Qualities Of Sankhya

Author(s): Johannes Bronkhorst
Publisher: Johannes Bronkhorst
Catalog link: https://jainqq.org/explore/269584/1

Page #1

THE QUALITIES OF SANKHYA
By Johannes Bronkhorst, Lausanne*
1. Bhartrhari's commentary on the Mahābhāsya contains, in the first Ahnika, the following remark concerning the Sānkhya philosophy (CE IV/23,21-23): na hīdam śāstram kasya cid ekasya sahāyabhūtam sarvasādhāraṇam | yathaiva sankhyādinām dravyâd eva prati pattiḥ rūpādisamavāyo ghato 'rthāntarabhūto veti yasya yo ghatas tasmin ghataśabdam prayunkte 'For this science (of grammar) is common to all and does not side with anyone. For example, according to the Sankhyas and others the understanding derived from a substance is that a vase is a collection of colour(s) and so on, or something else; (the grammarian) uses the word "vase" with regard to that which constitutes a vase for the person with whom he is in discussion).'
What interests us in this passage is the passing reference to the Sānkhya position, according to which a vase is a collection of colour(s) and so on. A similar statement occurs in the Vākyapadīya (VP III 13,14):
sarvamūrtyātmabhūtānām sabdādīnām gune-gune /
trayah sattvādidharmās te sarvatra samavasthitah // "Those three characteristics, sattva etc., which are found in each quality from among sound etc. which constitute all corporeal objects, are present everywhere.' The mention of the three characteristics, sattva ete.' - i. e., sattva, rajas, and tamas – leaves no doubt that the system of thought referred to is, again, Sankhya?.
Bhartrhari does not stand alone in attributing to Sānkhya the position that material objects are collections of the qualities colour, sound, etc. Punyarāja's commentary on the second Kānda of the Vākyapadīya may refer to the same view in the following passage (VPT 63,16f. [ad VP II 135):
* I thank Eli Franco and Albrecht Wezler for useful suggestions. 1 = AL 28,11-13; SW 33,22-24; MS 9c5–7.
2 The commentator Helārāja, interestingly, tries to show that sound etc. only seem to constitute corporeal objects (VPP 11/138,21f.): vyatireke 'pi dravyasya samavāyava sāt tadatmakam iva.
WZKS 38 (1994) 309–322

Page #2

310
J. BRONKHORST
vaiseșikenāvayavinam prati pădayitum ghatasabdah prayuktaḥ sānkhyair gunasamahāramātram abhimanyate jainasaugataih paramāņusamcayamatram iti "The Vaiseșika uses the word "vase" to designate the whole; the Sānkhyas think that it is used to designate the collection of gunas and nothing else; the Jainas and Buddhists, only a heap of atoms.' There is some ambiguity in this statement in as far as the Sānkhyas are concerned: the term guna does not only mean quality in this system of thought; it can also refer to the three constituents (sattva, rajas, tamas) of primary matter.
No such ambiguity attaches to Dharmapāla's introductory remarks to Aryadeva's Catuḥsataka 301 (tr. TILLEMANS 1990: 1/135): "[The Sāmkhya philosopher] Kapila asserts [the following]: Things such as vases and cloths are established simply as [visual] forms3 ... and other such [properties]; the natures ... , which are the objects of the sense organs, do really exist". Simhasūri, similarly, ascribes to Sānkhya the view that vases etc. (ghatādi) are collections of colours etc. (rūpādisamūha)
All these statements - as well as others from Malla vädin's Dvādaśāra Nayacakra and Kaiyata's commentary on the Mahābhāsya, to be considered below - support Bhartrhari's claims according to which the Sankhyas looked upon material objects as being constituted of 'colour(s) etc.' (rūpādi), or of 'sound etc.' (sabdādi). It seems moreover clear that 'colour(s) etc.' and 'sound etc.' in these statements refer to the five qualities colour, taste, smell, touch and sound.
2. It is not easy to reconcile the contents of these statements with classical Sānkhya doctrine as presented in the Yuktidīpikā, the most elaborate commentary on Isvarakrsna's Sānkhya Kārikā. There, it may be recalled, the material world is conceived of as having evolved out of prakrti through a number of intermediate stages. Material objects are considered to consist of the five elements: earth, water, fire, wind and ether. Qualities are not even mentioned among the 25 tattvas which constitute the world. In fact, the elements that do figure among the 25 tattvas possess qualities: ether possesses only sound; wind possesses sound and touch; fire possesses sound, touch and colour; water possesses sound, touch, colour and taste; earth, finally, possesses sound, touch, colour, taste and smell.
These five elements are believed to have directly evolved out of five tanmātras, which carry the names of the five qualities without
3 TILLEMANS thus translates the Chinese equivalent of skt. rūpa, 'form, colour'.
4 DNC 1/266,9. For Simhasūri's interpretation of this statement, see section 3 below (p. 313f.).

Page #3

The Qualities of Sānkhya
311
being qualities themselves. The distinction between tanmātras and qualities is clear from the following passage (YD 118,14-16): sabdagunāc chabdatanmātrād ākāśam ekaguņam | sabdasparsagunāt sparsatanmātrād dviguno vāyuḥ śabdasparsarūpaguņād rūpatanmatrāt trigunam tejah | sabdasparsarūparasaguņād rasatanmātrāc caturgunā āpah sabdasparsarūparasagandhaguņād gandhatanmātrāt pañcaguņā prthivi | 'From the tanmātra [called] "sound", which has sound as quality, ether [is born), which has [that] one quality. From the tanmātra [called] "touch", which has sound and touch as qualities, wind [is born), which has [these] two qualities. From the tanmātra [called] "colour", which has sound, touch and colour as qualities, fire [is born), which has [these] three qualities. From the tanmātra [called] "taste", which has sound, touch, colour and taste as qualities, water is born), which has [these four qualities. From the tanmātra [called] "smell", which has sound, touch, colour, taste and smell as qualities, earth [is born), which has [these] five qualities. '5
Interestingly, it is not certain that the Yuktidīpikā correctly represents the position of the Sänkhya Kārikā in this respect. The Sānkhya Kārikā leaves us in doubt whether it distinguishes between the tanmatras and the qualities 'colour', 'sound', 'smell', 'taste' and 'touch'. This can be seen as follows. Recall first that several early texts, such as Ašvaghoşa's Buddhacarita (XII 18f.) and some portions of the Mahābhārata (XII 203,25-29; 294,27-29; 298,10-21; XIV 49,34f.), knew a form of Sānkhya in which the five qualities figure among the tattvas; they are here among the final evolutes, and derive from the five elements. Here, then, the qualities do figure among the five fundamental tattvas. It is true that they did not occupy the same position as the tanmātras in classical Sankhya. It is yet conceivable (though not provable, as far as I can see) that the five tanmätras, at one phase of the development of San
5. A similar passage occurs in the Mătharavrtti (on SK 22 [MV 37.5-91). The Gaudapādabhāsya and the commentary translated into Chinese by Paramărtha (see TAKAKUSU 1904) simply derive the elements from one tanmātra each, without mentioning qualities. See further n. 9, below.
6 This has been known at least since STRAUSS 1913; see also FRAUWALLNER 1927.
7 Occasionally one gets the impression that the idea of qualities as constituting the very end of the evolutionary list of tattvas is not completely unknown to classical Sānkhya. An example is the following line, quoted YD 117,13f.: upabhogasya sabdădyupalabdhir ādih gunapurusopalabdhir antah. Interestingly, YD 64,19ff. states that the qualities sound etc. are pervaded (samanu+gam) by, or have the same nature as (svarūpa), the three constituents (here called sukha, duhkha and moha), as does DNC 1/265; SK 38, on the other hand, makes a similar observation regarding the elements (bhuta), using the terms sänta, ghora and mūdha (YD 119,20f. adds that the tanmätras are not sānta, ghora and mūdha).

Page #4

312
J. BRONKHORST
khya, were the five qualities. This possibility is not contradicted by the Sānkhya Kārikā. That is to say, this text allows, besides the 'orthodox' interpretation, of an interpretation in which the tanmātras are the five qualities. Consider first SK 28ab: rūpādisu (v.l. sabdado) pañcānām ālocanamatram isyate vrttih / 'The function of the five (sense organs] with regard to colour (v.l. sound) etc., is deemed to be mere perception'. Here it is possible to take colour (v.l. sound) etc.' to be the five qualities of those names. SK 34ab, on the other hand, has: buddhindriyāni teşām pañca viseşāvisesavisayāni / 'Of the [tenfold external organ] the five sense organs have the višesas and the avisesas as objects'. The meanings of visesa and avisesa are explained in SK 38a-c: tanmātrāny avisesās tebhyo bhūtāni pañca pañcabhyah / ete smrtā višeşāḥ "The tanmātras are the avisesas. From those five [arise] the five elements; these are known as the visesas'. According to SK 34ab, then, the sense organs have as objects the five elements and the five tanmātras. If it is true that five qualities are the objects (SK 28a, as interpreted above), one might think that the five tanmātras are the five qualities. Nothing in the Sānkhya Kārikā militates against this view, as far as I can see.
It would be premature to draw far-reaching consequences from the lack of clarity of the Sānkhya Kārikā. It is not at all certain that it looked upon the tanmātras as qualities. But if it did, this would not be without interest in connection with the various quotations maintaining that in Sānkhya material objects are collections of qualities..
3. We must now consider a passage in Mallavādin's Dvādaśāra Nayacakra which criticizes the Sānkhyas. This passage reads, in the reconstruction of Muni JAMBUVIJAYA (DNC I/268,1f.):
8 Most of the commentaries hasten to add that the tanmätras are not grasped by the sense organs of ordinary mortals (often asmadādi). It is here further to be noted that SK does not appear to justify the translation "subtle elements" or the like for tanmatra. Stanza 39 rather speaks of a subvariety of the visesas that are sūksma 'subtle'; these sūksma visesas 'subtle elements' are clearly not aviseşas, i. e., tanmatras.
9 FRAUWALLNER (1953: 355f.; also 1927: 2 = 141]) claims that in early Sankhya the different tanmātras each had only one quality (cp. YD 91,7 [ekarūpāni tanmātrānīty anye | ekottarānīti vārsaganyah 1] and 118,12f., also Vācaspati Misra's Tattvakaumudi on SK 22 [sabdatanmātrād akāśam sabdagunam sabdatanmātrasahitāt sparsatanmātrād vāyuh sabdasparsagunah sabdasparsatanmātrasahitād rūpatanmātrāt tejaḥ sabdasparsarūpagunam | sabdasparsarūpatanmātrasahitād rasatanmātrādāpah sabdasparsarūparasagunah sabdasparsarūparasatanmātrasahitād gandhatanmātrāc chabdasparsarūparasagandhagunā prthivi jāyata iti) and similar statements in the Candrikā and Jayamangalā). This position is of course but one step removed from the above tentative suggestion that the five qualities once occupied the place of the tanmātras.

Page #5

The Qualities of Sankhya
313
atha katham ekakāranatva pratişedhānantaram sabdai kagunapravrtti viyad abhyu pagamyate na pravartetaivam asamdruteḥ puruşavad vandhyāputravad vā | 'But how can the Sānkhyas) accept that ether is produced from the single quality sound, immediately after rejecting [the possibility] that something has one single cause? It cannot be produced in this way, because [ether] is not a collection, just as a soul (purusa) or the son of a barren woman is not a collection).' The commentator Simhasūri cites in connection with the term asamdruteh 'because ether is not a collection', the following phrase from the Mahābhāsya: gunasamdrāvo dravyam 'a material object is a collection of qualities'. We shall pay further attention to this phrase below (p. 317). Here it is sufficient to note that Simhasűri is most probably correct in attributing to Mallavādin the belief that the Sănkhyas looked upon the material objects as collections of qualities.
WEZLER (1985b: 3ff.) interprets the above passage in the light of the YD passage cited in section 2 above (p. 317). To quote his own words (p. 5): "The gist of Mallavādin's counterargument is hence that ether cannot originate in the manner asserted by the Sāmkhyas because it does not correspond to their definition of dravya, i. e. because it is not a dravya or rather because its cause, the sabdaguna sabdatanmatra, is not a dravya just like the soul or the son of a barren woman". This interpretation is not, however, free from difficulties. First of all, the words sabdai kagunapravrtti viyat in the above passage translate most naturally as 'ether is produced from the single quality sound'. The alternative translation 'ether is produced from the sabdatanmātra] which has sound as its only quality is decidedly more artificial. Moreover, if the latter interpretation had been intended by Mallavādin, his remark 'immediately after rejecting [the possibility) that something may have one single cause' (ekakāraṇatva pratişedhānantaram) would be besides the point. As can be seen from the YD passage cited above (p. 31!), each of the elements, not only ether, is there presented as deriving from a single cause, viz., from the corresponding tanmātra.
It will hardly be necessary to point out that Mallavādin's passage allows of an interpretation in the light of what we have discussed in section 1 above. The material world is constituted of the qualities sound etc.; these qualities are accordingly the causes of all material objects. Ether has but one quality, sound, and therefore but one cause. This, however, goes against the rule that every product must have more than one single cause.
WEZLER's interpretation of Mallavādin's passage can, in view of the above, be replaced by one that does more justice to its precise wording. Interestingly, WEZLER's interpretation appears to coincide with the one offered by Mallavādin's commentator Simhasūri. This can be deduced from some phrases in the latter's Nyāyāgamānusāriņī.

Page #6

314
J. BRONKHORST
Consider first the following passage (DNC I/268,4-6): yady anekātmakaikakāraṇatvam isyate evam ekakāranatvapratisedhānantaram ... katham sabdaikaguņapravrtti viyad abhyu pagamyate. The difficulty connected with ekakāranatva pratiședhānantaram, pointed out above, is here avoided by superimposing a different interpretation on this term. The "rejection of [the possibility) that something may have a single cause' becomes here the requirement that something has a single cause which has a multiple nature. This requirement fits, of course, the different tanmātras which are single causes of the corresponding elements, but have several qualities. Simhasūri is equally careful to avoid the difficulty presented by the word sabdaikagunapravrtti. He cites an unfindable Dhātupāthalo in order to interpret the problematic gunao as "number". The aim of this procedure seems, once again, to force the orthodox version of Sankhya upon a recalcitrant text. It appears, then, possible that Simhasūri, unlike Mallavādin, is acquainted with a form of Sānkhya in which tanmātras, and not qualities (guna), figure in the list of evolutes, or perhaps one in which tanmātras and gunas had come to be differentiated. Be it noted that another passage of his Nyāyāgamānusāriņi (11/470,13) enumerates mahat, ahamkāra and the tanmātras, three evolutes which succeed each other in classical Sānkhya.
If our interpretations of Mallavādin and Simhasūri are correct, we have stumbled upon an interesting difference between these two authors. Mallavādin, it appears, was not yet acquainted with Sānkhya in its 'classical' form. Simhasūri, on the other hand, was no longer aware of the earlier form of Sānkhya known to Mallavādin, and felt obliged to reinterpret the latter's words so as to arrive at an understanding that was in agreement with the form of Sankhya that he knew.
4. The conclusion we have to draw from the preceding sections is that a number of classical authors appear to have known the Sankhya system of thought in a form which was in at least some points different from the classical system as it has been handed down to us. The Sānkhya known to Bhartrhari, Mallavādin and others had, we have been led to believe, the qualities sound, colour, taste, touch and smell among its evolutes. Interestingly, this position is primarily known to us through texts that were no school-texts of the Sānkhyas, most notably a number of passages in the Mahabharata. We have seen, however, that the Sānkhya Kārikā itself may have held a similar position.
10 See DNC 1/268n. 3 and WEZLER 1985b: 27n. 14.

Page #7

The Qualities of Sānkhya
315
It seems probable that Bhartrhari and the other authors we have discussed found the position they attributed to the Sankhyas in one or more texts belonging to that school. And there can hardly be any doubt that that text - or one of those texts – is the one called vārsagana tantra by Simhasūri, and which FRAUWALLNER (1958: 94 [= 233]) identifies as the Şastitantra of Vrşaganall. This text was known to Dignāga and Mallavādin, as FRAUWALLNER has shown. If indeed Bhartrhari was acquainted with it, its date of composition must precede him, too.
How is it possible that Simhasūri who, like Mallavādin, knew the Sastitantra, gives evidence of being acquainted with a different version of Sankhya? FRAUWALLNER 1958 may provide the elements for an answer. Already Dignāga's commentator Jinendrabuddhi, FRAUWALLNER argues (p. 109 [= 248] and 113 [= 252]), knew at least two, possibly three commentaries on the Şaştitantra. It is not impossible that one of these commentators was, or was close to, the author of the Yoga Bhāsya (p. 114f. [= 253f.]).
It is not our task at present to take position with regard to FRAUWALLNER's conclusions, which contain inevitably a speculative element. Be it however noted that Simhasüri's deviation from Mallavādin in the interpretation of Sānkhya doctrine fits in very well with the assumption that Sankhya philosophy evolved – and therefore changed - through the reinterpretation(s) by its commentators of its classical text, which may have been called şastitantra. This assumption would, of course, agree very well with the hypothesis presented in section 2 above, according to which the Sankhya Kārikā would still precede the modification which finds expression in its commentaries.
5. The above reflections suggest that a major change took place in Sankhya doctrine, perhaps some time in the 5th century of our era. What could possibly have been the reason of this change? Why should Sankhya abandon the idea that material objects are nothing but collections of qualities?
These questions do not, at present, allow of a certain and indubitable answer. There are simply no texts from the period that might provide such an answer. It is yet very tempting to suspect a connection with the satkāryavāda, the doctrine according to which effects (or products) pre-exist in their causes. This doctrine of classical Sānkhya
11 For the authorship of this text, see OBERHAMMER-16. The name of its author may rather have been Vārşaganya, see CHAKRAVARTI 1951:137f., LARSON 1987: 624n. 21 and WEZLER 1985a: 14n. 6.

Page #8

316
J. BRONKHORST
is already known to Aryadevall and Mallavādin (DNC 1/271).13 It must therefore have co-existed with the view that material objects are nothing but collections of qualities for at least some time. Yet the two are strange bedfellows.
In order to accomodate the doctrine of satkāryavāda, classical Sānkhya views the world as a continuous series of modifications parinama) of substrates which do not lose their essence.14 The Yuktidīpikā defines pariņāma in the following stanza (YD 75,6f.):
jahad dharmāntaram pūrvam upādatte yadā param/
tattvād apracyuto dharmi parināmaḥ sa ucyate // When the substrate (dharmin), without abandoning its essence, drops the earlier property (dharma) and accepts the next one, that is called modification (parināma).' Essential in this definition is that the substrate remains in each modification, without abandoning its essence. That is to say, material objects are more than mere collections of properties; there is necessarily something more to them, viz. the all-important substrate 15
12 E.g., Catuḥsataka XI (LANG 1986, esp. p. 106f.); see further HONDA 1974.
13 Several authors (FRANCO 1991: 127; JOHNSTON 1937: 25; LARSON 1969: 165; LIEBENTHAL 1934: 9n. 11) have drawn attention to the fact that satkāryavāda is without clear precedents in the earlier literature, and must be a relatively late development in Sankhya. Regarding the origin of this doctrine we may recall LIEBENTHAL's question, "ob nicht vielleicht satkārya selbst nur ein Aspekt einer Diskussion mit Madhyamika-Buddhisten ist" (1934: 4).
14 The ultimate substrate is, of course, known by the name prakrti or pradhāna.
15 This is how we must read YD 51,17f.: asmākan tu kāra pamätrasyaiva samghätäd äkäräntaraparigrahād vā kriyāgunānām pracitir vyaktivise so bhavatiti bruvatäm adosah "But [this] reproach is not valid for us because what we teach is that a particular manifest thing originates as the accumulation of movements and qualities on account of the cause and nothing but the cause having coagulated or having assumed another shape" (WEZLER 1985b: 22). This passage occurs in a discussion about the question whether the effect pre-exists in its cause, the famous satkāryavāda. The opponent argues that if the effect were there, it should be observable, which it is not; and if it is not observable, one should be able to infer it on the basis of its movements and qualities, which, again, is not the case. Here the author of the Yuktidīpikā responds that one can only search for the movements and qualities of an effect as distinct from those of the cause, if one assumes that cause and effect themselves are distinct, which Sankhya denies; cf. YD 51,15-17: kāryakāranaprthaktvavādinas tatkriyāgunānām prthaktvam anumătum yuktam ity atas tantvavasthāne patakriyāgunagrahanād anumānābhāva ity ayam upalambhah sävakāśaḥ syät | 'For him who holds that effect and cause are separate, it is appropriate to infer that their movements and qualities are separate. For this reason the reproach can be made that, in the state of a mere] thread (and no cloth), no [cloth can be inferred on the basis of the observation of the movements and qualities of [that] cloth (precisely because

Page #9

The Qualities of Sankhya
317
The Yoga Bhasya offers a similar definition of pariņāma at the very end of its commentary on YS III 13 (Ybh. 255,7f.): avasthitasya dravyasya pārvadharmanivṛttau dharmantarotpattiḥ pariņāma iti I 'parinama is the production of a new property in a substance which remains the same, while the earlier property is destroyed'. It is true that the Yuktidipika finds fault with this definition, but its criticism concerns the use of the terms 'production' (utpatti) and 'destruction' (nivṛtti)16, certainly not the part which states that the substance remains the same. Is it conceivable that Sankhya changed its view about the nature of material objects under pressure from the satkāryavāda?
6. To conclude this article we have to consider two statements that occur in the Mahabhāṣya. This text, whose author is called Patañjali, is one of the very few texts of early India that can rather precisely be dated: it belongs almost certainly to the middle of the second century B.C.E.
The first statement that interests us is gunasamdravo dravyam (Mbh. II/366,26); the second one reads guṇasamudayo dravyam (II/ 200,13f). Both phrases are practically synonymous, and state that material objects are collections of qualities. There is no reason to believe that they express the opinion of the author of the Mahabhäṣya, yet they prove that this view existed in his days. Mbh. II/198,5 specifies what is meant by gunas: sound (sabda), touch (sparsa), colour (rupa), taste (rasa), and smell (gandha). There is no reason to think that the gunas that constitute material objects are different from these five.
Can we conclude from these two phrases that some form of Sankhya was known to the author of the Mahābhāṣya? This would not be without danger, the more so since the Mahabhasya contains, to my knowledge, no clear indications to that effect. What is more, the view of matter as a collection of qualities was not the exclusive property of
these latter are not observed)' (or, reading with WEZLER [1985b: 21] patakriyagunagrah, 'no [cloth can] be inferred because no movements and qualities of [that] cloth are observed'). Our phrase follows immediately after this remark. It will be clear that there is no question anywhere in this discussion of objects being nothing but accumulations of movements and qualities. Movements and qualities come in because they distinguish the effect from its cause, not because they constitute either or both of the two. Essentially effect and cause are not distinct, precisely because they are not made up of movements and qualities. Note, to conclude, that the Yuktidīpikā cites a stanza which describes bodies, as well as vases, etc., as nothing but collections of sattva etc. (YD 133,1f.): tasmāt samghātamātratvāt sattvādīnām ghaṭādivatā brahmanaḥ parijñāya dehānām anavasthitim ||.
16 See HALBFASS 1992: 200f.

Page #10

318
J. BRONKHORST
the Sānkhyas: the Sarvāstivādins held similar views, as has been correctly pointed out by WEZLER (1985b: 32n. 82). And whereas the Mahābhāsya contains no clear indication that its author knew the Sānkhya doctrine, there is reason to believe that he was acquainted with the teachings of the early Sarvāstivādins17. This is not, however, the place to discuss this question in further detail.
Appendix
A solution to our problem of early Sankhya has been suggested by Nāgesa Bhatta, author of the Uddyota, a subcommentary on the Mahābhāsya. It occurs in his comments on Kaiyata's Pradīpa on the Mahābhāsya on P. IV 1,3. Kaiyata states (MP III/447): sattvarajastamāmsi gunāḥ | tatparināmarūpāś ca tadātmakā eva sabdādayaḥ pañca gunāḥ | tatsamghātarūpam ca ghatādi na tu tadvyatiriktam avayavidravyam astīti sānkhyānām siddhāntah "The doctrine of the Sānkhyas is as follows]: The gunas are sattva, rajas and tamas; the (so-called] five gunas, (viz.) sound etc., are modifications of those [three guņas] and [therefore) identical with these; and vases etc. are collections of those [five gunas), not material wholes different from those [five gunas)'.
This statement repeats the position also expressed by Bhartrhari and the other authors studied above. Nāgesa comments as follows on the word sānkhyānām (ib.): sānkhyānām iti | seśvarasānkhyānām ācāryasya patañjaler ity arthah ! gunasamūho dravyam iti patañjalir iti yogabhāsye spastam “Of the Sānkhyas" means: of Patañjali, a teacher belonging to the Sānkhyas with God. It is clear in the Yoga Bhāsya that according to Patañjali a material object is a collection of gunas'. The reference is to the Yoga Bhāsya on YS III 44, which reads (Ybh. 299,6): ayutasiddhāvayavabhedānugataḥ samāho dravyam iti patañjalih | 'According to Patañjali's, a material object is an aggregate of different component parts which do not exist separately'. Nāgesa interprets this to mean that a material object is a collection of gunas. Is this correct? And what does he mean by guna?
The statement from the Yoga Bhāsya must be read in context. It is preceded by a discussion, the most important points of which (for our present purposes) are:
17 See BRONKHORST 1987: 56ff.
18 According to HALBFASS (1992: 106n. 8), the reference is to the grammarian Patañjali. This seems doubtful, and is indeed not the opinion of Nāgesa, as we have seen.

Page #11

The Qualities of Sankhya
A material object is a collection of samanya(s) and viseṣa(s) 19. What are sāmānyas and viseṣas? The viseṣas are sound etc. – belonging to earth etc. together with their properties, shape etc. 20. The sāmānyas are corporeality (which is earth), viscosity (which is water), heat (which is fire), moving forward (which is wind)", and going everywhere (which is ether)22. The text adds that sound etc. are the viseṣas of a samanya.
There can be little doubt that both sāmānyas and viseṣas are qualities of some sort23; we may speak, with DASGUPTA (1924: 168), of generic and specific qualities. Material objects are, therefore, aggregates or collections of qualities, which are, moreover, inseparable. We may assume that we have to do here with a development of the preclassical form of Sankhya outlined above".
Does this mean that we have to believe, following the lead of Nageśa, that Bhartṛhari and the other authors cited at the beginning of this article referred to the Yoga Bhāṣya, or perhaps to a work by the mysterious Patanjali mentioned there? It seems doubtful. The sāmānyas in the Yoga Bhāṣya are never referred to as gunas25; yet Bhartṛhari uses this term in connection with 'sound etc.'.
AL
Bibliography and Abbreviations
319
BRONKHORST 1987
Mahābhāṣyadīpikā of Bhartṛhari, crit. edd. K.V. ABHYANKAR V.P. LIMAYE. [BORI Post-Graduate and Research Department Series, No. 8]. Poona: Bhandarkar Oriental Research Institute, 1970.
JOHANNES BRONKHORST, Three Problems Pertaining to the Mahābhāṣya. [Post-graduate and Research Department Series, No. 30]. Poona: Bhandarkar Oriental Research Institute, 1987.
19 Ybh. 298,6 sāmānyaviseṣasamudayo 'tra dravyam |.
20 Ybh. 297,8 parthivadyaḥ sabdadayo viseṣaḥ sahäkärädibhir dharmaiḥ. 21 FRAUWALLNER (1953: 357, 404) translates pranamitā as "Vorwärtsbewegung".
22 Ybh. 298,2f. svasāmānyam murtir bhumiḥ sneho jalam vahnir uṣṇatā vayuḥ pranāmi sarvatogatir ākāśaḥ. Ybh. 339,4ff. (ad YS IV 14) enumerates the same samanyas as murti, sneha, ausnya, pranamitva and avakāśadāna.
23 Some of the samanyas of the Yoga Bhāṣya figure among the dharmas of the elements enumerated at YD 118,21f.
24 Buddhist influence cannot be ruled out either; cf. Abhidharmakosa Bhāṣya I 12: kharaḥ prthividhatuḥ | sneho 'bdhatuḥ | uṣnata tejodhatuḥ | īraṇā vayudhatuḥ, the similarity of which with the Yoga Bhāṣya is undeniable.
25 Except, of course, by Nageśa in the passage cited above.

Page #12

320
J. BRONKHORST
CE IV
Mahābhāsyadīpikā of Bhartrhari. Fascicule IV: Ahnika I, crit. ed. J. BRONKHORST. [BORI Post-Graduate and Research Department Series, No. 28). Poona: Bhandarkar Oriental Research Institute, 1987. PULINBIHARI CHAKRAVARTI, Origin and Development of the Sāmkhya System of Thought. Repr. New Delhi: Munshiram Manoharlal, 1975.
CHAKRAVARTI 1951
DASGUPTA 1924 DNCI
SURENDRANATH DASGUPTA, Yoga as Philosophy and Religion. Repr. Delhi: Motilal Banarsidass, 1973. Dvădaśāram Nayacakram of Acārya Sri Mallavādi Kşamāśramana. With the commentary Nyāyāgamānusā - rini of Sri Simhasūri Gaņi Vādi Kşamāśramaņa. Pt. I (1-4 Aras), ed. Muni JAMBOVIJAYAJT. Bhavnagar: Sri Jain Atmanand Sabha, 1966.
FRANCO 1991
ELI FRANCO, Whatever happened to the Yuktidipikā?.
WZKS 35 (1991) 123-137. FRAUWALLNER 1927 ERICH FRAUWALLNER, Zur Elementenlehre des Sām
khya. WZKM 34 (1927) 1-5 (= Kleine Schriften, edd. G. OBERHAMMER - E. STEINKELLNER. Wiesbaden 1982,
p. 140-144). FRAUWALLNER 1953 Id., Geschichte der indischen Philosophie: I. Bd.: Die
Philosophie des Veda und des Epos - Der Buddha und der Jina - Das Samkhya und das klassische Yoga
System. Salzburg: Otto Müller, 1953. FRAUWALLNER 1958 id, Die Erkenntnislehre des klassischen Sämkhya
Systems. WZKS 2 (1958) 84-139 (= Kleine Schriften, p. 223-278).
HALBFASS 1992
HONDA 1974
JOHNSTON 1937
LANG 1986
WILHELM HALBFASS, On Being and What There Is. Classical Vaiseșika and the History of Indian Ontology. Albany, NY: State University of New York Press, 1992. MEGUMU HONDA, Arya Deva's Critique Against Samkhya. JIBS 23,1 (1974) 491-486 (= 7-12). E.H. JOHNSTON, Early Sāmkhya. An Essay on its Historical Development according to the Texts. Repr. Delhi: Motilal Banarsidass, 1974. KAREN LANG, Aryadeva's Catuhsataka. On the Bodhisattva's Cultivation of Merit and Knowledge. [Indiske Studier VII]. Copenhagen: Akademisk Forlag, 1986. GERALD JAMES LARSON, Classical Sāmkhya. An Interpretation of its History and Meaning. Delhi: Motilal Banarsidass, 1969. Id.-RAM SHANKAR BHATTACHARYA (edd.), Encyclopedia of Indian Philosophies. Vol. IV: Sămkhya. A Dualist Tradition in Indian Philosophy. Delhi: Motilal Banarsidass, 1987 WALTER LIEBENTHAL, Satkārya in der Darstellung seiner buddhistischen Gegner. Die prakrti-parīksā im Tattvasamgraha des Sāntirakṣita zusammen mit der Panjikā des Kamalasila übersetzt und ausführlich interpre
LARSON 1969
LARSON 1987
LIEBENTHAL 1934

Page #13

The Qualities of Sānkhya
321
Mbh.
MP
MS
MV
P.
SK
STRAUSS 1913
SW
tiert. Beiträge zur indischen Sprachwissenschaft und Religionsgeschichte, Heft IX). Stuttgart - Berlin: W. Kohlhammer, 1934. The Vyakarana-Mahābhāsya of Patañjali, ed. F. KIELHORN. Third Edition revised and furnished with additional readings, references and select critical notes by K.V. ABHYANKAR. Vol. I-III. Poona: Bhandarkar Oriental Research Institute, 1962-1972. Vyakarana-Mahābhāsya with Kaiyata's Pradīpa and Nägesa's Uddyota, ed. VEDAVRATA. Vol. I-V. Rohtak: Harayaņā-Sahitya-Samsthāna, 1962-1963. Mahābhāsyadīpikā of Bhartrhari. Manuscript Reproduced. Poona: Bhandarkar Oriental Research Institute, 1980. Sânkhya Karika by Iswara Krishna with a commentary by Mathara Charya, ed. P. VISHNU PRASAD SARMA.
The Chowkhamba Sanskrit Series, No. 296). Benares: Chowkhamba Sanskrit Series Office, 1922. Pāṇinian Sūtra Isvarakrsna, Sankhya Kārikā: see MV and YD. OTTO STRAUSS, Zur Geschichte des Samkhya. WZKM 27 (1913) 257-275 (= Kleine Schriften, ed. F. WILHELM. Wiesbaden 1983, p. 154-172). Mahābhāsya Tikā by Bhartrhari. Vol. I, ed. V. SWAMINATHAN. [Hindu Vishvavidyalaya Nepal Rajya Sanskrit Series, Vol. 11). Varanasi: Banaras Hindu University, 1965. J. TAKAKUSU, La Samkhyakārikā étudiée à la lumière de sa version chinoise. BEFE0 4 (1904) 1-65 & 978-1064. Tom J.F. TILLEMANS, Materials for the Study of Aryadeva, Dharmapāla and Candrakīrti. The Catuḥsataka of Aryadeva, Chapters XII and XIII, with the Commentaries of Dharmapala and Candrakirti: Introduction, Translation, Sanskrit, Tibetan and Chinese Texts, Notes. Vol. I-II. (WSTB 24,1-2). Wien: Arbeitskreis für tibetische und buddhistische Studien Universität Wien, 1990. Bhartrharis Väkyapadīya. Die Mülakärikäs nach den Handschriften hrsg. und mit einem Pada-Index versehen von W. RAU. JAKM XLII/4). Wiesbaden: Franz Steiner - Deutsche Morgenländische Gesellschaft, 1977. Väkyapadiya of Bhartrhari with the Prakirnakaprakāśa of Helarāja. Kända III, Pt. ii, crit. ed. K.A. SUBRAMANIA IYER. Poona: Deccan College, 1973. Vākyapadiya of Bhartrhari (An ancient Treatise on the Philosophy of Sanskrit Grammar) Containing the Tīkā of Punyarāja and the Ancient Vrtti. Kānda II, ed. K.A. SUBRAMANIA IYER. Delhi: Motilal Banarsidass. 1983.
TAKAKUSU 1904
TILLEMANS 1990
VP
VPP II
VPT
WEZLER 1985a
ALBRECHT WEZLER, A Note on Vārsaganya and the Yogācārabhūmi. JASC 27,2 (1985) 1-17.

Page #14
322 J. BRONKHORST WEZLER 1985b Ybh. Id., A Note on Mahabhasya II 366.26: gunasamdravo dravyam (Studies on Mallavadin's Dvadasaranayacakra II). In: Buddhism and Its Relation to Other Religions. Essays in Honour of Dr. Shozen Kumoi on His Seventieth Birthday. Kyoto 1985, p. 1-33. Patanjala-Yogasutra-Bhasya-Vivaranam of SankaraBhagavatpada, crit. edd. P. SRI RAMA SASTRI - S.R. KRISHNAMURTHI SASTRI. [Madras Government Oriental Series, No. XCIV]. Madras: Government Oriental Manuscripts Library, 1952. Yuktidipika. An ancient Commentary on the SamkhyaKarika of Isvarakrsna, ed. R.CH. PANDEYA. Delhi: Motilal Banarsidass, 1967. Yoga Sutra YD YS