tantrayuktih /; that is to say, he himself clearly states the principle he proceeds upon in accepting the « opinions of others ». As for this tantrayukti, cf. NBhāşya 16.3 f. (on NS 1.1.4).
147. Viz. NBhūs 445.28 and cf. fn. 146a. The reading 'vidyupāyaih given there is, however, a misprint for 'vidhyupāyaih.
148. Nyāyasāra of Āchārya Bhāsarvajña with the Nyāyasārapadapañchikā of Vā. sudeva of Kashmir, Poona, 1922, p. 84.
149. Viz. NBhūş '584.9 ff. where Bhāsarvajña, with reference to his definition of upāsanāvidhi, not only refers to NS 4.2.45, but also to the means (upāya) taught in the « Patañjalādiśāstrāņi», stating that this upāya has also to be accepted to the extent that it does not stand in contradiction to Nyāya tenets and is hence not expressly objected to (pratişiddha). Note also the remark NBhūş 584.18 19: dvādaśaprameyajñānam api not pannamātram asahāyam rāgādinivsttau samartham / kim tarhi? yogopāyānuşthānasahitam, by which Bhāsarvajña plainly states his particular position: true knowledge of the objects of valid cognition is not by itself as soon as it has been gained capable of bringing all desire to an end, it needs support, viz. practising the means of Yoga. Cf. in this connection also NBhūş 441.14 where YS 2.15 is quoted as the statement of an « adhyātmavid ».
150. As far as I know it is presently being studied by G. Oberhammer.

Page #48

336
A. Wezler
5.3. Returning now to the central problem, one has to put oneself the following question: what are the historical conclusions to be drawn from the observation that the quadruple division of the doctrine of salvation is attested in a substantially identical form not only in the Indian science of medicine, certain Yoga texts and the Four Noble Truths of the Buddha, but also in the Nyāya tradition? There is but a hypothesis I am able to offer, viz. the assumption that suggests itself quite naturally when the relative chronology of the respective earliest sources is taken into account and when a casual convergent development is a priori excluded: the quadruple division was discovered by the Buddha; it was thereafter taken over by the author of the relevant sūtras of the YS (which shows strong Buddhist influences in other respects, too), who also chose other, i.e. neutral, terms for the four different elements. Paksilasvāmin borrowed them probably from the YS 151 as the terminology used by him is practically identical with that of the YS. Similarly, it found, though probably earlier, its way into the Carakasamhitā which directly or indirectly influenced the author of the Yogabhāşya and that of the Buddhist Vyādhisūtra so that they decided to refer to the science of medicine by way of comparison.
This hypothesis, of course, implies that the assumption made above (p. 323 f.), viz. that the science of medicine may claim priority as regards the « discovery » of the quadruple division, has to be abandoned. Yet, I think, it offers a more plausible explanation of the undisputable fact that this division was never of real importance in the tradition of the Cikitsāśāstra. There is another point I should like to make in this connection: is not the widespread belief that the quadruple division as taught in medicine it was that inspired the Buddha perhaps suggested also by the assumption that the Buddha could not have conceived his systematically divided four āryasatyāni had he not had in mind a concrete and wordly example which he need only apply to his doctrine of salvation? And might not the idea, « natural » though it seems, that this « discovery >> cannot but have been made first by people concerned with the somatic aspect of man's existence turn out to be a preconceived one, in the case of Indian culture at least; is it not perhaps even due to the fact that a rather modern and Western conception of medicine is unconsciously taken as basis? Could not just the reverse be true, viz. that in India it was a person deeply. reflecting on man's existence as such that led to this « discovery », i.e. that it was the extraordinary
151. Similarly G. OBERHAMMER, Paksilasvamin's Introduction to his Nyāyabhasyam, in « Asian Studies », vol. 2 (1964). pp. 311-15. Cf. also his Strukturen yogischer Meditation, Wien, 1977, p. 210.

Page #49
On the Quadruple Division of the Yogasastra 337 analytical capacity and sensibility of the Buddha to which Indian medicine in its turn owes the recognition of its caturvyuhatva? tad etad anyair api kathinyarataih suksmadarsibhir yatnena vicaryatam ity alam ativistarena 152. 152. I likewise withstand the temptation to enter on my own into an independent comparative study of medicine and doctrines of salvation the comparison of which could, of course, be carried much farther than it was ever done by Indian authors.