STHIRAMATI'S VIBHÄSÄ: A PRELIMINARY SURVEY

STHIRAMATI'S VIBHÄSÄ: A PRELIMINARY SURVEY

Jowita KRAMER

Among the copies of Buddhist Sanskrit manuscripts kept in the China Tibetology Research Center (Beijing), one of the more important texts is Sthiramati 's commentary on Vasubandhu's Pañcaskandhaka (PSk), the Pañcaskandhakavibhä$ä (PSkV). A Tibetan translation of the PSkV is preserved in the Tibetan bsTan 'gyur as no. 5567 of the Peking edition (P). A Chinese version of a PSk commentary ascribed to Sthiramati (6th c.), which bears the title Da Cheng guang wu yun lun is available in Taishö 1613. This text does not seem to be a direct translation of the PSkV, as it is much shorter and its contents do not exactly correspond to the Sanskrit and the Tibetan versions of the PSkV. Apart from Sthiramati 's commentary, two other commentaries on the PSk are available in Tibetan translation: the Pañcaskandhavivarana by Gunaprabha (P 5568) and the Pañca skandhabhäya by *Prthivïbandhu (P 5569). This paper aims at investigating some important passages of the PSkV. The first section presents a general overview of the contents of the text, the second section deals with the purpose of the teachings of the PSkV, and the third section examines the characterization of the skandha "matter" (rüpa). The main concern of the PSk and the PSkV is the analysis of the five aggregates (skandha), twelve bases (äyatana), and eighteen elements (dhätu). As one of the central ideas of Buddhist philosophy,

This article is a revised version of a lecture I gave at Otani University (Kyoto) on March 3, 2008. I am grateful to Profs. Kazunobu Matsuda, Lambert Schmithausen, and Peter Skilling as well as to my husband Ralf who read previous drafts of this paper and made a number of valuable suggestions. I would also like to acknowledge the support received from the German Research Foundation (DFG) since 2007, which enabled me to complete this article.

the five skandhas (as well as the twelve äyatanas and the eighteen dhätus) appear to be a well studied concept. Nevertheless, a thorough investigation of the PSk(V) makes it evident that the concept of the five skandhas as it is known from early canonical sources changed when it was integrated into the Abhidharmic system of the Yogäcäras. 3 In the process of systematizing all factors of existence within the Abhidharmic literature, the early concept of five skandhas which aimed at describing merely the physical and mental parts of living beings might have appeared not adequate anymore. As it was impossible to abandon this traditional model, the meaning of some of the skandhas was extended to subsume the whole outside world. 4 The archaic scheme had to be harmonized with new teachings for instance, in the context of Yogäcära tradition, with the "store mind" (älayavijñäna) or the "notion of I" (kli$tamanas). In Vasubandhu's PSk and Sthiramati's PSkV this process of reshaping the skandha model becomes visible.

I. Subject Outline of the PSkV

The structure of the PSkV follows the arrangement of Vasubandhu's root text and can be divided into seven main parts as depicted in the following outline: 1. Introductory remarks (1b1-3b2)

2. Explanation of the five skandhas (3b2-61a1)

3. Explanation of the twelve äyatanas (61a1-64a6)

4. Explanation of the eighteen dhätus (64a6-65b4)

5. Reasons for explaining the skandhas, ayatanas and dhätus (65b467a4)

6. Mätrkä of qualities (67a4-73b1)

7. Concluding matter (73b2-5)

3 A very detailed investigation of the five skandha model in the canonical period is offered by Tilmann Vetter [2000]. Vetter made accessible all passages of the Vinayapitaka and the main Nikäyas of the Suttapi!aka (as they were known to him) mentioning the five skandhas. Through his minute approach Vetter succeeded in presenting a very precise characterization of the five skandha concept as found in the Päli canon.

4 Some important observations regarding this development are found in the studies by Erich Frauwallner [1963] and [1994: 110ff.]. However, Frauwallner focuses in his investigation almost exclusively on the Abhidharmic tradition of early Buddhism.

Sthiramati opens his commentary with a worshipful invocation and some introductory remarks. The latter include information on the author's motive for composing the text, the reasons for the number and the order of the five skandhas, and the etymology of the compounds rüpaskandha, vedanäskandha, and so on. The concluding matter consists of a final sentence mentioning the name of the author and the text as well as of a short colophon of three lines that has been added in a different script. The main and longest section of the PSkV is the second one, where the five skandhas are described and analysed in detail. An investigation of the first part of this section, which is the explanation of rüpa, is presented below (in section Ill).

The third section of the PSkV deals mainly with those two äyatanas that are not included in rüpaskandha. These are the mental faculty (manas) and its objects, the factors (dharma). The latter are said to comprise eight different entities: the three skandhas vedanä, sarnjñä, and sanskära, the "not making known" (avijñapti), and the four unconditioned (asamskrta) factors space (äkäša), cessation not resulting from consideration (apratisamkhyänirodha), cessation resulting from consideration (pratisamkhyänirodha), and the true reality (tathatä). Remarkably, the lists of the unconditioned constituents of the dharmäyatana/-dhatu presented in the Abhidharmasamuccaya and the Abhidharmakošabhäya vary from the four entities given in the PSk. In the Abhidharmasamuccaya three different kinds of true reality are mentioned: the tathatã of the wholesome factors (kušaladharmatathatä), of the unwholesome factors (akušaladharmatathatä), and of the neutral factors (avyäkrtadharmatathatä). These three entities are combined with äkäša, apratisamkhyänirodha, pratisarnkhyänirodha, the state of motionlessness (äniñjya) and the cessation of ideations and feelings (sarnjñävedayitanirodha) to form a group of eight. In AKBh 3, 16-19 äkãša, apratisamkhyänirodha, and pratisamkhyanirodha are mentioned as three kinds of unconditioned factors. Vasubandhu apparently tried to find a compromise in the PSk between the eight entities as found in the Abhidharmasamuccaya and the three entities listed in the Abhidharmakošabhä$ya. He adopted the group of three and adjusted it to the Yogäcära context of the PSk by adding tathatä to it, a central philosophical concept of the Yogäcära tradition. As twelve of the eighteen dhätus are identical to the twelve äyatanas, the fourth section of the PSkV covers only one folio of the text. The most significant information of this passage is that the manaäyatana and the six "elements of mind" (cittadhätu), i.e., the six vijñänadhätus, are to be regarded as being identical to the vijñänaskandha. 8 In the fifth section of the PSkV Sthiramati comments on Vasubandhu's explanation that the three concepts of five skandhas, twelve äyatanas, and eighteen dhätus were taught in order to eliminate the three atmagrähas. This part of the PSkV is dealt with in more detail in section Il, below.

Covering seven folios, the sixth section of the PSkV is the second longest. It includes a list of qualities (mätrkä) which are applied consecutively to the eighteen dhätus. The mätrkä treats the dhätus under the following eleven aspects:

material (rüpin) — immaterial (arüpin) visible (sanidaršana) — invisible (anidaršana) penetrable (sapratigha) — impenetrable (apratigha) pure (säsrava) — impure (anäsrava) belonging to [the sphere of] sensual pleasures/to the material [[[sphere]]]/to the immaterial [[[sphere]]] (käma-/rüpa-/ärüpyapratisamyukta) — not belonging to [any of the three spheres] (apratisamyukta)9 wholesome (kušala) — unwholesome (akušala) — neutral (avyä- krta) internal (ädhyätmika) — external (bähya) having an object (salambana) — having no object (anälambana) having a conception (vikalpaka) 10 having no conception (avikalpaka) appropriated (upätta) — not appropriated (anupätta) similar (sabhäga) — similar to itself (tatsabhäga) ll

8 PSkV 65bl: yo vijñänaskandha,þ sa manaäyatanam cittadhätavaš ca sapta cak§urvijñänadhätur yävan manovijñanadhätur iti.

9 After this category the root text of the PSk mentions the categories skandhasa?tgrhïta and upädänasamgrhïta (See PSk 7a6).

10 In PSk 7b3 this category is listed under the term savikalpa.

11 For further information on the meaning and translation of these two terms, see Kramer [2005: 128, n. 1441.

The mätrkä found in the PSk is much shorter than mätrkäs included in the Abhidharmakošabhäya, the Abhidharmasamuccaya, or in the Yogäcärabhümi, in which the lists in some cases comprise more than twenty items. Why Vasubandhu has chosen these particular qualities and excluded others is not certain. It seems that his list includes mainly qualities which — in both cases (i.e., rapin and arüpin) — describe worldly phenomena in order to classify them in material or in mental entities, and not to differentiate them from phenomena of the absolute level, such as the tathatä or nirvana. This might be one of the reasons why we find qualities like "material" or "immaterial", "internal" or "not internal", "having an object" or "not having an object" in the mätrkä and not qualities like "mundane" (laukika) or "supramundane" (lokottara), "conditioned" (samskrta) or "unconditioned" (asanskrta), "being the highest" (anuttara) or "not being the highest" (sottara). These latter qualities separate worldly, conditioned, and inferior entities like the skandhas from the unconditioned, supreme tathatä or nirvana. As our text does not deal with these supreme entities in the first place but is intended to be an analysis of the conditioned parts of a person, Vasubandhu might have chosen only those qualities that are suitable for specifying these worldly factors.

Il. The purpose of teaching the skandhas, äyatanas, and dhätus

According to Vasubandhu there is a reason to teach the three different models of skandhas, äyatanas, and dhätus: The understanding of each of the three concepts eliminates one of the three wrong views of the self (ätmagräha). The skandha analysis is taught as an antidote to the apprehension of the self as a unity (ekatvagräha). Sthiramati adds that people who claim the existence of a substantial self hold that there is a unitary self which sees, hears, smells, tastes and touches,

thinks and perceives. But there is no unitary self besides the five skandhas within the personal existence (ätmabhäva) of the living being, as neither its own nature (svarüpa) nor its effects (kärya) can be grasped. There is no seer apart from the faculty of seeing (cak$uh). In contrast, the understanding of the twelve äyatanas is taught in order to eliminate the apprehension of the self as the one that experiences (bhoktrtvagräha) the object of the sense faculties and the results of good and bad karma. But there is no such self. Actually those äyatanas that are the six sense faculties have the six kinds of experiences (upalabdhi). 17 Finally the classification of the eighteen dhätus is taught as an antidote to the apprehension of self as an independent agent (kartrtvagräha). Sthiramati explains that there is no self that effects wholesome and unwholesome karma. There is no action besides the effect of causes and results, and without an action no agent is possible.

This traditional interpretation of the three models offered by Vasubandhu and Sthiramati is just one of several possibilities to explain why it appeared necessary to add the models of the twelve äyatanas and the eighteen dhätus to the concept of the five skandhas. Another reason for the extension of the five skandha model is certainly the fact that it was impossible to include unconditioned entities, like nirvana, in it, as all the skandhas are conditioned. Within the concept of the ayatanas and dhätus unconditioned entities could be incorporated into the category of the dharmas, the object of the mental faculty (manas). Another reason for the extension of the five skandha model becomes visible when we look at the traditional

division of the eighteen dhätus. As mentioned above, the PSk explains that ten of the dhätus are to be regarded as matter and seven as mind (citta or vijñäna). Obviously the contrast between matter and mind could be expressed more clearly within the model of the eighteen elements.

It is notable that the other three skandhas, i.e., feeling (vedanä), ideation (sarpjñä), and impulses (sarnskära), lost their relevance in the context of the äyatanas and dhätus, where they were included in the dharma category. As constituents of the latter they do not have the same value and position as rüpa and vijñäna, but are merely objects of thinking. In contrast, the position of rüpa and even more so that of vijñäna strengthened. These two are the categories that underwent the most important modifications during the evolution of the five skandha model. In particular the function of vijñänaskandha — the original role of which was actual perception — was widened through the inclusion of subliminal forms of mind, like the "store mind" (älaya vijñäna) and the "notion of I" (kli$tamanas). The strong emphasis placed by Sthiramati on vijñäna is evident, for instance, when he states that ordinary people — those who have not perceived reality — regard the vijñäna as the self (atman), whereas they view the other four skandhas as "mine" (ätmïya). Interestingly Sthiramati also mentions alternative concepts of the self, for example that of the Sämkhyä tradition. According to his understanding, the Sämkhyas only regard rüpaskandha as ätmïya, and all the other four skandhas as atman. He thus claims that for the Sämkhyas the self is not only identical to vijñäna but also consists of the factors accompanying the mind (caitasika).

Ill. Definition ofrüpa

The sections on rapa- and on vijñänaskandha are of almost the same length and each cover more or less eleven folios in the PSkV. In the following I present an outline of the section on rapa and investigate its contents in detail. In this context I also try to explain some of the difficulties which the distinguishing of rapa and vijñäna posed to authors of Abhidharmic treatises. A more comprehensive study of the vijñäna section is beyond the scope of the present paper. The following outline gives a preliminary overview of the contents of the rüpa section:

1. Is the question "What is matter?" not incoherent (asambaddha)?
(3b2-4)
2. Matter of the four basic elements (mahäbhüta) and matter dependent (upädäya) on them (3b4-14a5)
2.1 Four basic elements (3b4-4b1)
2.2 Matter dependent on the four basic elements (4b1-6)
2.3 Definitions of the four basic elements, their effects (karma) and their own nature (svabhäva) (4b6-6a2)
2.4 Definition of matter dependent on the basic elements (6a214a5)
2.4.1 Sense faculties (6a3-7a2)
 2.4.1.1 Sense of sight (6a3-5)
2.4.1.2 Pellucid matter (rüpaprasäda) (6a5-7a2)
2.4.2 Objects of the sense faculties (7a2-11b6)
2.4.2.1 General remarks (7a2-4)
2.4.2.2 Visible matter (rüpa): colour, shape, "making known"
(vijñapti) (7a4-9b1)
2.4.2.3 Sound (9b1-6)
2.4.2.4 Odour (9b6-10a2)
2.4.2.5 Taste (100-3)
2.4.2.6 Tangible (100-11b6)
24.3 "Not making known" (avijñapti) (11b6-14a5)
2.4.3.1 Definition (11b6-12a1)
2.4.3.2 Different kinds of vijñapti and avijñapti (12a1-12b2)
2.4.3.2.1 The avijñapti of the sphere of sensual pleasures (kama[dhätu]) (120-6)
2.4.3.2.2 The avijñapti of the material sphere (rüpa[dhätuD and the pure (anäsrava) avijñapti (12a6-12b2)
 2.4.3.3 Avijñapti being invisible (anidaršana) and penetrable
(apratigha) (12b2-4)
2.4.3.4 Additional remarks on the avijñapti belonging to kamadhätu, the avijñapti belonging to rüpadhätu and the pure avijñapti (12b4-14a5)

3. Reasons for the term rüpa (14a5)

Sthiramati opens his commentary on the rüpaskandha section with the statement that it is impossible to determine the own nature (svabhäva) of the skandhas due to their lack of svabhäva. But it is possible to expose the basis [of their designation] (upädäna), which in the case of rüpaskandha is rüpa. Therefore, in the next passage of the text a detailed definition of the constituents of rapa is given. The PSk describes rüpa as the four basic elements (mahäbhüta) and the matter that is dependent (upädäya) on them. 26 Sthiramati adds in his commentary that space (äkäša) is not to be regarded as an additional basic element, as it is nothing other than the mere non-existence of impenetrable (sapratigha) rüpa. In connection with the explanation of matter dependent on the mahäbhütas, Sthiramati mentions five kinds of how matter derived from the elements (bhautika) is dependent on them: generating (janana), basis [of change] (nišraya), continuity (prati§tha), support (upastambha), and nourishing (brmhaaa). The first of these five indicates that derived matter could not arise without the existence of the mahäbhütas. The fact that the mahäbhütas are the basis of upädäyarüpa means, according to Sthiramati, that dependent matter changes in the same moment as the elements change. As long as the elements are produced in a certain continuum, the series of derived matter will not be interrupted — this is the meaning of the third kind of dependence, "continuity". 31 The dependence consisting in "support" indicates that the continuum of upädäyarüpa can only ex ist where the mahäbhütas exist. 32 "Nourishing" points to the fact that dependent matter can only increase if the mahäbhütas it depends on grow.33

The definitions of the four mahäbhütas given by Sthiramati (in accordance with Vasubandhu) in the following section of the PSkV closely resemble the explanations of the Abhidharmasamuccaya and the Abhidharmakošabhäya. Solidity (khakkhatatva) is said to be the nature of earth, humidity (sneha) the nature of water, heat (üynä) the nature of fire, and lightness and motion (laghusamudïranatva) the nature of wind. 34 The respective activities of the four mahäbhütas are support (dhrti), cohesion (sarpgraha), ripening (pakti), and shifting (vyühana).35 In the section on matter dependent on the basic elements, Sthiramati explains the meaning of pellucid matter (rüpaprasäda), which is the nature of the five material sense faculties. 36 He gives the example of images being reflected in a clear mirror or in a water vessel. In the same way the pellucid matter of the five sense faculties reflects their

also ASBh 3,7f. 31 PSkV 4b4f.: sthänahetutvanz bhütänämz sadršasantänotpattau bhautikasyäpi sadršasantänänucchedahetutvät. See also ASBh 3,8f.

32 PSkV 4b5: upastambhahetutvarp tadvašenänucchedät. See also ASBh 3,6.

33 PSkV 4b5f.: brmhapahetutvarn bhûtavrddhau tadä§ritopadäyarapavrddhik. See also ASBh 3,9f.

34 PSkV 5a3-5. Though the terminology found in the Tibetan translation of the Abhidharmasamuccaya is identical to the Tibetan rendering of the PSkV (see AST 53a3f., PSkT 13alf., and PSkVT 5a6-8: sra ba nyid, gsher ba nyid, tsha [ba nyid], yang thing g.yo ba nyid), the terms given by Pradhan in his reconsruction of the Sanskrit text differ from those in the PSkV: kathinatä, niyandatä, uy.zatä, kampanatä (see AS* 3,14-16). Pradhan's retranslation is probably to be rejected here. The respective terms found in AKBh 8,20 are khara, sneha, uy:zatä, Tram.

35 PSkV5b1-2.

36 According to the PSk upadayarapa consists of the five sense faculties, their five objects, and the avijñapti. The explanation of AS* 3,16-18 (AST53a4f.) is very similar, though the last category, the avijñapti, is extended to "matter included in the dharmäyatana" (dharmäyatanasamgrhïta). The answer to the question, what the five indriyas are, given in the PSk is: "subtle matter having colour as its object , having sound as its object", etc. (see PSkV 7a2f.: cak$urindriyam katamat/ varaavi$ayo rapaprasädak/ šrotrendriyaž?i katamac chabdavi$ayo rapaprasäda). The definition found in AS* 3,18f. (AST 53a5f.) differs somewhat. There it is said that the faculty of seeing is pellucid matter dependent on the four mahabhütas and the basis for visual perception (calqurvijñäna). See also AKBh 5,27ff., where a very similar explanation is found.

objects. Sthiramati mentions in this context that manas and faith (šraddha) are also considered as prasäda, but they are to be distinguished from the indriyas as they do not have matter (rüpa) for their nature.

It is worth to investigate the subsequent passage of the PSkV which defines the objects of each sense faculty in detail, insofar as it reveals interesting dissimilarities between the PSkV, the Abhidharmakošabhä$ya, and the Abhidharmasamuccaya. First of all, the object of the faculty of seeing is discussed. The PSk assigns three different categories to the visible (rüpa): colour (varna), shape (samsthäna), and "making known" (vijñapti). All three categories are mentioned in the equivalent description of the Abhidharmasamuccaya, whereas the Abhidharmakošabhäya only points out varm and samsthäna as parts of the visible and obviously places käyavijñapti under samsthäna. When going into details the PSkV, the Abhidharmasamuccaya, and the Abhidharmakošabhäya agree on four kinds of colour and eight kinds of shape. But they disagree on the question whether entities like a cloud, smoke, or the sunlight are to be classified as separate categories, or whether they are already included in the categories of colour and of shape. In this context AKBh 6,13 mentions eight additional entities: cloud (abhra), smoke (dhüma), dust (rajas), mist (mahikä), shade (chäyä), sunlight (ätapa), (other) light (äloka), and darkness (andhakära). In the Abhidharmasamuccaya space (abhyavakäša), vijñapti, and the colour of the sky (nabha) as well as two further shapes, namely fine (rdul phra mo) and rough (rags pa) shape, are added to the eight entities found in the Abhidharmakošabhäya. Sthiramati lists the entities from abhra to nabha (omitting vijñapti, which he mentions in a separate section), but he rejects them as separate constituents different from colour and shape. He argues that clouds, smoke etc. are either included in the category samsthäna, in case they are limited, or in the category varna, in case they are not limited.42

While the PSk analyses vijñapti as a third kind of visible rapa beside colour and shape, in the Abhidharmakošabhäsya the definition of vijñapti is included in another context, namely in chaper 4, in which karma is characterized. The Abhidharmasamuccaya mentions vijñaptirüpa as one of the 25 separate categories forming the visible, but it neither explains what the matter of vijñapti includes nor mentions it in its chapter on kan•na. The examination of vijñapti presented in the PSkV and the Abhidharmakošabhäsya shows that there was a dispute over the characterization of the bodily vijñapti (käyavijñapti) within the different traditions. This discussion was related to the general question of whether shape (sarnsthäna) was to be regarded as a real entity (dravya), in the same manner as varwa, or as a designation for an accumulation of colour atoms arranged in a certain way. The first view (i.e., sansthäna being dravyasat) was held by the Sarvästivädins, who regarded käyavijñapti as a kind of shape and therefore as a real entity. They classified kayavijñapti as shape without there being colour. The opposite view was taken by the Sauträntikas, who maintained that samsthäna was a mere designation and that kayavijñapti was shape that does not exist as a real entity. 46 A very similar position is held by Sthiramati in the PSkV. He also explains that shape does not exist as an entity and argues that this is evident for example from the fact that there are no differently shaped atoms in the same way as there are atoms of various colours. Therefore there are for instance no long-shaped atoms in an accumulation of atoms having a long shape. According to Sthiramati shape and vijñapti exist as mere designations and are hence not objects of the faculty of seeing, which can only have a real entity as its object. What is perceived by the latter is colour having different shapes. Remarkably, Sthiramati adds that according to ultimate reality (paramärthatah) varm — in the same way as samsthäna is not the object of the faculty of seeing. He argues that this is due to the fact that mind (vijñäna) does not have any outer objects, because the existence of (colour) atoms, like that of shape, is not possible from the viewpoint of that level. This remark by Sthiramati is one of the very few indications found in the PSkV pointing to the idealist doctrine of "mind-only" (vijñaptimatratä).

The classifications of sound (šabda), the object of the faculty of hearing, differ in the PSkV, the Abhidharmasamuccaya, and the Abhidharmakošabhä$ya. In this context it becomes obvious that Sthiramati in some cases refutes the traditions of the Abhidharmasamuccaya and of the Abhidharmakošabhäya and develops his own theories. All three texts agree on two kinds of sound:

1. the sound caused by the basic elements (mahäbhütahetuka) that are appropriated (upätta)

2. the sound caused by the basic elements that are not appropriated (anupätta)

The first kind of sound is identified as the sound of the voice (väc) in the PSkV, the Abhidharmakošabhä§ya, and the Abhidharmasamuccayabhäya. The PSkV and the Abhidharmakošabhäya

additionally mention the sound of the (clapping) hand (hasta). The sound that is not appropriated is, according to the PSkV and the Abhidharmakošabhä$ya, the sound of the wind (väyu), of the trees (vanaspati), or of the river (nadï). The categorization of the remaining types of sound appears to have been controversial. The Abhidharmasamuccaya and the PSkV distinguish a third category of sound: the sound that is both, upätta and anupätta. As an example for this kind of sound the texts mention the sound of a hand and a drum (mrdañga), i.e., a hand hitting a drum. Remarkably, this class of sound is rejected in the Abhidharmakošabhãya as a separate category. There it is stated that "others" (apare) say that a sound can be appropriated and not appropriated at the same time, but this is not accepted, as it is not admitted that one atom is based on two tetrads of the basic elements (i.e., the four basic elements of the hand and the four basic elements of the drum). This rejection of the sound that is both, appropriated and not appropriated, in the Abhidharmakošabhäya is remarkable insofar as the author of the Abhidharmakošabhäyya is generally accepted to be identical to that of the PSk. What is more, two other kinds of sound mentioned in the Abhidharmakošabhã$ya are absent in the PSk. In AKBh 6,24 the class of sounds of living beings (sattva) and, finally, the sounds not belonging to living beings are listed additionally. The first of these two classes refers to the "making known" of speech (vägvijñapti), the second is described as including all other kinds of sound. The Abhidharmako§abhä$ya adds that all four of the mentioned sounds can be pleasant (manojña) or unpleasant (amanojña), which makes a total of eight different categories of sounds. In contrast, the Abhidharmasamuccaya lists five additional classes of sound besides the two categories mentioned above. They include sounds known in the world (lokaprasiddha), i.e., common talk (laukikabhä§ä), sounds produced by the siddhas

PAÑCASKANDHAKA VIBHÄ$Ä

(siddhopanïta), fabricated (parikalpita) sounds, sounds belonging to the common practice of the Äryas (äryavyävahärika) or to the common practice of the non-Äryas (anäryavyävahärika). All these sounds can be not only pleasant or unpleasant, as indicated in the explanation of the Abhidharmakošabhäsya mentioned above, but also neutral.

The sections on odour and taste consist of only three lines in the PSkV. Nevertheless they are noteworthy as they differ from the respective passages in the Abhidharmasamuccaya and the Abhidharmakošabhäsya. The latter mentions four types of odour: good (su-) and bad odour (durgandha) which can both be either constant

(sama) or inconstant (visama). It is added, however, that in the šästra (i.e., in the Prakarana) three kinds of odour are taught: good, bad, and neutral (samagandha). An almost identical statement is made by Vasubandhu in the PSk, the "neutral" odour being indicated with the phrase "other [odours]" Sthiramati does not mention this triad in his commentary, but instead refers to another definition, which describes the odour as natural (sahaja), like the smell of sandal wood (candana), as arising from contact (sämyogika), like the smell of incense (dhüpavarti), and as arising from change (parinämika), like the smell of ripe mango fruits (pakvämraphala). All six types of odour are listed in the Abhidharmasamuccaya, and in its commentary the last three are with the same examples of sandal wood, incense, and ripe fruits. That means that the triad of sahaja, sämyogika, and päripämika was probably unknown to the author of the Abhidharmakošabhäya and also to the author of the PSk. In his commentary Sthiramati, possibly following the tradition of the Abhi-

dharmasamuccaya, decided to go beyond Vasubandhu's definition and include the triad in his own explanation. In the case of taste, Sthiramati (in accord with Vasubandhu's root text) follows the classification of the Abhidharmakošabhäya, where six types are listed: sweet (madhura), sour (amla), salty (lavaaa), pungent (katuka), bitter (tikta), and astringent (kayäya). He does not mention that taste, according to the Abhidharmasamuccaya, can also be divided into the classes pleasant (manojña), unpleasant (amanojña), and neutral or, in analogy to the categories of odour, into sahaja, sämyogika, and pärinämika.

When analysing the nature of the tangible, Sthiramati again is closer to the Abhidharmakošabhäyya than to the definition of the Abhidharmasamuccaya, parts of which he even refutes explicitly. In AKBh 7,9f. eleven entities are mentioned as being tangible: the four mahäbhütas, softness (šlakuatva), hardness (karkašatva), heaviness (gurutva), lightness (laghutva), cold (šïta), hunger (jighatsä), and thirst (pipäsä). The explanation given in the PSk is more differentiated because, in contrast with the Abhidharmakošabhäya, rapa is divided into the matter of the four mahäbhütas and the matter dependent (upädäyarüpa) on them. As the tangible is explained in the context of upädäyarüpa, the mention of the four mahäbhütas as constituents of this category would contradict the classification (of mahäbhüta versus upädayarüpa) made earlier. Therefore it is said in PSk 2a1 that only a part of the tangible (spra§tavyaikadeša) is explained in this context. The Abhidharmasamuccaya does not mention the mahäbhütas under the topic of the tangible and explains that the latter consists of upädäyarüpa including in addition to the seven entities softness etc. fifteen other categories like strength (bala), weakness (daurbalya), or fainting (mürchä). Sthiramati explains that these additional categories are not listed in the PSk because they are already included in the remaining ones, like for example strength is included in hardness and heaviness and fainting in lightness.

The third constituent of matter dependent on the basic elements is, according to the PSk, the avijñapti. In the root text (PSk 2a1f.) avijñapti is explained as "invisible and penetrable matter arisen from vijñapti and meditative absorption" (vijñaptisamädhijan•z rüpam anidaršanam apratigham). A similar statement is given in AKBh 8,9f., where avijñapti is described as wholesome (kušala) or unwholesome (akušala) matter, having arisen from vijñapti or samädhi. Remarkably, the term avijñapti is not used in the AS. Instead, the term sämädänika is applied to this kind of karmic matter in the context of the definition of five classes of matter belonging to the dharmäyatana. The Abhidharmasamuccaya does not specify, however, what exactly the matter belonging to the sämädänika category is.

According to Sthiramati avijñapti that has arisen from vijñapti belongs to [the sphere of] sensual pleasure (kamäpta) and can be divided into four classes: (1) the restraint (samvara) of the code of precepts (prätimoksa), (2) the restraint of the Bodhisattva, (3) the non-restraint (asanvara), which includes practice harming others, and (4) neither restraint nor non-restraint, which is described as the appearance of partly beneficial and partly harmful [[[activities]]] (ekadešenänugrahopaghätapravrttth). This classification of avijñapti seems to be an extended adaptation of the analysis of it as found in the Abhidharmakošabhäya. There avijñapti is structured into the three classes samvara, asanwara, and naiva samvaro näsarnvarah. The class of the bodhisattvasarnvara is, as might be expected, missing in the Abhidharmakošabhäsya. It is notable in this context that the same structure of three classes is presented in the Abhidharma samuccaya; not, however, explicitly describing avijñapti, but the divisions of karma. What is surprising here, is the fact that the Abhidharmasamuccaya does not mention the restraint of the Bodhi-

sattva either.

The first category, sarpvara, is divided into three subclasses in the Abhidharmakošabhäsya and the Abhidharmasamuccaya. The first of these is prätimokyasanwara. The other two classes are the restraint of contemplation (dhyänasamvara) and the restraint of the uncontaminated (anäsravasamvara). These are the two categories that are described by Sthiramati as arising from meditative absorption (samädhija). The avijñapti that arises from meditative absorption belongs either to the material [[[sphere]]] (rüpäpta) or is the uncontaminated (anäsrava) avijñapti. If it belongs to the rüpadhätu then it originates from contaminated meditative absorption (säsravasamädhija) of the four contemplations (dhyäna), of [the stage] before attaining [the first dhyãna] (anägamya), and of the states between the [first two] dhyänas (dhyänäntara). 72 The avijñapti that is uncontaminated is produced by uncontaminated meditative absorption (anäsravasamãdhija). In the Abhidharmakošabhäya the analysis of avijñapti is very comprehensive and the discussion of opposing views with regard to the nature of avijñapti is rather complex and difficult to understand. In this context one of Vasubandhu's main concerns seems to be to oppose the Sarvästiväda theory of avijñapti as an existing entity (dravya) and a component of the category rüpa. Naturally, both these assumptions were rejected by the Sauträntikas, whose arguments Vasubandhu employs to support his position. Sthiramati refers only briefly to this discussion and seems (in accord with the PSk) 76 to accept the view that avijñapti belongs to the category of matter. He does not indicate, however, whether this also means that avijñapti is to be regarded as dravya. What seems to be more controversial to

Sthiramati is the question whether there are any other kinds of matter comparable to avijñapti. The manner in which Sthiramati discusses this problem shows that the boundary between material and mental factors was disputed among the authors of Abhidharmic texts like the PSkV, the Abhidharmasamuccaya, and the Abhidharmakošabhä§ya.

In the Sañgïtisütra it is stated that there are three kinds of rüpa: one that is visible (sanidaršana) and impenetrable (sapratigha), one that is invisible and impenetrable, and one that is invisible and penetrable. The sütra does not explain, however, what exactly is meant by these three classes. The passage is quoted in the Abhidharmakošabhaya (196,8f.), and Sthiramati seems to refer to it in the PSkV when he explains that all matter is threefold. He identifies visible and impenetrable matter with the object of the sense of sight, invisible and impenetrable matter with the five sense faculties as well as the objects of the other four senses (apart from the sense of sight), and finally the invisible and penetrable matter with the matter of dharmäyatana. According to the Abhidharmasamuccaya, the matter of dharmäyatana includes five entities: compressed matter (äbhi- sarnk$epika), matter of space (äbhyavakãšika), matter of commitment (sämädänika), imagined (parikalpita) matter, and matter produced by those with [[[meditative]]] power (vaibhütvika). The commentary explains matter that is äbhisamksepika as the matter of atoms (paramärpu), whereas abhyavakäšika is defined as referring to the matter of atoms being separated from other obstructing tangible [[[Wikipedia:matter|matter]]] (tadanyaprativärakaspra§tavyarahita). The term sämädänika is explained as an alternative expression for the matter of avijñapti, and parikalpita is defined as indicating matter of meditative images (pratibimba), as for instance the image of a skeleton (asthisamkalika). Vaibhütvika is explained as referring to objects of those who are absorbed in the eight liberations

KuMER

gocara). This last category probably includes the formless objects visualized during the practice of the eight liberations, such as the four colours. Sthiramati rejects four of these categories as belonging to rüpa and explains why he accepts only the avijñapti as invisible and penetrable matter. According to his argument, the first two categories, the äbhisamkyepika and the äbhyavakäšika, refer to matter of atoms and thus belong to the category of colour (which is part of the visible matter). The objects of the last two categories, the parikalpita and the vaibhütvika, are nothing other than mental images and are therefore to be regarded as being part of the mind. Thus in the case of invisible and penetrable matter Sthiramati explicitly rejects the teaching of five different entities given in the Abhidharmasamuccaya and follows the tradition of just one entity falling under this category of matter. This tradition is also evident in the Abhidharmakošabhäya where it is presented as the position of the Sarvästivädins, who state that there is no other invisible and penetrable rüpa than avijñapti. It is notable that in this context Vasubandhu mentions the view of some Yogäcäras who claim that an image that is perceived in contemplation is matter that is invisible and penetrable. This position seems to be accepted by the Sauträntikas (and probably also by Vasubandhu himself who tends to favour the Sauträntika view of avijñapti). However, the fivefold division of invisible and penetrable rüpa as explained in the Abhidharmasamuccaya and the PSkV is not mentioned in the Abhidharmakošabhäsya, possibly because it was unknown to Vasubandhu.

IV. Concluding Remarks

As one of the very few Abhidharmic works of the Yogäcära school extant in Sanskrit, Sthiramati's PSkV represents a very important source for the technical terminology of this tradition and contains valuable information on philosophical development and on the process of reshaping the canonical concept of the skandhas, äyatanas, and dhätus. In the context of the elaborate system of mind of the Yogäcära tradition, new theories such as those of älayavijñäna or kli$tamanas had to be integrated into the traditional structure. In a few instances Sthiramati even incorporates into the PSkV some aspects of the idealistic position of vijñaptimätratä, for instance when he mentions that colour, from the viewpoint of the highest reality, is not the object of the eye faculty, because vijñäna does not have any external objects. Apart from one or two remarks pointing in this idealistic direction, it is evident that the PSkV basically follows the more traditional teachings of the Abhidharmakošabhäya and the Abhidharmasamuccaya. Although Sthiramati does usually not explicitly mention his sources, it is obvious that his positions are closely related to those of the Abhidharmakošabhäya and the Abhidharmasamuccaya. With regard to the latter, however, it is notable that Sthiramati in some cases rejects views that very clearly have their origin in the Abhidharmasamuccaya. This is for example the case when Sthiramati explains the matter of dharmäyatana: He rejects its fivefold division as found in the Abhidharmasamuccaya, and instead accepts only the avijñapti in this category of matter. Remarkably, Sthiramati seems to follow a position here that in the Abhidharmakošabhä$ya is ascribed to the Sarvästivädins. According to the latter, only the avijñapti is to be regarded as invisible and penetrable matter. The contrary view, namely that meditative images are to be considered as this kind of matter, was obviously held by some (early) Yogäcäras (as represented by some portions of the Yogäcärabhümi) and the Sauträntikas.

On other occasions Sthiramati seems to compromise on the divergent teachings of the Abhidharmakošabhäya and the Abhidharmasamuccaya, and as a result develops his own interpretations. A similar procedure can already be observed in Vasubandhu's PSk, where for instance the number of unconditioned entities is established as four most probably a compromise between the three

KRAMER

unconditioned factors of the Abhidharmakošabhä§ya and the eight of the Abhidharmasamuccaya. Notable in this context is Vasubandhu's (and Sthiramati 's) treatment of the object of the faculty of hearing, the sound, which is defined in the PSk as being threefold. This division disagrees with the respective explanation of the Abhidharmakošabhäyya, in which the last of these three kinds, sound that simultaneously is upätta and anupätta, is rejected. Here Vasubandhu seems to follow the tradition of the Abhidharmasamuccaya, which accepts this kind of sound. But at the same time neither Vasubandhu nor Sthiramati mention the additional five classes of sound, like the sound "known in the world", listed in the Abhidharmasamuccaya in the same context. This tendency of developing interpretations which are neither in accordance with the Abhidharmakošabhäya nor with the Abhidharmasamuccaya is also visible when Sthiramati denies the positions held in the Abhidharmakošabhäya and the Abhidharmasamuccaya that there are more constituents of the visible (rüpa) than just varm and sarpsthäna, namely clouds, smoke, dust etc. Finally it may also be noted that Sthiramati occasionally supplements the definitions of the PSk, for example in the case of odour, where he lists three additional kinds that are absent in Vasubandhu's root text and also in the Abhidharmakošabhäya but are listed in the Abhidharmasamuccaya.

BIBLIOGRAPHY

PRIMARY SOURCES:

AKBh: Abhidharmako§abhäya (Vasubandhu), ed. P. Pradhan, Patna: K. P. Jayaswal Research Institute, 1967. AS: Abhidharmasamuccaya (Asañga), ed. P. Pradhan, Santiniketan: Visva-Bharati, 1950. AS*: Passages of AS retranslated into Sanskrit (from the Tibetan and Chinese translations of the AS) by Pradhan. ASBh: Abhidharmasamuccayabhäya (Buddhasimha or Jinaputra), ed. N. Tafia, Patna: K. P. Jayaswal Research Institute, 1976. AST: Tibetan translation of AS: P 5550. Prakarawa(pada): T 1542. PSk: Pañcaskandhaka (Vasubandhu) (manuscript, 7 fols., copy kept at the China Tibetology

Research Center [[[Wikipedia:Beijing|Beijing]]]).

PSkr: Tibetan translation of PSk: P 5560. PSkV: Pañcaskandhakavibhä§ä (Sthiramati) (manuscript, 73 fols., copy kept at the China Tibetology Research Center [[[Wikipedia:Beijing|Beijing]]]). PSkVT: Tibetan translation of PSkV: P 5567. Sañgrtisatra: See Stache-Rosen [1968]. YT: Tibetan translation of the Yogacarabhami: P 5536-5543.

SECONDARY SOURCES:

Frauwallner, Erich 1963 "Abhidharma-Studien: I. Pañcaskandhakam und Pañcavastukam," Wiener Zeitschrift fiir die Kunde Süd- und Ostasiens 7: 20-36. Karunadasa, Y. 1989 Buddhist Analysis ofMatter, Singapore: The Buddhist Research Society. Kramer, Jowita 2005 Kategorien der Wirklichkeit im frühen Yogäcära: Der Fünf-vastu-Abschnitt in der Vinišcayasamgrahapï der Yogäcarabhûmi, Wiesbaden: Dr. Ludwig Reichert Verlag.

In progess. Sthiramati's Pañcaskandhakavibhä§ä (Part I: Critical Edition; Part II: Diplomatic Edition), Beijing/Vienna: China Tibetology Publishing House — Austrian Academy of Sciences Press. Kritzer, Robert

2005 Vasubandhu and the Yogäcärabhami: Yogäcära Elements in the Abhidharmakošabhaya, Tokyo: The International Institute for Buddhist Studies. Schmithausen, Lambert

1973 "Spirituelle Praxis und philosophische Theorie im Buddhismus," Zeitschrift fiir Missionswissenschaft und Religionswissenschaft 3: 161-186. 1987 Älayavijñäna: On the Origin and the Early Development ofa Central Concept of Yogacära Philosophy, 2 vols., Tokyo: The International Institute for Buddhist Studies. Skilling, Peter

1980 "On the Five Aggregates of Attachment (III)," Linh-Son — Publication d'études bouddhologiques 13: 30-37. Stache-Rosen, Valentina 1968 Das Sañgrtisütra und sein Kommentar Sañgïtiparyäya, Berlin: Akademie-Verlag. Vetter, Tilmann

2000 The 'Khandha Passages' in the Vinayapitaka and the Four Main Nikäyas, Wien: Verlag der Österreichischen Akademie der Wissenschaften. Institute for Indology and Tibetology University of Munich Germany