Once Again Vaisesika Sutra 3 1 13

Author(s): Johannes Bronkhorst
Publisher: Johannes Bronkhorst
Catalog link: https://jainqq.org/explore/269610/1

Page #1

ONCE AGAIN VAISEŞIKA SUTRA 3.1.13*
Johannes BRONKHORST, Lausanne
The study of philosophical Sūtra works is beset with difficulties. Apart from the condensed style, which makes them sometimes difficult to understand even where no other problems intervene, we often have reason to suspect that these texts may have undergone interpolations and other modifications. In practice this means that, in order to understand a Sūtra text, we should know as much as possible of its history, of the vicissitudes it has undergone from its beginning until today.
Such detailed knowledge of the history of individual Sūtra texts is not normally available. This is the reason why we have to be content, in most cases, with a global understanding of the kind of influences that Sūtra texts undergo. Here we will concentrate on one such influence, viz., the one exerted by the commentary or commentaries that accompany them. It is known that Sūtra texts are frequently extracted from commentaries that contain them. During this process of extraction mistakes can easily creep into the Sūtra text: a sütra may be overlooked; or, more probably, a statement properly belonging to the commentary may be taken to be a sūtra. Confusions of this kind were facilitated by the fact that commentaries of around the middle of the first millennium C.E. often fail to contain clear indications as to what is sutra, and what commentary. The use of the socalled Vārttika style could not but add to the confusion.
The extraction of a Sūtra text from a commentary could lead to an incorrect result in other ways, too. There is evidence to show that commentators of around the middle of the first millennium occasionally felt free to comment upon the sūtras in an order which deviates slightly from the 'correct' one. In itself this need not be looked upon as an attempt to change the order of the sūtras. But whatever the intentions of these commentators, the effect of such a procedure might very well be that the Sūtra text which someone else subsequently extracted from such a commentary would have some of the sūtras in a modified order.
Usually Indian Sūtra texts are handed down to us in one single 'line of descent', at least where their early period is concerned. It is only on rare
• 1
I thank T. Tillemans for help and advice. See Bronkhorst, 1992, for a brief survey.

Page #2

666
JOHANNES BRONKHORST
occasions that we can show with certainty that commentators did actually comment upon the sūtras in a changed order. There are, however, some clear cases, which I will now present.
Consider first the Samkhya Kärikā. This work does not consist of sūtras, but of kārikās. A considerable number of more or less early commentaries on it have been preserved, and the exact chronological relationship between them is not easy to determine.
The most elaborate and interesting of these commentaries is the Yuktidipikā. This text is not only interesting from the point of view of its contents. Its form, too, is special; it constitutes a perfect example of what I have called the Vārttika style. In the present study we are interested neither in its contents nor in its style, but in the manner in which it deals with the text it comments upon.
The author of the Yuktidipikā is aware of the fact that the Sāmkhya Karikā consists of kārikās. This we must conclude from his use of the term saptati 'seventy', hence 'work consisting of seventy kārikās', to refer to the Sāmkhya Kārikā in his introductory verses. This same term saptati, along with the term āryā which refers to the metre of the work, occurs again in the concluding verses of the Sāmkhya Kārikā as they are found, and paraphrased, in the Yuktidipikā. In spite of this, the Yuktidipikā, unlike all other surviving commentaries, treats the Samkhya Kārikā as if it consisted of sūtras, not of kārikās.3 It frequently divides the kārikās into smaller parts, which it comments upon and refers to as sūtras. Indeed, it never gives the slightest hint that these 'sūtras' together constitute kārikās, so much so that its third Ahnika ends right in the middle of the discussion of what we call kārikā 15; the remainder of karikā 15 is commented upon in Ahnika 4. Sometimes sūtra and kārikā coincide; in such cases a whole kārikā can actually be referred to as sūtra; an example is kārikā 19, which is called sūtra in its discussion in the Yuktidipikā (p. 84 1. 7-8).5 Interestingly, on two occasions the 'sūtras' of the Yuktidipikā do not occur in the order of the kārikā concerned, as these latter are known from all the other surviving commentaries.
2
Solomon (1974) studies eight of them. In this respect the Yuktidipikā has parallels in the Abhidharmakośa Bhasya and Madhyāntavibhāga Šāstra; see Bronkhorst, 1992. See, e.g., YD p. 91. 10, p. 671. 2, p. 98 1. 3. Note that Sadyojyotis' commentary (before 9th century) on the Svayambhuvasūtrasangraha calls the verses of this text 'sūtras', as does the title itself. According to Filliozat (1991: xvii), the term sūtra here "réfère plutôt à la parole d'un être à qui l'on attribue la plus haute autorité".

Page #3

ONCE AGAIN VAIŠEŞIKA SŪTRA 3.1.13
667
Consider kārikā 4. This reads, in all the commentaries except the Yuktidipikā: drstam anumānam aptavacanam ca sarvapramāṇasiddhatvāt/ trividham pramāņam istam prameyasiddhih pramāņād dhill. The Yuktidipika (p. 29 f.) comments, in this order, on the following parts: (i) prameyasiddhih pramāņād dhi, (ii) trividham pramānam iştam, (iii) sarvapramānasiddhatvāt, (iv) dystam anumānam āptavacanam ca. These parts constitute the kārikā, but their order has been reversed. No need to add that in this order nothing remains of the āryā metre. .
In the case of kārikā 4 one might think that the author of the Yuktidipikā took this kārikā as a single unit, and commented upon its parts in a different order. No such position can be maintained in connection with kārikās 6 and 7. These kārikās occur, in almost the same form, in all the surviving commentaries, and must therefore be looked upon as integral parts of the Samkhya Karikā, at least at the time of composition of the Yuktidipikā. The Yuktidipikā has these two kārikās, but it has interposed kārikā 7 between the first and second half of kārikā 6. That is to say: the normal order is 6a-6b-7a-7b, but the Yuktid pikā has 6a-7a-7b-6b. Again, the āryā metre is thoroughly disturbed in this manner.6
There can be no doubt that the author of the Yuktidipikā consciously changed the order of the 'sūtras' of the Samkhya Kärikā; or perhaps: he consciously decided to comment upon them in an order which differs from the original one. The tradition preserved in all the other commentaries guarantees this sufficiently. This certainty makes the procedure of the
6
K. Preisendanz has kindly sent me a portion of her forthcoming book, in which she mentions the possibility that the author of the Yuktid Ipikā did not (yet?) look upon SK 7 as a karikā. She makes this suggestion because of the modified order of the kārikās, and because kärikä 7, in the interpretation of the Yuktid Ipika, is invoked by an opponent. I find her suggestion nonetheless problematic, mainly because already the Samkhya Karika as translated into Chinese by Paramārtha contains this kārikā. (For a discussion of the date of the Yuktidspikā, see Bronkhorst, 1985: 93-94.) Preisendanz's suggestion further seems to necessitate the assumption that the Yuktidipika is not only older than all the other commentaries, but also that it was looked upon by the authors of the other commentaries as in some way authoritative. This again is hard to harmonize with the doctrinal differences which exist between the Yuktidipika and some of the later commentaries. One such difference concerns the question whether the tanmātras have one quality each, or an increasing number from one to five, depending on which tanmatra one is talking about; and the related question whether the tanmatras produce one element each, or whether they produce the elements jointly; see Bronkhorst, 1994, for details. The problems mentioned by Preisendanz can, of course, equally well be solved by the assumption that the author of the Yuktidipika felt free, not only to comment upon the kārikās in a slightly modified order, but also to interpret one kārikā as representing the opinion of an opponent.

Page #4

668
JOHANNES BRONKHORST
Yuktidipika all the more interesting. It shows beyond reasonable doubt that at least some commentators in the first millennium felt free to change the order of the sutras on which they commented.
The Samkhya Kärikā is certainly not the only text the order of whose sūtras has been changed. It may however be the only text where there is so little occasion to look for alternative explanations. It is known, for example, that the Brahma sutras occur at some places in a different order in the commentaries of Sankara and Ramanuja.7 Here, too, it is reasonable to assume that someone changed the original order. However, the Brahma sūtras as they survive today are written in such a manner that it is virtually impossible to decide what this original order may have been.
Let us now turn to Vaiseṣika sutra (VS) 3.1.13.8 It reads:
ätmendriyamano'rthasannikarṣād yan niṣpadyate tad anyat
A number of authors, among them the oldest whose testimony has been preserved, see in this sūtra a definition of perception:
(i) Dignaga remarks in his Pramāṇasamuccaya: "For the Vaiseṣikas there is a definition, mentioned in the Sutra, of perception in respect to substance (dravya), [which is made meaningful] by a certain relation [to the preceding sūtras]. It says: "That [cognition] which is brought about by the contact of the soul (atman), the sense (indriya), the mind (manas), and the object (artha) is [perception as] a separate one [of the pramāṇas]'." There can be no doubt that the sutra quoted by Dignaga is VS 3.1.13. Dignaga's remark to the extent that the sutra "[is made meaningful] by a certain relation [to the preceding sutras]" is noteworthy and must be kept in mind; we'll return to it later.
7 See Bronkhorst, 1981: 317-18 n. 6.
8 This is its number in the version contained in Candrananda's commentary, edited by Jambuvijaya (C). It is 3.1.20 in the version of Bhaṭṭa Vadindra, also contained in the anonymous Vṛtti, both edited by A. Thakur (V), 3.1.18 in the version contained in Sankara Miśra's Upaskāra, reproduced and translated in Sinha, 1911 (U). Where we use only one number, the reference is to C. The present sutra contains the word manas in versions C and V and in a number of quotations of this sūtra in other works; manas is lacking in version U and in "one demonstrably wrong translation of the Pramanasamuccaya Vṛtti" (Isaacson, 1990: 27).
9 Translation Hattori, 1968: 42; the two Tibetan versions on the basis of which the translation was made are reproduced ibid. p. 198-199.

Page #5

ONCE AGAIN VAIŠEŞIKA SŪTRA 3.1.13
669
(ii) Simhasūri quotes the following definition of perception in his Nyāyāgamānusāriņi (ed. Jambuvijaya, vol. I, p. 110):
ātmendriyamano'rthasannikarsād yan nispadyate tad anyat, ätmā manasā mana indriyeņa indriyam artheneti catustayatrayadvayasannikarşad utpadyamanam pratyakşam
This is our Vaiseșika sūtra along with an explanation. The sūtra was apparently quoted - in order to be rejected - in Mallavādin's Dvādaśāra Nayacakra, which Simhasüri comments. 10
(iii) The Yuktidipikā (p. 34, 1. 29-30) cites this sūtra besides other definitions of perception.
(iv) Jayantabhatta's Nyāyamañjari cites the sūtra in the following, slightly amplified, form (p. 280, 1. 15-16):
yad api kaiścit pratyaksalaksaņam uktam 'ärmendriyamano'rthasannikarsad yad utpadyate jñānam tad anyad anumānādibhyah pratyakşam'iti...
Here too there can be no doubt that the sūtra is read as a definition of perception.
(v) The anonymous Vịtti on the Vaiseșika Sūtra edited by A. Thakur (1957), too, explains the sūtra as a definition of perception. So does the commentary by Bhatta vadîndra edited by Thakur (1985), of which the former is an abbreviation.
(vi) The anonymous Sarvasiddhāntapraveśaka, in its chapter on Vaišeşika, contains the following passage (Jambuvijaya, 1961: 145):
āha pratyakşalaksanam kim iti cet, tadāha 'ātmendriyamano'rthasannikarşād yan nispadyate tad anyat '/asya vyākhyā: ātmā manasa yujyate mana indriyena indriyam artheneti/ tataś catustayasannikarşād ghatarūpādijñānam, trayasannikarsāc chabde, dvayasannikarşāt sukhadişul evam pratyakşam nirdistam/
10 Be it noted in passing that Simhasūri appears to quote in matters Vaišeșika usually from
the katandi, a Vaišesika work probably written before Dignāga's Pramāṇasamuccaya (see Bronkhorst, 1993). Do we have to conclude that also the present explanation given by Simhasūri derives from the Katandi? It is not possible at this point to address this question.

Page #6

670
JOHANNES BRONKHORST
(vii) Akalanka's Tattvārtha-Vārttika (p. 53, 1. 32.) cites VS 3.1.13 in an enumeration of definitions of perception.
This interpretation, though supported by early authorities - among them our earliest, is not accepted by some Sanskrit authors, and by several modern scholars (Hattori, 1966; Oetke, 1988: 303-319; Preisendanz, 1989: 150 f.; Nozawa, 1989: 71).11 They reject this interpretation for various reasons, which all boil down to one single factor: the context. The whole of Āhnika 3.1 is believed to be intended to prove the existence of the soul. 12 It is in this way that the three commentators of the sūtras explain this section, and it is clear that a definition of perception in the midst of such an argument would be out of place.
Here the question can legitimately be raised whether we are entitled to discard the massive and ancient evidence provided by Dignāga and the other authors mentioned above on the basis of mere considerations of context. We have no surviving commentary on the Vaiseșika Sūtra that is even approximately as old as Dignāga, nor do we possess certain knowledge of what the context of VS 3.1.13 looked like during his time. (Recall that according to Dignāga the interpretation of 3.1.13 as a definition of perception is supported by its context; in the present situation of the text this can hardly be said to be the case.) Arguments based on context show a marked degree of confidence in the reliability of the Sūtra text as it has been handed down to us, and this without supporting evidence.
The (present) context of VS 3.1.13, when looked at more closely, presents a number of peculiar features, which justify a certain suspicion with regard to the present order of the sūtras. Āhnika 3.1, as stated above, is believed to prove the existence of the soul (ātman). But strangely enough, the proof of the existence of the soul is again taken up in VS 3.2.4, without the slightest hint that this is the second time the topic is addressed. 13 Sūtras 3.2.1-3, which are situated between the two sections purportedly dealing with the proof of the soul, discuss the proof of the existence of the mind (manas). If
11 Honda (1990: 144 (29)) accepts the sutra as a definition of perception. 12 Hattori is explicit about this (p. 897 (100)): "Although VS is not skilful in its arrange
ment of topics, there certainly is an order, which does not allow any arbitrary interpretation to be put on a sūtra." In Oetke's interpretation, 3.1.13 is a reply to an objection
which finds expression in 3.1.7. The intervening sütras are perhaps interpolated. 13 In version V sūtra 3.2.4 concerns only the proof of the existence of the soul in others.
This reading of the sutra (prānāpānanimeson mesaj fvanamanogat indriyantaravikārah parātmani lingam) disagrees however with the evidence of the Padarthadharmasangraha and its commentaries. See below.

Page #7

ONCE AGAIN VAIŠEŞIKA SŪTRA 3.1.13
671
we follow the (relatively recent) indigenous commentaries and several modern scholars, we are asked to believe that Adhyāya 3 contains two sections dealing with the proof of the soul, which are separated, for no obvious reason, by a section which establishes the existence of the mind. 14
The difficulties do no end here. Sūtras 3.1.1-12, which allegedly establish the existence of the soul (ātman), do not once mention the word ātman, nor indeed any other word for soul. This constitutes a marked contrast with the second section concerned with the proof of the soul: VS 3.2.4 enumerates a number of ātmalingas, thus leaving no doubt as to its intentions.
The preceding observations show that the context of VS 3.1.13 is by far not as clear as some may maintain. It is instructive at this point to study how the first discussion on the existence of the soul is initiated.
The discussion starts with sūtra 3.1.1: prasiddhā indriyārthāh No commentator claims that this sūtra by itself introduces the topic of the soul. The sūtra is rather presented as an introduction to sūtra 3.1.2 (3.1.3 in the version of Bhatta Vädindra) which, it is claimed, presents an inference proving the existence of the soul. 15 Sūtra 3.1.2 begins with the word indriyārthaprasiddhi, which obviously refers back to sūtra 3.1.1. What strikes us here, is that there is no need in this context of sūtra 3.1.1. In presenting "the knowledge / establishment of senses and objects / objects of the senses." as logical ground for something else, presumably the soul (or the difference of the soul from other things), it is superfluous to have this preceded by another sūtra which states that "the senses and objects l objects of the senses are known / established".
The only reasonable explanation I can think of for the presence of 3.1.1 at this place is that this is a sūtra which the author of 3.1.2 (or 3.1.2-3) used as excuse and pretext for the introduction of one or more new sūtras. This is of course only possible if 3.1.2 (3.1.2-3 in the case of Bhatta Vādindra) is a later addition to the Sūtra text, newly composed when 3.1.1 was already considered to constitute part and parcel of the traditionally accepted Sūtra text. Seen in this way, 3.1.1 does not, and never did, constitute part of the proof of the soul, but it could be used as point of departure for such a discussion.
14 Oetke (1988: 304) admits: "Dass das gesamte erste Ahnika des dritten Adhyāya dem
Nachweis der Existenz einer Seele gewidmet ist, ist keineswegs so selbstverständlich, wie es von manchen indischen Kommentaren und europäischen Interpreten angenommen wird." The emphasis here is not, however, on Seele, but on Existenz, as is clear from the following remark on the very next page: "Es sieht so aus, dass die zu beweisende These nicht die Existenz einer Seele ist, sonder der Umstand, dass der Atman ein (von den am Anfang des Werkes angeführten Padārthas) verschiedener Gegenstand ist." Or rather, with Oetke, that the soul is different from the padārthas enumerated thus far.

Page #8

672
JOHANNES BRONKHORST
It may be possible to explain in this way the use that was made of 3.1.1 by a later commentator. But what was its function before this commentator used it to introduce a discussion on the existence of the soul? An easy explanation can be provided if we are willing to consider that the commentator concerned did not only use 3.1.1 for his purposes, but moved it away from its original context. Explaining 3.1.1 in its present position is difficult, irrespectively of how one wishes to interpret the remainder of Ahnika 3.1. The fact that it deals with indriya and artha, suggests that it originally belonged between 3.1.13 - which deals with indriya, artha, manas and ātman - and the discussions of manas and ātman in sūtras 3.2.1 ff. If we remove 3.1.14 (which is suspect because it seems to presuppose that the preceding sūtra dealt with the existence of the soul), 16 we arrive at the following sequence of sūtras (the variants in the other two versions are minor and do not affect the argument):
3.1.13: ātmendriyamano'rthasannikarşād yan nispadyate tad anyat 3.1.1: prasiddhā indriyārthāh 3.2.1: ātmendriyārthasannikarse jñānasyābhāvo bhāvas ca manaso lingam 3.2.2: dravyatvanityatve vāyună vyākhyāte 3.2.3:
prayatnāyaugapadyāj jñānāyaugapadyāc caikam manah 3.2.4: prāṇāpānanimeşonmeşajivanamanogatindriyāntaravikārāh sukha-1
duhkhe iccha-dveşau prayatnascety ātmalingāni 3.2.5: dravyatvanityatve vāyună vyākhyāte
This would then mean:
- That which comes about as a result of contact (sannikarşa) between soul
(ātman), sense organ (indriya), mind (manas) and object (artha), is a different [kind of cognition) (viz., perception) (3.1.13) [From among these four factors) sense organs and objects are well-known [and need no further explanation) (3.1.1)
16 Sūtra 3.1.14 appears to deal with the proof of the existence of a soul in others. It cannot
therefore have found its present position until after the rest of Ahnika 3.1 had been given an interpretation that concerns the existence of the soul. This does not necessarily imply that 3.1.14 is a late sutra. Nozawa (1989) has argued that it is old, dating from the time when the soul was still thought of as of limited size. (See however Bronkhorst 1993a: 87 f. on the size of the soul in early Vaišeşika.) All we can say is that its present position cannot be all that old.

Page #9

ONCE AGAIN VAIŠEŞIKA SŪTRA 3.1.13
673
- The inferential mark for the existence] of a mind is that there is (some
times] cognition and (sometimes] not, even though there is contact between soul, sense organ and object (3.2.1) The fact that (mind) is a substance and eternal is explained by the same
arguments as in the case of] wind17 (3.2.2) - There is (only) one mind (in each body] because (several] efforts do not
occur simultaneously, nor do (several] cognitions (3.2.3) - The inferential marks (for the existence] of a soul are: breathing in and breathing out, shutting and opening the eyes, life, movement of the mind, the modifications of the other senses, pleasure and pain, desire and aver
sion, and volition (3.2.4) - The fact that (soul] is a substance and eternal is explained by (the same
arguments as in the case of] wind18 (3.2.5)
This sequence makes sense, is coherent, and even clear enough to allow us to interpret the sūtras without the help of a commentary.
Recall that the reason why someone should comment upon 3.1.1 at its present position, i.e., at the beginning of Adhyāya 3, is quite clear: this allowed him to address the question of the existence of the soul at that place. We will see below why this could be a concern to this commentator.
First, however, we must consider the question what originally preceded sūtra 3.1.13. If we accept, with the ancient witnesses cited above, that this sūtra was a definition of perception, the question can be answered with a fair amount of confidence. The sutra must have been preceded by a related discussion. This is shown by the peculiar form of 3.1.13. Recall that this definition of perception does not mention the word 'perception'; instead it has anyat '[something different / the other one'. It clearly continues a discussion, which distinguished (at least) two kinds of knowledge or cognition (ñāna(?); or perhaps pramāņa? see below). The form of 3.1.13 indicates that it was preceded, in all probability, by one or more sūtras about inferential knowledge. This is what Jayantabhatta suggests in so many words (see above),19 and indeed, even in its present shape the Vaiśesika Sūtra has some sūtras somehow dealing with inference immediately preceding 3.1.13.
17 This refers to sütras 2.1.11 adravyavattvād dravyam and 2.1.13 adravyavattvena
nityatvam uktam, both of which occur in the discussion of wind. 18 See preceding note. 19 Similarly Jinendrabuddhi; see below.

Page #10

674
JOHANNES BRONKHORST
It would be sheer temerity to pretend to be able to reconstruct the original form and history of sūtras 3.1.3-12. One thing seems however certain: sūtra 3.1.13 was once preceded by sūtras dealing with inferential knowledge. The word anyat in 3.1.13, moreover, suggests that at one time these preceding sūtras contained some such neuter noun as jñānam. No such noun is at present to be found in 3.1.3-12. Nor do any of these sutras seem to introduce, or define, inferential knowledge.
Here, however, the following is to be observed. The first part of sūtra 3.1.8 reads: samyogi, samavāyi, ekārthasamavāyi, virodhi ca. The commentators seem to think that these adjectives characterize the word linga 'inferential mark', which is not mentioned in the sutra. But there is another sūtra (9.18) which reads: asyedam karyam kāraṇam sambandhi ekārthasamavāyi virodhi ceti laingikam. The similarities with 3.1.8 are striking, yet 9.18 does not speak of inferential marks, but of inferential knowledge. It can be translated: "Inferential [knowledge is characterized by the relation:] 'this is the effect of that', 'this is the cause of that', 'this is related to that', 'this inheres in the same object as that', 'this is opposed to that'." It is therefore conceivable that 3.1.8, too, introduced inferential knowledge. The person who changed the order of the sutras may have adjusted sūtra 3.1.8 to his purposes by trimming it. Alternatively we may consider the possibility that the scribe who extracted the sūtra from its commentary failed to extract the whole sūtra.
· Whatever the exact original shape of the sūtras, it seems probable that Adhyāya 3,20 prior to the changes pointed out above, discussed inferential and perceptual knowledge before turning to the mind (manas) and the soul (ātman). Mind and soul being the last two of the nine substances (dravya) enumerated in sūtra 1.1.4, Adhyāya 3 completes the discussion of the substances, the earlier ones having been enumerated in Adhyāya 2, as follows: earth (prthivi) 2.1.1; water (ap) 2.1.2; fire (tejas) 2.1.3; wind (vāyu) 2.1.4; ether (ākāsa) 2.1.5; wind 2.1.9 f.; ether 2.1.26 f.; time (kāla) 2.2.6 f.; space (dis) 2.2.12 f.
Why was the treatment of inferential and perceptual knowledge inserted into the discussion of the substances? The answer is obvious: sūtra 3.2.1, which proves the existence of the mind, refers back to the definition of perception. Even when there is contact sannikarsa) between soul (ātman), sense organ (indriya) and object (artha), there may or may not be knowledge; this fact indicates the existence of a fourth factor, viz., the mind
20
It is not, of course, claimed here that the original Vaiseșika Sütra was already divided into Adhyāyas and Ahnikas.

Page #11

ONCE AGAIN VAIŠEȘIKA SŪTRA 3.1.13
675
(manas).21 The fact that the soul is dealt with after the mind is explained by the fact that sūtra 3.2.4, which proves the existence of the soul, presents as one of the arguments the movement of the mind (manogati).
What was the purpose of the commentator who changed the order of the sūtras? Again it is not difficult to divine the answer. This commentator apparently wanted the discussion of the substances to continue without interruption. After the treatment of space (dis) the next substance mentioned in sūtra 1.1.4 was the soul (ātman). He introduced this topic in the way we now know, i.e., before the mind.
One final question must be addressed: When did the change of order take place? Better perhaps: when was the commentary written which commented upon the sūtras in their modified order? Here22 we have to consider the following statement in Praśastapāda's Padārthadharmasangraha (Ki p. 97 1. 25-26, Ny p. 219 1. 3, Vy I p. 134 1. 17-18): atmalingādhikāre buddhyādayah prayatnāntāh siddhāh "In the section on inferential marks of the soul (the qualities) from consciousness (buddhi) to effort (prayatna) have been established." The early commentators on the Padārthadharmasangraha all agree that this statement refers to the Vaiseșika Sūtra. Sridhara (Ny p. 219 1. 9) specifies that the reference is to the prāņāpānādisutra. Udayana provides the following commentary (Kip. 98 1. 10-11): prāņādisütre buddhyādayah prayatnāntāh siddhāh yady api buddhis tatra kantharavena nästi tathāpi sukhādaya eva svakāraṇatayā tām ākşipanti/ "In the prāņādisutra the qualities from consciousness to effort have been established. Although consciousness does not figure explicitly in that (sütra), [the qualities) happiness (sukha) etc. suggest it as it is their cause." These remarks show that the three early commentators on the Padārthadharmasangraha, as perhaps Praśastapāda himself, knew sūtra 3.2.4 more or less in the form which we find in versions C and U (pränäpānanimeşonmesajivanamanogatindriyantaravikārāḥ sukhaduḥkhe icchädveşau prayatnas cety ātmalingāni / sukhaduhkhecchādveşaprayatnās cātmano lingāni), and not as we find it in V (prānāpānanimeşonmesajīvanamanogatindriyāntaravikārāh parātmani lingam). But there is a problem connected with the identification proposed by the commentators, as pointed out by Udayana. Sūtra 3.2.4 does not mention consciousness (buddhi). Udayana's solution to the prob
21 VS 3.2.1: átmendriyārthasannikarșe jñānasyābhāvo bhāvas ca manaso lingam.
Hattori (1966: 893 (104)) has already drawn attention to the parallelism that exists between the interpretations of VS 3.1.1-2 offered in the three oldest commentaries, and a passage in the Padarthadharmasangraha (Ki p. 84 & 86, Ny p. 176, 179 & 182, Vy p. 133-34).

Page #12

676
JOHANNES BRONKHORST
lem is not convincing. A far more convincing solution presents itself if we assume that Prasastapāda already knew the beginning of Adhyaya 3 more or less in the form in which we know it. Sūtra 3.1.2 (3), in particular, uses indriyārthaprasiddhi as inferential mark to prove the existence of the soul. Indriyārthaprasiddhi is, of course, a kind of knowledge (jñāna) or consciousness.23 Sūtra 3.1.13, too, is interpreted by some commentators (C, U) as presenting consciousness as an inferential mark of the soul.
There is no evidence, on the other hand, to believe that already the author of the Carakasamhita knew the present order and interpretation of Adhyāya 3 of the Vaiseṣika Sutra. This text enumerates a number of inferential marks of the highest self (lingani paramātmanah) in Sarirasthana 1.70-72.24 A. Comba (1987: 54 f.) has pointed out that this enumeration draws upon two sources, the one being VS 3.2.4, the other Yajnavalkyasmrti 3.174-175 (Stenzler, 1849: p. 99 of the edition).25 The elements of VS 3.2.4, Comba suggests, were taken as basis, to which the elements of the Yajnavalkyasmrti have been added. The fact that buddhi 'consciousness' figures in the list of the Caraka-samhita does not, therefore indicate that Ahnika 3.1 of the Vaiseşika Sutra was read and understood as it is at present, for buddhi occurs in the list of the Yajnavalkyasmṛti. The fact that buddhi is added after the elements occurring in VS 3.2.4, agrees with the general procedure of the author of the Carakasamhita, drawn attention to by Comba, to add the elements of the Yajnavalkyasmṛti after those taken from the Vaiseşika Sutra.
It looks, then, as if the commentator who used sūtra 3.1.1 as an introduction to a discussion of the proof of the soul, lived before Prasastapāda, though perhaps not very long before him. Some facts suggest that his new interpretation of the sutras of Ahnika 3.1 could not impose itself immediately. There is, on the one hand, the ongoing tradition of authors who look upon VS 3.1.13 as a definition of perception. Equally interesting is the fact that both the commentators Śrīdhara and Udayana fail to understand that
23 This is precisely what Sarkara Miśra says in his Upaskära (p. 85 1. 17-19): yady api jñānam eva lingam iha vivakṣitam tathap Indriyārthaprasiddhe rūpādisäkṣätkärasya prasiddhatarataya tädrūpyeṇaiva lingatvam uktam.
24 Caraka Sa. 1.70-72: prāṇāpānau nimeṣādyä jIvanam manaso gatiḥ/ indriyantarasamcaraḥ preranam dharanam ca yat// desantaragatiḥ svapne pañcatvagrahanam tatha/ dṛṣṭasya dakṣinenākṣṇā savyenävagamas tatha// iccha dveṣaḥ sukham duḥkham prayatnas cetana dhṛtiḥ/ buddhiḥ smṛtir ahankäro lingāni paramātmanaḥ//.
25 Yajnavalkyasmṛti 3. 174-75 (ed. Stenzler): ahankaraḥ smṛtir medha dveşo buddhiḥ sukham dhṛtiḥ/ indriyantarasañcara iccha dharaṇaj Ivite// svargaḥ svapnaś ca bhāvanām preraṇam manaso gatiḥ/ nimesas cetanāyatna ādānam päñcabhautikam//.

Page #13

ONCE AGAIN VAIŠEŞIKA SŪTRA 3.1.13
677
Prasastapāda's buddhi in the statement cited above refers to sūtra 3.1.2 and/ or 13 (Vyomaśiva's position cannot be determined with certainty). Do we have to conclude that they still knew the earlier interpretation, perhaps even the earlier order of the sūtras in Ähnika 3.1? Did the two interpretations of Ahnika 3.1 exist for a while side by side? In this connection it is interesting to cite Hattori's (1968: 134-35 n. 4.3) paraphrase of some remarks from Jinendrabuddhi's commentary on Dignāga's Pramāṇasamuccaya: "Jinendrabuddhi remarks that the relation of VS, III, i, 13, to the preceding sūtras is variously interpreted by different commentators. He refers to the following two interpretations: (1) The universal apprehension (prasiddhi) is nothing other than knowledge (jñāna). It therefore follows that it is an attribute (guna), and is non-eternal (anitya). That which is non-eternal has a cause (kārana). Thus the sūtra in question indicates the cause of knowledge and also mentions that knowledge as an effect is different from its causes, as a pot as an effect is different from its cause, clay. (2) Since the preceding sūtras explain anumāna, one might consider anumāna as the only pramāņa. VS, III, i, 13, forestalls this by mentioning pratyakşa as a separate pramāņa. As Jinendrabuddhi says, VS, III, I, 13, can be understood as providing the definition of pratyaksa according to the second interpretation but not the first. Dignāga's implication when he says "by a certain relation [to the preceding sūtras)' (kenacit sambandhena) should be understood as referring to these different interpretation; (Pramāṇasamuccaya-ţikā] [Sde-dge ed., Tohoku, No. 4268] 53a.3-536.1 ([Peking ed., Tibetan Tripitaka, No. 5766] 59b.4-60a.3)." Note in particular the remark, in the second interpretation, "since the preceding sūtras explain anumāna" (rtags las byung ba tshad mar ba rjod la, which Muni Jambuvijaya (1961: 174 1. 5) translates into Sanskrit laingikapramāne 'bhihite). This seems to confirm our earlier supposition that a discussion of inferential knowledge (laingikam jñānam; perhaps better laingikam pramānam?) once preceded sūtra 3.1.13, not a discussion of the inferential mark (linga) as maintained, for example, by Candrānanda. Also Akalanka must have known two interpretations of VS 3.1.13, as has been pointed out by K. Preisendanz (1989: 152).26
Preisendanz (1989: 151 n. 39) also refers to *Vimalākşa's commentary on Müla-Madhyamaka-Kārikā 14.1, translated by Walleser from the Chinese into German (1912: 90). She observes that here "VS 3.1.13 is obvi
26 Compare Akalarka's remark cited above with Tattvārtha-Värttika p. 46 1. 6-8: yasya
matam - ātmano jñānākhyo guṇaḥ, tasmāc cărthäntarabhūtah, "ātmendriyamanorthasannikarsät yan nispadyate tad anyar" iti vacanäd iti ...; also p. 50 1. 9-14.

Page #14

678
JOHANNES BRONKHORST
ously used to demonstrate the difference between perception, perceptible object, and perceiver". She thinks that this is "a related line of interpretation" to her own, in which "3.1.13 most probably serves to show that cognition, here specifically perception, is different (anyar) from the soul, the latter being a factor involved in its production" (p. 150). This, if correct, would push the reordering and reinterpretation of Ahnika 3.1 back to a date well before Kumārajiva (344-413), who translated this commentary into Chinese.
However, *Vimalākşa's remarks do not force us to draw such a conclusion. It is true that an opponent in his commentary - presumably a Vaiśeşika - uses VS 3.1.13 to demonstrate the difference between perception, perceptible object, and perceiver.27 But this can very well be done, even if one looks upon this sūtra as a definition of perception. Also when interpreted as a definition of perception, this sūtra distinguishes between the self (ātman), the object of knowledge (artha), and the perceptual knowledge which results from their contact (sannikarşa) with the mind (manas) and the sense organ (indriya). We must therefore conclude that *Vimalākşa's remarks do not constitute evidence that the "new" interpretation of sūtra 3.1.13 existed already in the fourth century C.E.
Similar remarks should be made with regard to the following passage in Vasu's commentary on the Sataśāstra:28
The unbeliever says: A disciple of Ulūka, who reads the Vaiseșika Sutra, says that knowledge and atman are different, and that therefore the atman does not fall into the state of non-eternity, and yet that it is not without knowledge. Why? "Because atman and knowledge are united just like the possessor of an ox." For example, if a man is united with an ox, he is called the possessor of an ox. In the same way, from the union of the atman, the senses, the manas, and the objects, to the atman there is an occurrence of knowledge. Because of the union of the atman with knowledge, the ātman is called a possessor of knowledge.
27 See Walleser, 1912: 90: "Frage: Selbst (ātman), Vorstellung (manas), Sinn (indriya),
Sinnesbereich (gocara): da (diese) vier Dinge vereinigt sind, ist Entstehen des Erkennens. Man kann Krug, Tuch usw., alle Dinge erkennen. Deshalb ist Sehen, zu Sehendes,
Seher." The Chinese is to be found T. 1564 (vol. 30) p. 19a 1. 13-15. 28 I thank M. Nozawa, who drew my attention to this passage, and provided me with a
translation - different from Tucci's (1929: 23-24) - which I here reproduce (with minor modifications). Nozawa points out that the latter part of the underlined portion (which corresponds to VS 3.1.13) follows Ui's and Hatani's Japanese translation. An alternative translation might be "the ātman arises as a possessor of knowledge". The Chinese occurs T. 1569 (vol. 30) p. 1715 l. 7-12.

Page #15

ONCE AGAIN VAIŠEŞIKA SUTRA 3.1.13
679
Here VS 3.1.13 is used to prove the difference between the self and knowledge. This can be done, even if 3.1.13 is considered a definition of perception.
We have come to the end of this article. It must be admitted that the reconstruction of the context and of the interpretation here presented of VS 3.1.13 cannot be definitely proved to be correct. They do, however, solve a number of problems which other interpretations had failed to solve. The least one can deduce from them is that the arguments which discard the interpretation of 3.1.13 as a definition of perception on the basis of its context are not conclusive. VS 3.1.13 can be interpreted as a definition of perception, as maintained by our earliest witnesses; and other difficulties surrounding Adhyāya 3 can be solved, if only we are willing to consider the possibility that the order of sūtras which we find in the surviving versions of the Vaisesika Sūtra may in one point deviate from their original order. This possibility in its turn, as we have seen, is supported by the fact that other commentators on Sūtra texts are known to have occasionally changed the order of sūtras on which they comment. In view of all this, we may conclude with a variant of an observation made by Oetke (1988: 310): The hypothesis that the sūtras of Ahnika 3.1 have reached us in their original order may not be less speculative than the opposite assumption.
References
Akalanka: Tattvartha-Vārttika. Edited by Mahendra Kumar Jain. 2nd ed. Delhi: Bharatiya
Jnanpith. 1982. (Jnanpith Murtidevi Jaina Granthamala, Sanskrit Grantha no. 10.) Bronkhorst, Johannes (1981): "Yoga and seśvara samkhya." Journal of Indian Philosophy
9, 309-320. Bronkhorst, Johannes (1985): "A possible quotation from the Niruktavārttika known to
Durga in the Yuktidepikā." In: Proceedings of the Fifth World Sanskrit Conference
(Varanasi, October 21-26, 1981). New Delhi: Rashtriya Sanskrit Sansthan. Pp. 90-100. Bronkhorst, Johannes (1992): "Two literary conventions of classical India." Etudes Asiatiques
Asiatische Studien 45(2), 1991 (1992), 210-227. Bronkhorst, Johannes (1993): "The Vaišesika vākya and bhāşya." Annals of the Bhandarkar
Oriental Research Institute 72-73 (1991-92), 145-169. Bronkhorst, Johannes (1993a): "Studies on Bharthari, 5: Bharthari and Vaišeşika." Etudes
Asiatiques / Asiatische Studien 47 (1), Proceedings of the First International Confer
ence on Bharthari, pp. 75-94. Bronkhorst, Johannes (1994): "The qualities of Sāmkhya." Wiener Zeitschrift für die Kunde
Südasiens 37 (Festschrift G. Oberhammer), 309-322.

Page #16

680
JOHANNES BRONKHORST
Carakasamhita: The Charakasamhita by Agnivesa, revised by Charaka and Dridhabala, with the Ayurveda-Dipika Commentary of Chakrapāṇidatta. Edited by Vaidya Jadavaji Trikamji Acharya. Third edition. Published by Satyabhāmābāi Pandurang, for the Nirnaya Sāgar Press, Bombay. 1941. Comba, Antonella (1987): "Carakasamhita, Särirasthāna I and Vaiśeşika philosophy." In: Studies on Indian Medical History. Ed. G. Jan Meulenbeld and Dominik Wujastyk. Groningen: Egbert Forsten. (Groningen Oriental Studies, 2.) Pp. 43-61. Filliozat, Pierre-Sylvain (1991): Le Tantra de Svayambhu, vidyapada, avec le commentaire de Sadyojyoti. Genève: Droz.
Hattori, Masaaki (1966): "Studies of the Vaiseşikadarśana (1): On the Vaiśeşikasūtra III, i,
13." Journal of Indian and Buddhist Studies 14, 902-890/(95)-(107). Hattori, Masaaki (1968): Dignaga, On Perception. Being the Pratyakşapariccheda of Dignaga's
Pramanasamuccaya, from the Sanskrit fragments and the Tibetan versions translated and annotated. Cambridge, Mass.: Harvard University Press. (Harvard Oriental Series, 47.)
Honda, Megumu (1990): "A reading in the Vaiseşika Sutra." Dobô Daigaku Kiyo 4, 172-79 (=(1)-(94)).
Isaacson, H. (1990): A study of early Vaiseşika: the teachings on perception. Thesis University of Groningen.
Jambuvijaya, Muni (ed.) (1961): Vaiseṣikasūtra of Kanāda, with the Commentary of Candrananda. Baroda: Oriental Institute. (Gaekwad's Oriental Series, 136.) Jambuvijaya, Muni (ed.) (1966): Dvādasaram Nayacakram of Acarya Sri Mallavādi Kṣamāśramana, with the commentary Nyāyāgamānusāriņt of Śn Simhasūri Gani Vādi Kṣamāśramana. Part I. Bhavnagar: Sri Jain Atmanand Sabha. (Śr Atmanand Jain Granthamālā Serial No. 92.)
Jayantabhaṭṭa: Nyāyamañjarī. Vol. I. Edited by K.S. Varadacharya. Mysore: Oriental Research Institute. 1969.
Nozawa, Masanobu (1989): "On the interpretation of Vaiseṣikasūtra 3.1.14." Hokkaido Journal of Indological and Buddhist Studies 4, 62-74.
Oetke, Claus (1988): 'Ich' und das Ich. Analytische Untersuchungen zur buddhistischbrahmanischen Atmankontroverse. Stuttgart: Franz Steiner. (Alt- und Neu-Indische Studien, 33.)
Prasastapāda: Padarthadharmasangraha. 1) Ki: Prasastapādabhāṣyam with the commentary Kiraṇāvali of Udayanācārya, edited by Jitendra S. Jetly. Baroda: Oriental Institute. 1971. 2) Ny: Nyayakandalı [of Sridhara], being a commentary on Prasastapadabhāṣya, with three sub-commentaries, edited by J.S. Jetly and Vasant G. Parikh. Vadodara: Oriental Institute. 1991. 3) Vy: Vyomavati of Vyomasivacārya, edited by Gaurinath Sastri. Varanasi: Sampurnanand Sanskrit Vishvavidyalaya. 2 vols. 1983-1984. Preisendanz, K (1989): "On ätmendriyamanorthasannikarṣa and the Nyaya-Vaiseṣika theory of vision." Berliner Indologische Studien 4/5, 141-213.
Sankara Miśra: Upaskāra. In: Vaisesikadarsane maharṣipravara-Prasastadevācāryaviracitam Prasastapādabhāṣyam vidvaccuḍāmanis-Sankara-Misravinirmitaḥ Upaskāras ca. Ubhayatra Kasistha-Vedavidyalayadhyapaka-Nyayopadhyaya Pam. Dhuṇḍhirājaśästrikṛtam Vivaranam. Sarvam etat samskṛtam uktaśāstriṇaiva. Kāśī: Caukhambā. 1923. Sinha, Nandalal (tr.)(1911): The Vaiseṣika Sutras of Kanāda with the Commentary of Sankara Misra and Extracts from the Gloss of Jayanārāyaṇa, together with notes from the commentary of Chandrakanta and an introduction by the translator. Reprint. Delhi: S.N. Publications. 1986.

Page #17
ONCE AGAIN VAISESIKA SUTRA 3.1.13 681 Solomon, Esther (1974): The Commentaries of the Samkhya Karika - A Study. Ahmedabad: Gujarat University. Stenzler, Adolf Friedrich (1849): Yajnavalkya's Gesetzbuch. Sanskrit und Deutsch. Neudruck der Ausgabe 1849. Osnabruck: Biblio Verlag. 1970. Thakur, Anantalal (ed.) (1957): Vaisesikadarsana of Kanada, with an anonymous commen tary. Darbhanga: Mithila Institute. Thakur, Anantalal (1966): "Studies in a fragmentary Vaisesikasutravstti." Journal of the Oriental Institute (Baroda) 14 (1965/66), 330-335. Thakur, Anantalal (ed.) (1985): Bhattavadindraracita-Vaisesikavartika-Krsnabhupalaracita Trisutriprakasa'jnatakartrkavrttibhir vilasitam Maharsi-Kanada-prantam Vaisesika darsanam. Darbhanga: Kamesvarasimha-Darabhanga-Samsksta-Visvavidyalaya. Tucci, Giuseppe (1929): Pre-Dinnaga Buddhist Texts on Logic from Chinese Sources. Sec ond edition. Madras: Vesta Publications. 1981. Walleser, Max (1912): Die Mittlere Lehre des Nagarjuna. Nach der chinesischen Version ubertragen. Heidelberg: Carl Winter. Yuktidipika. Edited by Ram Chandra Pandeya. Delhi: Motilal Banarsidass. 1967.