Indian Abhidharma Literature in Tibet: A Study of the Vijñ"ana Section of Sthiramati’s Pañcaskandhakavibh"a s"a *

14 Indian Abhidharma Literature in Tibet:

A Study of the Vijñana Section of Sthiramati’s Pañcaskandhakavibh"a_s"a *

Jowita Kramer

INTRODUCTION

most of the indian Abhidharma texts available in Tibetan translation were produced in the ninth century by the Tibetan translators ska ba dpal brtsegs, ye shes sde, and cog ro klu’i rgyal mtshan in collaboration with the indian pan] dita] s Jinamitra, Śīlendrabodhi, Prajñāvarman, dānaśīla, and Viśuddhisimha. { They prepared translations of the Abhi dharmasamuccaya (P 5550) and its commentaries, the Abhidharmasamuccayabhās ya_ (P 5554) and the Abhidharmasamuccayavyākhyā (P 5555), of parts of the Prajñaptiśāstra (P 5587-5589), of the Abhidharmakośa(bhā_sya) (P 5590 and 5591) and its commentary, the Sphu_tārthā

Abhidharmakośavyākhyā (P 5593), of the Sārasamuccaya (P 5598), as well as of the Pañcaskandhaka (P 5560) and its three commentaries, namely, sthiramati’s Pañcaskandhakavibhā_sā (P 5567), gu]naprabha’s Pañcaskandhavivara]na (P 5568) and *P_rthivībandhu’s Pañcaskandhabhā_sya (P 5569).1 The latest translation of an indian Abhidharma treatise into Tibetan was probably produced by the Tibetan grammarian and translator chos skyong bzang po (1441-1527/28), also known by his sanskrit name dharmapālabhadra, who rendered into Tibetan sthiramati’s extensive commentary on the Abhidharmakośabhā_sya, the Abhidharmakośabhā_sya_tīkā Tattvārthā (P 5875).

despite the existence of a great variety of Abhidharma works in the Tibetan canon, only two of these texts made their way into the general curriculum of Tibetan monastic education and were transmitted in a continuous lineage: the Abhid harmakośa(bhā_sya) and the Abhidharmasamuccaya.2 remarkably, the teaching transmission of the Abhidharmasamuccaya is considered to have continued uninterruptedly from the ninth century, even through the “grey period” of Tibetan history (i.e. through post-imperial times from 842 until the end of the tenth century), while the Abhidharmakośabhāsya _ was reintroduced to Tibet by the indian pa]n]dita sm_rti in the late tenth century after a gap in the early transmission lineage.3

Besides the Abhidharmasamuccaya, the Abhidharma of the yogācāras is also represented in the Tibetan tradition by a collection of indigenous works that deal exclusively with the yogācāra concepts of the “notion of ‘i’” (kli_s_tamanas, nyon mongs pa can gyi yid) and the “store mind” (ālayavijñāna, kun gzhi rnam par shes pa). This literary tradition is based on indian yogācāra sources, as, for instance, the Mahāyānasa ^mgraha and the Tri^mśikā. its starting point goes back to the famous dge lugs pa scholar Tsong kha pa (1357-1419), who authored a work on this topic entitled Yid dang kun gzhi’i dka’ ba’i gnas rgya cher ’grel pa (P 6149). so far, seven commentaries on Tsong kha pa’s treatise have been located, most of them written by dge lugs pa scholars active in the major monastic centres of Tibet in the eighteenth and nineteenth centuries.4

Although the Pañcaskandhaka and its commentaries do not seem to have been major sources for Tibetan scholarship, the works do, nonetheless, mark important steps in the evolution of the Buddhist understanding of the person and are some of the very few texts that present this topic from the viewpoint of the yogācāra Abhidharma. Thus, they are of crucial importance for understanding the development of the Buddhist view of the person and the indian influence on related Tibetan literature. This particularly holds true for the study of the Buddhist notion of mind. i, therefore, provide below an analysis of the vijñāna section of the Pañcaskandhaka and its extensive sixth-century commentary, the Pañcaskandhakavibhā_sā. Prior to this investigation, i give a description of the sanskrit manuscript of the Pañcaskandhakavibhā_sā, which came to light in the collection of microfilm copies kept at the china Tibetology research center (cTrc) in Beijing a few years ago,5 and i also present some noteworthy scribal peculiarities.

By composing the Pañcaskandhaka, Vasubandhu produced a handy manual on the five constituents of a person (skandha) as understood from the viewpoint of the yogācāras. The conciseness of the work, covering seven manuscript folios,6 was motivated, according to the commentator sthiramati, by the intention to meet the needs of its potential readers: the householders who do not have enough time to read extensive treatises because of their various duties and the contemplating monks who should not be distracted by reading lengthy works.7

DESCRIPTION OF THE MANUSCRIPT General Description

The manuscript of the Pañcaskandhakavibhā_sā is complete, consisting of 73 palm-leaf folios with six lines of writing per folio.8 The pagination is written in the centre of the left margin of versos. remarkably, folio 8 appears twice in the manuscript. The reason for this duplication seems to be the scribe’s omission of a part of the folio. As he tried to include the missing section in the additional folio 8, the script of the second version of the folio is more condensed, the recto containing seven instead of six lines. Each leaf has been punched with a single hole on the left. The text is written in an upright and elegant Proto-Bengālī script, very similar to the script of the Vimalakīrtinirdeśa and the Jñānālokāla^mkāra, published by Taisho university.9 The very clear and confident script of the Pañcaskandhakavibhā_sā is different from the script of its root text, the Pañcaskandhaka, which is written in the hook-topped Nevārī script of the early twelfth century.10 The leaves of the Pañcaskandhakavibhā_sā manuscript are extremely well-preserved, without showing any serious damage. Thus, the physical state of the manuscript is much better than the condition of the other two yogācāra works kept (as facsimiles) at the cTrc: the Pañcaskandhaka and the Abhidharmakośabhā_sya_tīkā Tattvārtha, which are both incomplete and illegible in several passages.

Scribal Peculiarities

in the following, i describe some of the scribal and palaeographic characteristics of the manuscript, including various ways of erasing wrong aksara_ s, corrections in the margin, markings of the da]n_da, and peculiarities in the writing of certain ak_saras. There are three possibilities of how the space of a deleted ak_sara may appear in the manuscript. in most cases, the space was simply left blank after the ak_sara had been erased. in these instances, parts of the ak_sara may still be visible: . if there is no visible remainder of the deleted ak_sara – as in the following case: – it is difficult to decide whether the empty space indicates an erased ak_sara or a gap in the text inserted for another reason. The second possibility to indicate a deleted ak_sara is to mark it with one or two small strokes: . Thirdly, the place of the erased ak_sara can be indicated with a line-filling sign, of which there are three variants:11

it is unclear why three different signs are employed and why the scribe or corrector did not use the line-filling signs regularly. compared to the mere erasure of aksara_ s, the usage of the line-filling signs seems more efficient insofar as they make it very obvious to the reader that the ak_sara has been removed intentionally and that no ak_sara is missing in the manuscript. While the most common of all these possibilities of deletion is the erased ak_sara with some visible parts, the application of two strokes above the deleted ak_sara is the least frequent alternative. The line-filling signs seem to be employed in two functions: to fill the gap of an erased ak_sara or to fill a space left empty by the scribe, presumably because he was unsure about the reading of a passage. This difference is obvious because, occasionally, remains of the erased ak_saras are visible under the line-filling signs, while, in other cases, the latter seem to be written on an entirely blank surface.

Another striking feature of the Pañcaskandhakavibhā_sā manuscript exists in corrections to the text written in a different hand in the lower and upper margins by a proof-reader. They are placed above or below the section they refer to and are often followed by a numeral indicating the emended line. in addition, an upward- and/or downward-pointing kākapada occasionally marks the place where the corrected aksara _ has to be inserted. While the erased aksara_ s and line-filling signs might have been corrected by the scribe of the text himself, the emendations in the margins, written in a script different from that of the main

text, are most likely to be later additions, possibly added not only by proof-readers but also by later readers of the manuscript. Thus, the corrections in the margins and the various erasures mentioned above presumably reflect several layers of emendation. of interest is, moreover, the marking of the da]n_da, which very often is reduced to the half of a stroke or even to a mere dot. What makes the treatment of the da]n_da even more difficult for the editor is the fact that some of the strokes have obviously been added by a later corrector or reader to indicate phrase- or word-endings, and not to separate sentences. occasionally, the da]n_da is also employed in order to avoid the application of (difficult) sandhis, as for instance in folio 39a2: sa puna|h | rāgānukūla|h

(with sandhi: sa punā rāgānukūla|h). The heterogeneous employment of the da]n_da suggests – similar to the inconsistent correction of ak_saras – different stages of development of the manuscript. The unclear marking of the da]n_da makes it rather difficult to find an efficient system for reproducing the da]n_da in the diplomatic edition of the Pañcaskandhakavibhā|sā. is it reasonable to transcribe every dot and fine stroke appearing in the manuscript, even if it obviously is a later addition by a reader, possibly only included as a reading aid? other special characters appearing in the manuscript are the line-filling signs used at the end of the line or before a string-hole ( ) and the markers of passage endings ( ). remarkable are also the two ways of presenting an

anusvāra: one is a dot ( ) and the other a small circle ( ). Whenever this circle is written with a virāma ( ), it represents a stylised m with a virāma.12 moreover, there are two variant ways of indicating a visarga: one is formed in a figure eight ( ) and one is shaped with two circles ( ). Notable is also the way in which some of the vowels are marked in the manuscript. This is what an -ā usually looks like (here: kā): . But, occasionally, the vowel has the form of a hook and appears on top of the ak_sara (here: <ngā): . in the case of -e, the vowel sign can either have the following shape (here: ve): or look like this (here: tme): . The vowel signs written on top of the ak_saras might, again, be later additions to the manuscript. furthermore, there is a sign appearing several times in the manuscript which resembles an unfinished va or deleted dha but does not fit into the text (vijñaptir dviº): . This ak_sara has not been identified so far.

EXPLANATION OF VIJñANA

in the following, i present an overview of the vijñāna section of the Pañcaskan dhakavibhā|sā.13 Perception (vijñāna) is treated under three aspects in the Pañca skandhaka: “the actual perception” (prav|rttivijñāna), “the store mind” (ālayavijñāna) and “the notion of ‘i’” (kli | s | tamanas). Vasubandhu defines vijñāna in his Pañcas kandhaka as the “making known of objects” (ālambanavijñapti).14 “making known” means, according to sthiramati, grasping (graha|na), perceiving (avabodha) and seizing (pratipatti).15 Thus, the eye perception is, for instance, defined as “the making known of the visible on the basis of the

eye faculty.”16 As synonyms of vijñāna, the Pañcaskan dhaka men tions citta and manas.17 A similar explanation is found in the Abhidharmasamuccaya,18 where the citta aspect of vijñāna is defined as being the ālayavijñāna, whereas the manas aspect of vijñāna is paraphrased as kli|s_tamanas.19 it is notable that the Pañcaskandhaka initially gives an alternative interpretation of manas explaining that vijñāna is to be considered manas because it is the basis for the following moment of the mind (citta).20 The function of being the mental moment preceding present perception is also ascribed to manas in the Abhidharmakośabhā _

sya.21 however, Vasubandhu also mentions in the Pañcaskandhaka that the main function of manas in the context of the vijñānaskandha is to be the vijñāna that has the ālayavijñāna as its object and which is always associated with the contaminations “wrong attitude towards the self” (ātmamoha), “false view of the self” (ātmad_r_s_ti), “conceitful conception of the self” (ātmamāna), and “self-love” (ātmasneha).22 in his commentary on this passage, sthiramati specifies this kind of vijñāna as the “contaminated notion” (kli_s_ta^m mana]h) which continuously takes the ālayavijñāna in the form of the self (ātman) as its

object.23 sthiramati also points out that there is a difference between the manas which is the basis for mental perception (manovijñāna) and the manas which has the nature of conceit (manyanā).24 This ambivalent meaning of manas and the question of whether there is a direct relation in doctrinal development between the concepts of manodhātu and klis_ tamanas_ requires further investigation, and will be treated in a separate article which is currently under preparation.25 Vasubandhu continues his explanation of manas with the statement that the latter is of only one kind (ekajātīya).26 sthiramati comments that manas can only be

contaminated (kli_s_ta), as it is permanently connected with the four contaminations mentioned above. he adds that kli_s_tamanas cannot be beneficial (kuśala) or (exclusively) neutral (avyāk_rta).27 The association of kli _ s _ tamanas with contaminations has also been expressed differently, for example, in the Mahāyānasa^mgraha and in Tsong kha pa’s Yid dang kun gzhi’i dka’ ’grel where it is characterised as niv _ rtāvyāk _ rta, i.e. “obstructed [by contaminations but] neutral.”28 When stating that kli_s_tamanas cannot be neutral, sthiramati obviously refers to a quality of avyāk_rta other than the quality of the concept

nivrtāvyāk_ rta_ . The latter indicates the karmic indefiniteness of a factor which may, at the same time, be spiritually bad. Thus, possessing kli_s_tamanas does not necessarily result in karmic disadvantages but it hinders one’s path to liberation. in contrast, the term avyāk_rta, as used in sthiramati’s commentary, indicates the complete neutrality of a factor in the sense of anivrtāvyāk_ rta _ (a quality which is ascribed to the ālayavijñāna). The explanation of kli_s_tamanas ends in the Pañcaskandhaka(vibhā_sā) with the mentioning of states in which it is not active: arhatship (arhattva), the path of the noble ones (āryamārga) and the equipoise of cessation (nirodhasamāpatti).29

The citta aspect of vijñāna is also explained in a twofold way in the Pañcaskandhaka. Vasubandhu starts with an “etymological” interpretation, stating that the word citta is derived from citra, i.e. “[being of] various [kinds].”30 sthiramati explains in his commentary that actual perception (prav_rttivijñāna) can be of manifold kinds depending on its various objects.31 however, the main meaning of citta is, according to Vasubandhu, being the store mind (ālayavijñāna), filled with seeds of all conditioned factors (sa ^ mskāra).32 sthiramati describes the seeds as imprints (vāsanā), which are nourished by the repeated

occurrence of impulses (sa^mskāra).33 This means the more often a sa^mskāra is active, the fatter its vāsanā will get. The process beginning in the moment of perceiving an object and ending in the moment of creating an imprint in the ālayavijñāna is described as follows in the Pañcaskandhakavibhā_sā:34 After the actual perception (prav_rttivijñāna) has perceived an object, the investigative (parye_saka) mental perception (manovijñāna) arises. After that, the classifying (vyavasthāpaka) manovijñāna appears.35 And finally, the imagining (vikalpaka) manovijñāna arises. This is the moment when contamination (sa{mkleśa) and purification (vyavadāna) occur (depending on the object), and the moment when meritorious (punya| ), non-meritorious (apunya| ) or neutral (aniñjya)

intention (cetanā) is produced. After the intention is completed, it leaves (or “nourishes”) an imprint in the ālayavijñāna that is either an “imprint of maturation” (vipākavāsanā) and is the basis for the arising of the ālayavijñāna in future existences, or an “imprint of outflow” (ni _ syandavāsanā), based on which (future) sa ^ mskāras emerge from the ālayavijñāna.36

The longest part of the vijñāna section of the Pañcaskandhakavibhāsā _ consists of a detailed investigation of the ālayavijñāna, in which sthiramati points out its divergence from the prav_rttivijñānas and provides several arguments for the necessity of its existence. The Pañcaskandhaka mentions three qualities of the ālayavijñāna that are contrary to the characteristics of actual perception:37

1. The object of the ālayavijñāna and the mode [in which it is apprehended] is not clearly determined (aparicchinnālambanākāra). 2. The ālayavijñāna is of one kind (ekajātīya). 3. The continuity of the ālayavijñāna is not interrupted (santānānuv_rtti).

in his commentary, sthiramati explains that the objects of actual perception are clearly definable as the visible, the sound and so on. in contrast, the mode of apprehension and the objects of the ālayavijñāna are not clearly determined. The objects are also described as difficult to understand (duravadhāra) due to their subtlety.38 sthiramati identifies them as the external world (bhājana) and the appropriation (upādāna), i.e. the [subtle] matter of the [material] sensefaculties (indriyarūpa) together with its [gross] basis (sādhi_s_thāna), and the impression (vāsanā), which consists in the sticking to the imagined character [of reality] (parikalpitasvabhāvābhiniveśa).39

The second quality of the ālayavijñāna, which is ekajātīya, indicates, according to sthiramati, that the ālayavijñāna is always morally neutral (avyāk _ rta), whereas the actual perception might be classified as neutral, beneficial (kuśala) or contaminated (kli _ s _ ta). The moral neutrality of the ālayavijñāna is constituted by its being exclusively [the result of] maturation (vipāka) of previous karma, which can by no means have a future karmic effect itself.40 The last difference between the ālayavijñāna and the pravrttivijñāna_ s consists in the uninterrupted homogeneous continuity of the ālayavijñāna, on the one hand, and the constant alternation of actual perceptions, on the other. While the continuity of the ālayavijñāna is not interrupted from the moment of “linking up” (pratisandhi) a new existence (i.e. from the moment of conception) until the moment of death, the quality of actual perception changes in every moment.41 A beneficial perception can be immediately followed by a nonbeneficial vijñāna and an eye perception might be succeeded by a perception of smell or taste and so on.

The existence of ālayavijñāna can be proved, according to sthiramati, on the basis of the authoritative scriptures (āgama) and of reasoning (yukti). in order to provide evidence from the former, he quotes relevant passages from the Sa ^ mdhinirmocana-sūtra and the Abhidharma-sūtra,42 the latter being currently unavailable in sanskrit or in Tibetan and only accessible through citations in other works. The logical arguments that confirm the necessity of ālayavijñāna are indicated in PSk 16,11-17,4 and may be paraphrased in the following way:

1. Actual perception reappears after a person has risen from unconscious states, as, for instance, the equipoise of cessation (nirodhasamāpatti). 2. Actual perceptions have different modes [of apprehension] (prakāra) depending on different kinds of object conditions (ālambanapratyaya). 3. Actual perception reappears after it has been interrupted. 4. An individual arises and ceases in sa^msāra.

sthiramati comments extensively on the first of the four arguments, explain- ing that it is impossible for actual perception to reappear after a person’s return from unconscious states of mind without a basis having the nature of the ālayavijñāna. As the prav _ rttivijñānas are interrupted during these states, they are not existent in the moment of rising from the unconscious condition and cannot be the basis for the newly-beginning perception.43 in his commentary, sthiramati also refers to the positions of some opponents as, for instance, the

sarvāstivādins. The first wrong view he argues against is that of the existence of past and future entities (bhāva), which the sarvāstivādins explain by means of their activity (kāritra). According to this theory, the entities are existent in all three time periods, being different merely as regards their activity: in the present, they are active; in the past, they are no longer active; and, in the future, they are not yet active.44 This argument is important in the context of the ālayavijñāna in that the existence of the latter would not be necessary if the continuity of mind could be assured by the permanent existence of

past and future phenomena. sthiramati refutes this model of the sarvāstivādins by stating that it would have unacceptable consequences. he opens his argument by pointing out that the entities would be constantly existent, whereas the activity would be existent exclusively in the present. Therefore, the entities would be different from their activity. As only the activity would arise and disappear (and not the entities themselves), the skandhas (also being bhāvas) would not be impermanent and, therefore, would not be part of suffering.45 The teaching that the skandhas are impermanent and characterised by suffering would

be a wrong view (viparyāsa). The result of this assumption would be that contaminations (kleśa) could not be removed as it is impossible to eliminate them through a wrong view.46 What is more, because of the arising and disappearing of the activity one would have to consider the latter as conditioned and, therefore, to be included among the five skandhas. however, as it is not subsumed under these skandhas, the only way out of this conflict would be the unacceptable adoption of an additional

(sixth) skandha consisting of activity.47 sthiramati also argues against the view that activity could be identical to the entities. According to the opponent, activity is defined as “taking hold of the effect” (phalaparigraha), which means that one entity, by taking hold of another as its effect, is the cause of this entity. The opponent states that, therefore, the activity is not different from the entity.48 in his answer, sthiramati explains that this position would have the consequence that either the past and future entities would not exist in the same way as their activity, or that the latter would be existent in all three

periods of time.49 however, these consequences would contradict the sarvāstivāda view that the entities exist in all three time periods and that the time periods are determined by activity. sthiramati adds that entities and activity must be different as, otherwise, the effect would be present at the same time as the cause. if all entities would exist at all times, they could not be causes for each other, as the cause must exist before the result arises. This would have the consequence that only activities would be caused by other activities but not the entities themselves.50 To those who hold the view that activity and entities are neither identical nor different, sthiramati responds that this position would also have unacceptable consequences. Activity cannot be identical to

entities because it is not the essential nature (svabhāva) of a certain entity. it cannot be different either as it is not the svabhāva of any other entity.51 Thus, it must be stated that activity is not the svabhāva of anything and is, therefore, not existent in the same manner as a hare’s horn (śaśavi_sā]na). An activity which is characterised in such a way cannot possibly be the factor that determines the three periods of time.52 Another position rejected by sthiramati is the claim of some opponents that the body (rūpa), endowed with material sense faculties, and the mind (citta) provide the seeds for each other. in this case, the mind could reappear on the basis of the body when a person arises from unconscious meditative states. This

is compared to the way in which the material body of someone leaving the immaterial sphere (ārūpyadhātu) reappears on the basis of his mind.53 Thus, the assumption of a continuous mind like the ālayavijñāna would not be necessary. sthiramati’s objection against this position is that it leads to the unacceptable condition that each living being would have two mental continua: one arising from the body and another one emerging from the mind.54 Additionally, it would be impossible to explain the arising from unconscious states of someone belonging to the immaterial sphere, as he does not have a material body that could be the basis for the reappearance of his mind.55

Vasubandhu’s second proof of the existence of the ālayavijñāna is explained by sthiramati as follows: due to the different kinds of object conditions (ālambanapratyaya) the actual perceptions appear in various modes. Therefore, a preceding perception usually cannot be the seed of the following one, e.g. a beneficial perception cannot function as the seed of a non-beneficial perception.56 Thus, a “store” like the ālayavijñāna is needed, containing the seeds from which different perceptions emerge. These seeds, in turn, are produced by previous actual perceptions that leave imprints (vāsanā) in the ālayavijñāna.57 Vasubandhu’s third argument pointing at the reappearance of actual perception after it has been interrupted refers, according to sthiramati, to the arising of the mind after deep sleep (middha) or a swoon (mūrchā).58 This argument is similar to the first proof of the reappearance of perception after rising from unconscious meditative states.

The last argument for the existence of ālayavijñāna offered by Vasubandhu and commented on by sthiramati is related to the dependent arising (pratītyasamutpāda) within sa^msāra and to liberation. sthiramati explains that the progressing of sa^msāra and the escaping from it would be impossible if there was no ālayavijñāna. Progressing of sa^msāra, i.e. rebirth, could not happen because the third member of the twelve-membered pratītyasamutpāda, the “linking up vijñāna” (pratisandhivijñāna), could not arise without ālayavijñāna.59 According to sthiramati, this vijñāna can neither emerge from the sa^mskāras of the previous existence, as these are no longer existent,60 nor can it arise from the prav_rttivijñānas because – without the ālayavijñāna – there would be

no possibility for the prav_rttivijñānas to leave an imprint (vāsanā) anywhere that could give rise to another vijñāna. it is neither possible that they produce an imprint in themselves nor that they leave the vāsanā in past or future moments of perception, as these are no longer or not yet existent.61 Thus, the only appropriate way to explain the progressing of sa^msāra is, according to sthiramati, as follows: the sa^mskāras arise due to ignorance and the vijñāna that is conditioned by the sa^mskāras is the [ālaya]vijñāna which is impregnated by them. The “linking up” which is conditioned by this vijñāna, is the nāmarūpa (i.e. the “mind and matter” which constitute the fourth limb of the pratītyasamutpāda).62

Escaping from sa^msāra would not be possible without the ālayavijñāna because liberation can only take place after the contaminations (kleśa) have been removed. if there was no ālayavijñāna, the contaminations would have to be eliminated in the moment of their actual appearance. however, this assumption is unacceptable as the actual contaminations cannot be present at the same moment as the path, their antidote which eliminates them. sthiramati also rejects the possibility that the bījas of the kleśas could be removed by their antidotes without the ālayavijñāna because the bījas and the mental moments which counteract them cannot be existent in one single mind series at the same time.63 Thus, a multi-layered mind stream is needed which allows the parallel existence of seeds and their antidotes.

CONCLUSION

despite the fact that Vasubandhu’s concise treatise on the five constituents of the person (skandha), the Pañcaskandhaka, and its three extensive commentaries by sthiramati, gu]naprabha and *P_rthivībandhu never achieved the same relevance within the scholarly tradition of Tibet as the Abhidharmakośa(bhāsya) _ and the Abhidharmasamuccaya, their position within indo-Tibetan scholasticism should not be underestimated. They not only testify to the variety of Abhidharma literature that was transmitted to Tibet but they also represent important constituents in the history of the development of the Buddhist concept of mind, which led to the emergence of a very particular literary collection in the Tibetan tradition dealing with the specific functions of the ālayavijñāna (kun gzhi rnam par shes pa) and the kli_s_tamanas (nyon mongs pa can gyi yid).

The main focus of the present paper is to present, on the basis of descriptions in the Pañcaskandhaka and comments in the Pañcaskandhakavibhāsā_ , the most important characteristics of vijñāna, the fifth skandha, which comprises the three functions of being the actual perception (prav_rttivijñāna), the “store mind” (ālayavijñāna), and the “notion of ‘i’” (kli_s_tamanas). of particular interest in this context is the Pañcaskan dhakavibhā_sā’s explanation of the process beginning in the moment of actually perceiving an object until the impressing of an imprint (vāsanā) in the ālayavijñāna by intention (cetanā).

however, the greatest part of the vijñāna section of the Pañcaskand haka (vi-bhā_sā) is devoted to a detailed description of the ālayavijñāna, mentioning three characteristics which mark the differences between the latter and the actual perception, and explaining the four arguments which make the existence of a “store mind” besides the mental stream of continuously fluctuating sense perceptions necessary. These four proofs are of particular interest because they differ from the eight proofs presented in the Yogācārabhūmi and the Abhidharmasamuccayabhā_sya, as well as from the arguments provided in the Mahāyānasa^mgraha.

The klis_ tamanas _ is treated far less exhaustively in the Pañcaskan dhaka(vibhāsā) | than the ālayavijñāna. remarkably, neither Vasubandhu nor sthiramati considered proving the existence of this aspect of the mind necessary, as can be deduced from the fact that they did not provide any proofs for it comparable to those for the ālayavijñāna.64 Notable in connection with the definition of kli_s_tamanas is the Pañcaskandhaka’s twofold definition of manas as the basis for the following moment of mind, on the one hand, and as the notion of the ālayavijñāna as the self, on the other. This explanation indicates the multilayered nature of this term, which is common in yogācāra literature but, nonetheless, requires further investigation.

NOTES

• i would like to thank Jens-uwe hartmann, Kazuo Kano, ralf Kramer, helmut Krasser, lambert schmithausen and Peter skilling for helpful comments and corrections. i am also grateful for the support received from the german research foundation (dfg) between 2007 and 2011, which enabled me to complete this article.

1. for a detailed list of works translated by Jinamitra, Prajñāvarman, and other indian pa|n|ditas, see skilling, Mahāsūtras, pp. 148ff. According to skilling, Mahāsūtras, p. 120, Jinamitra was probably the most important indian figure in the introduction of Abhidharma in Tibet. The question of whether Jinamitra and the three Tibetan translators were active in the eighth century (under Khri srong lde btsan) or in the ninth century (under ral pa can) has been discussed in martin, “gray Traces,” p. 339. martin argues in support of the later date.

2. it should be noted, however, that the most important reference work for the study of Abhidharma among the Tibetans is to be found in the Mchims mdzod, a thirteenthcentury Tibetan Abhidharma commentary based on Vasubandhu’s Abhi dharma- kośabhā|sya, yaśomitra’s Sphu_tārthā, and on various yogācāra sources. A brief description of this text is found in odani, “study of the Abhidharmakośa in Tibet.”

3. see martin, “gray Traces,” p. 337. in the gsan yig of the fifth dalai lama, Ngag dbang blo bzang rgya mtsho (1617-82), several alternative transmission lineages for the Abhidharmakośabhā_sya are listed, including a lineage introduced to Tibet by Jinamitra and continuing uninterruptedly from the imperial period onwards, and lineages entering Tibet through sm_rti and through the Kashmiri pa|n|dita Śākyaśrībhadra. see Thob yig, vol. 1, pp. 47f. however, according to martin, “gray Traces,” p. 344, these were originally Abhidharmasamuccaya lineages which were later appropriated and renamed as Abhidharmakośa lineages. martin states that by the time of Ngag dbang blo bzang rgya mtsho, the Abhidharmakośa had replaced the Abhidharmasamuccaya in the monastic curriculum to such an extent that it appeared appropriate to replace the incomplete lineage of the Abhid harmakośa with the unbroken lineages of the Abhidharmasamuccaya.

4. see the Yid dang kun gzhi’i dka’ gnad rgya cher ’grel pa legs bshad ’bru ’grel gsal sgron by dge ’dun bstan pa dar rgyas (1493-1568; mentioned in the collection of the Tibetan Buddhist resource center [W12601]), the Rnal ’byor spyod pa pa’i lugs kyi yid dang kun gzhi’i rtsa ’grel gyi dka’ gnas gsal byed nyi zla zung ’jug by Blo bzang ’jam dbyangs smon lam (eighteenth century; published in The Collected Works of Ke’u tshang sprul sku Blo bzang ’jam dbyangs smon lam, vol. 1, dharamsala, 1984), the Yid dang kun gzhi’i dka’ gnas rnam par bshad pa mkhas pa’i ’jug ngogs by gung thang dkon mchog bstan pa’i sgron me (1762-1823; published in The Collected Works of Gu<n-tha<n dkon-mchog-bstanpa’i sgron-me, vol. 2, New delhi, 1972), the Yid dang kun gzhi’i dka’ gnad rgya cher ’grel

pa legs par bshad pa’i rgya mtsho de’i ’bru ’grel snying po gsal ba’i sgron me by Blo bzang thugs rje (1770-1835; mentioned in the database of the Tibetan Buddhist resource center [W14101]), the Rnam rig pa’i lugs kyi yid dang kun gzhi’i don cung zad bshad pa ngo mtshar gzugs brgya ’char ba’i me long by Blo bzang dam chos rgya mtsho (18651917; published in The Collected Works of Blo bzang dam chos rgya mtsho, vol. 1, New delhi, 1975), the Kun gzhi’i thal phreng yig cha’i rjes su ’jug pa by Blo bzang chos dbyings (nineteenth century; xylograph scanned by the Tibetan Buddhist resource center [W1cZ899]), and the Yid dang kun gzhi’i rtsa ba’i mchan ’grel gser gyi lde mig by Blo bzang ’jigs med (published in Rje gung thang blo gros rgya mtsho’i drang nges dka’ ’grel sogs, lanzhou, 2000).

5. The history of the sanskrit manuscripts in Tibet has been outlined in steinkellner, Tale of Leaves. According to steinkellner (p. 23), the originals of the copies preserved in the cTrc were photographed in lhasa in 1987. for an investigation of the samskāra^ section of the Pañcaskandhaka(vibhā|sā), see Kramer, “study of the Sa]mskāra section.” The Tibetan version of the Pañcaskandhakavibhā_sā has recently been translated into English in Engle, Inner Science of Buddhist Practice.

6. The critical and diplomatic editions of the sanskrit text of the Pañcaskandhaka, facsimiles of which are also kept at the cTrc, have recently been published by Xuezhu li and Ernst steinkellner. 7. PSkV 1b3f.: g_rhasthānā]m bahuk_rtyavyāp_rtatvād vistaragranthe|sv abhi yogāsambhava|h | manasikārābhiyuktānā]m ca pravrajitānām api vistarag ranthābhiyogo vik _sepāyaiveti.

8. folio 1a is blank except for the Tibetan title of the text in dbu med script: phung po lnga’i ’grel pa. 9. for a table of the script used in these two manuscripts, see study group on Buddhist sanskrit literature, ed., Introduction, pp. 93-112. 10. see li and steinkellner, Vasubandhu’s Pañcaskandhaka, p. xi. 11. The first two signs are also used in the manuscript of the Vimalakīrtinirdeśa and the Jñānālokāla]mkāra. see study group on Buddhist sanskrit literature, Introduction, p. 111.

12. see macdonald, “manuscript description,” p. xxii. 13. A german translation of the Tibetan version of this section is available in Pahlke, “Vijñāna-Abschnitt.”

14. PSk 16,7. 15. PSkV 48b4. 16. PSkV 48b5: cak_survijñāna]m cak_surindriyāśrayā rūpaprativijñapti]h. A parallel statement is made in Y 4,6, AS 19,18, and AKBh 11,6. 17. see also AKBh 61,20: citta]m mano ’tha vijñānam ekārtham. 18. AS 19, 12-17.

19. in AS 19,16f., however, manas is mentioned as the immediately preceding moment of mind and is distinguished from the kli_s|tamanas. 20. PSk 16,8: mana ] hsanniśrayatām. see also PSkV 49a3f.: | sa ` n ` nā ^ m cak _ surādi vijñān- ānā^m yad yan nirudhyate | tat tad anantarasya vijñānasyotpadyam ānasyāśra ya bhāvena vyavati|s_thata ity ata]h samanantara cittasanniśra yatām upādāya mana ity ucyate. 21. AKBh 11, 21: yad yat samanantaraniruddha^m vijñāna^m tan manodhātur ity ucyate. 22. PSk 17, 7f.: prādhānyena mana ālayavijñānālambana ^ m sadātmamohātmad | r _ s | tyātmamānātma snehādi sampra yukta^m vijñānam. it is noteworthy that other sources (e.g. AS 19,15) mention asmimāna in their lists of the four kleśas instead of ātmamāna and avidyā instead of ātmamoha. These divergences are discussed in schmithausen, Ālayavijñāna, p. 442, n. 943.

23. PSkV 59a5f.: kli_s|ta^m hi mana ālayavijñānam ātmatvena nityam ālambate. 24. PSkV 59b1: yat _sa|n|nā^m cak_surādivijñānakāyānā^m samanantara niruddha^m vijñāna^m mana ity uktam | tat | sa | s | thasya manovijñānasyāśraya pra siddhyartham | na tu manyanākāratvāt.

25. see also schmithausen, Ālayavijñāna, pp. 122ff. 26. PSk 17, 8. 27. PSkV 59b4f.: kli|s_tajātīyam | na kuśalam avyāk_rta^m vā nityam ātmamo hādibhiś catu<r>- bhi]h kleśai]h sa^mprayuktatvāt. 28. MSg 6, 20f. and Yid kun 33a1. 29. PSk 17, 9f. and PSkV 59b5f.

30. PSk 16, 8. 31. PSkV 49a1f.: tatra prav_rttivijñānasyālambanaprativijñaptisvarūpatvād ālamba nasya ca prativijñānam anekākāratvāc cak|surādi vijñāna`m citram utpadyate. 32. PSk 16, 8-10: prādhānyena punaś cittam ālayavijñānam | tathā hi tac cita`m sarvasa`mskārabījai]h. here, Vasubandhu obviously understands citta as being related to cita, i.e. “filled.”

33. PSkV 49b1: pauna]hpu]nyena (read: °punyena) sa{mskārā]nā`m samudācārād bījap u|s_tir vāsanety ucyate. 34. PSkV 49b4-6: tadanantara ^ m parye | saka ^ m manovijñānam | parye _ sakād anantar a ^ m vyavasthāpakam | eva^m vyavasthāpa kānantara^m vikalpaka^m mano vijñānam utpadyate | tatra ca visayāt sa| mkliśyate vyavadāyate vā^ | tadavas thaś ca punyāpu| nyāniñjyān sa| mskārānaś ^ (read: sa { mskārā ] mś) cetanātmakān abhisa ] m skaroti | te ’bhisa ] msk | rtā nirudhyamānā ālayavijñāne vipākavāsanā]m vā pu_s`nanti ni|syandavāsanā]m vā. 35. The “investigative” (parye_saka) and the “classifying” (referred to as niścita) manovijñāna is mentioned in the context of an analysis of perception in the Yogācārabhūmi. see Y 58, 18.

36. PSkV 49b6-50a2: tatra ni _ syandavāsanām āgamyālayavijñānāt pu ` nyāpu ` nyā niñjyā ` h sa]mskārā`h pravartante | vipākavāsanām āgamyālayavijñānān nikāyas abhāgāntare py (read: °āntare_sv) ālayavijñānam eva vipākātma ka]m nirvartate. 37. PSk 16, 9-11. 38. PSkV 50a6-50b3: prav | rttivijñāna ] m hi rūpaśabdādyālambanatvāt svasāmānya- lak_sa`nākāratvāc ca paricchinnālamba na]m paric chinnākāra]m ca | ālayav ijñāna]m punar aparicchinnālamba nākāram | na hy asyālambana]m paricchettu]m śakyate nākāra`h | . . . etac cālamba na]m sūk_sm<atv>āl loka pa`n`ditair api duravad hāram.

39. PSkV 50b1-3: ālayavijñāna ] m dvābhyām ālambanābhyā ] m pravartate | adhyātmam upādāya(read: upādāna)vijñaptito bahirdhā paricchinnākāra bhājana vijñaptitaś ca | tatrādhyātmam upādānam parikalpita svabhāvābhiniveśa vāsanā svā] (read: sā)dhis| thāna_ m ] cendriya rūpam. see also schmithausen, Ālayavijñāna, pp. 90ff. While Vasubandhu seems to regard the ālayavijñāna’s object and way of perceiving to be not definable at all when he uses the term aparicchinna, sthiramati apparently shifts its meaning. According to him the objects of the ālayavijñāna are not clearly determined in the sense of being difficult to understand due to their subtle nature. Additionally, sthiramati applies the characteristic of being aparicchinna to the ālayavijñāna’s way of perceiving the external world: it is perceived without being clearly discerned. for further remarks on the term aparicchinnākāra, see schmithausen, Ālayavijñāna, pp. 389ff., n. 634.

40. PSkV 50b4-6: pravrttivijñāna| m kuśalakli] s_ tāvyāk_ rtajātīyam_ | ālayavijñānam tv ekajātīyam] . . . ālayavijñāna]m sasam prayoga]m pūrvakarma sa]mskāra hetu ka tvād ekāntena vipāka evety āvyāk|rta(read: avyāk|rta)jātīyam eva. on the assumption that a vipāka cannot have a karmic effect itself, see also Kramer, Kategorien der Wirklichkeit, p. 130, n. 155. 41. PSkV 50b6f.: atra hy ālayavijñāna]m nikāyasabhāgāntare_su pratisandhim upādāya yāvac cyuti]m tāvat k_sa`naprabandhapravāhena vartate | na tv anta rāntarā vicchidyate prav|rtti vijñānavat. 42. see PSkV 51a5-51b1.

43. PSkV 51b3: nirodhāsa ]mjñisamāpattisamāpannasyāsañjñike ca prādurbhūte nirodhaviśe _ se _ su (read: niravaśe _ se _ su) prav | rtti vijñāne | su niruddhe | su punar vyutthāna kāla ālayavijñānam antare`na prav|rttiv ijñānotpattir na yujyate. 44. PSkV 51b4f.: anāgatavartamānātītatva ]m tu bhāvānā ]m kāritradvārakam | tathā hy akurvan(read: akurvat)kāritro ’nāgata `h | kurva<t>kāritro vartamāna`h | uparata kāritro ’tīta ity ucyate.

45. PSkV 51b5-52a1: eva]m tarhi kāritra]m bhāvād anyad ity abhyupagantavyam | atītānāgatāvasthayo`h kāritre ’saty api bhāvābhyupagamāt | eva]m ca kāri trasyaivotpādo vyayaś ca na bhāvānām abhūtvā bhāvād bhūtvā cābhāvāt | na ca rūpavedanāsa ] mjñā- samskāravijñānānām] | sarvadā svarūpādhyāsitatvāt | tataś cotpāda vyayābhāvān nirodhavan nānityā rūpādaya`h skandhā`h syu`h | tataś ca du`hkhā api na syu`h. 46. PSkV 52a2: eva]m ca skandhān anityato du`hkhataś ca cintayato viparyāsa eva syāt | na ca viparyāsāt kleśap rahā`na]m yuktam. 47. PSkV 52a3f.: kāritrasya cotpāda vyayasambhavāt sa]msk|rtatvam | rūpādi_su ca skandhe_su sa^mgrahābhāvāt |sa|s_tha]h skandho ’bhyupagantavya]h. 48. PSkV 52a4-6: kāritram nāma phalaparigraha] h` | . . . hetubhāvavyavasthānam bhāvā nā] m ] phalaparigraha ] h | hetubhāvena ca sa eva dharmo vyavati _ s | thata iti dharme bhya | h kāritram ananyat.

49. PSkV 52b1f.: eva]m cātītānāgatayor adhvano|h kāritravad dharmābhāva|h | . . . kāritrasya vā dharmasvarūpavad dharmād ananyatvāt traiyadhvakatvam abhyupag antavyam. 50. PSkV 52b4-6: dharmakāritrayoś cānanyatve hetuphalābhāvaprasanga< h] | tathā hi phalasya kāra|navat pūrvam eva parini_spannatvāt phalasya ki^m kurvat kāra`na]m bhavaty akurvad vā katha]m kāra`nam | phalam api tenākriyamā`na]m katha]m tasya kārya]m bhavati | kāritrā`nām eva ca parasparato hetuphalabhāvo na dharmā`nām. 51. PSkV 53a2f.: kāritra ] m dharmasya tattvānyatvenāvācyam ity anye | kimartha ] m tattvenāvācyam | dharmasvabhāva] m na bhavatīti | kimartham anyatvena nocyate | anyasva bhāvam api tan naivety anyatvenāpi nocyate. 52. PSkV 53a3f.: eva]m tarhi yan na tatsvabhāva]m nānyasvabhāva]m tac chaśavi_sā|navan ni|hsvabhāvatvān nāsty eveti | . . . kāritrāt sarvathāpy adhvavya vasthā na sidhyati.

53. PSkV 53a5-53b1: sendriya]m rūpa]m citta]m cānyonyabījakam ity ato nirodhāsa]mjñisamāpattibhyām asamjñikāc ca vyutti] s_ thata_ h sendriyarūpāc _ (read: sendriyād rūpāc) cittam utpadyate | yathārūpyadhātoś cyavamānasya cittād eva ciraniruddham api sendriya]m rūpam utpadyata iti. 54. PSkV 53b1f.: eva]m tarhy ekaikasya sattvasya dvau cittasantānau syātām | eka_h sendriyād rūpāt | aparaś cittāt. 55. PSkV 53b2: ārūpye_su ca nirodhasamāpannasya rūpābhāvād vyutthānābhāva_h. 56. PSkV 55a1f.: na ca prav_rttivijñānānā]m parasparabījatva]m yujyate | kuśalākuśalāv yāk_rtānām hīnamadhyapra] nītadhātūnā| m laukikalokottarā] nā| m sāsravānā sravā] nā| m vijñānānā] m ] paryāye|nā (read: paryāye|na) bhāvāt. 57. PSkV 55a4-6: taiś ca prakārāntarav_rttibhi|h prav_rttivijñānair vāsitād vipākav ijñānāt punar apy ālambanāpek _ sā | ni kuśalākuśalādiprakārāntarav _ rttīni prav _ rtti vijñānāny utpadyante.

58. PSkV 55b1f.: acittakamiddhamūrchādyavasthāsu cchinne prav_rttivijñānasantāne punar acittakamiddhamūrchāpagate (read: °āpagata) ālayavijñānam antare | na prav | rttivijñānotpattir na yujyate. 59. PSkV 55b5: tatrālayavijñānānabhyupagame sa]mskārapratyayavijñānābhāvāt sa]msāraprav_rttir na yujyate. 60. PSkV 55b6f.: tatra na pratisandhivijñāna ] m pūrvajanmopac itasa ] mskārap ratyayam | pūrvajanmopacitasa ] mskārā | nā ] m ciraniruddhatvāt | niruddhasya cāsattvāt | asataś ca pratyayābhāvāt. 61. PSkV 56b2f.: na hi vijñāna ]m vipākavāsanā ]m ni _syandavāsanā ]m vātmany ādhātu ]m samartham | ātmani kāritravirodhāt | nāpy anantare tasya tadānut panna tvād anutpannasya cāsattvāt | nāpy utpanne pūrvakasya (read: pūrvasya) tadā niruddhatvāt. 62. PSkV 56b1: samskārās tv avidyāpratyayā] h | | tadadhivāsitam ca vijñāna] m sa] mskārapratyayam ] | tatpratyaya]m pratisandhau ca nāmarūpam evety e_saiva nītir niravadyā.

63. PSkV 57a3-6: na cālayavijñānam antare|na tatprahā|na]m yujyate | sa]mmukhī bhūto vā kleśa|h prahīyeta bījāvāstho (read: °āvastho) vā | tatra sa]mmukhī bhūta|h prahīyata ity ani_s_tir eveya]m tadā tatprahā|namārgābhāvāt | bījāvastho ’pi naiva prahīyate | na hi pratipak_sāt tadānī]m ki]mcid anyad abhyupagamyate | yatra kleśabīja]m vyavasthita]m tatpratipak_se|na prahīyeta | atha pratipak_sacitta<m> eva kleśabījānu_saktam i_syate | na hi kleśabījānu _saktam eva tatpratipak_so yujyate. 64. it seems that the oldest source containing proofs for the existence of the kli_s_tamanas is the Mahāyānasa]mgraha in which six arguments are presented to show that their rejection would result in unacceptable consequences (see MSg 5, 24-6, 8). REFERENCES Primary Sources

AKBh Abhidharmakośabhāsya | (Vasubandhu), edited by P. Pradhan. Patna: K.P. Jayaswal research institute, 1967. AS Abhidharmasamuccaya (Asa<nga), edited by V.V. gokhale, “fragments from the Abhidharmasamuccaya of Asa<nga,” Journal of the Bombay Branch of the Royal Asiatic Society 23 (1947): 13-38. MSg Mahāyānasamgraha ^ (Asanga), edited by É. lamotte, < La Somme du Grand Véhicule d’Asa<nga, vol. 1. louvain: Bureaux du muséon, 1938. PSk Pañcaskandhaka (Vasubandhu), edited by X. li and E. steinkellner. Beijing/ Vienna: china Tibetology Publishing house/Austrian Academy of sciences Press, 2008.

PSkV Pañcaskandhakavibhā|sā (sthiramati), manuscript, 73 folios. Thob yig Thob yig ganggā’i chu rgyun (Ngag dbang blo bzang rgya mtsho). gangtok: sikkim research institute of Tibetology, 1991-95. Y Yogācārabhūmi, edited by V. Bhattacharya. calcutta: university of calcutta, 1957. Yid kun Yid dang kun gzhi’i dka’ ba’i gnas rgya cher ’grel pa (Tsong kha pa), P 6149. Secondary Sources

Engle, Artemus B. The Inner Science of Buddhist Practice. ithaca: snow lion Publications, 2009. Kramer, Jowita. Kategorien der Wirklichkeit im frühen Yogācāra: Der Fünf-vastu-Abschnitt in der Viniścayasa{mgraha|nī der Yogācārabhūmi. Wiesbaden: dr. ludwig reichert Verlag, 2005.

———. “A study of the Samskāra ^ section of Vasubandhu’s Pañcaskandhaka with reference to its commentary by sthiramati.” in The Yogācārabhūmi and Its Adaptation in India, edited by ulrich Timme Kragh, pp. 986-1035. cambridge: harvard university Press, 2013.

li, Xuezhu and Ernst steinkellner. Vasubandhu’s Pañcaskandhaka. Beijing/Vienna: china Tibetology Publishing house/Austrian Academy of sciences Press, 2008. macdonald, Anne. “manuscript description.” in Jinendrabuddhi’s Viśālāmalavatī Pramā|nasamuccaya|tīkā (Chapter 1): Part II, Diplomatic Edition, edited by Ernst steinkellner, helmut Krasser, and horst lasic, pp. ix-xxxvi. Beijing/Vienna: china Tibetology Publishing house/Austrian Academy of sciences Press, 2005. martin, dan. “gray Traces: Tracing the Tibetan Teaching Transmission of the Mngon pa kun btus (Abhidharmasamuccaya) through the Early Period of disunity.” in The Many Canons of Tibetan Buddhism (PIATS 2000: Tibetan Studies: Proceedings of the Ninth Seminar of the International Association for Tibetan Studies, Leiden 2000), edited by helmut Eimer and david germano, pp. 335-51. leiden: Brill, 2002.

odani, Nobuchiyo. “The study of the Abhidharmakośa in Tibet as seen through the mChims mdzod.” in Tibetan Studies: Proceedings of the Fifth Seminar of the International Association for Tibetan Studies, Narita 1989, edited by shōren ihara and Zuihō yamaguchi, pp. 193-96. Narita: Naritasan shinshoji, 1992. Pahlke, michael. “der Vijñāna-Abschnitt des Pañcaskandhaka.” m.A. thesis, hamburg university, 2000.

schmithausen, lambert. Ālayavijñāna: On the Origin and the Early Development of a Central Concept of Yogācāra Philosophy. Tokyo: The international institute for Buddhist studies, 1987. skilling, Peter. Mahāsūtras: Great Discourses of the Buddha, vol. ii, parts i and ii. oxford: The Pali Text society, 1997. steinkellner, Ernst. A Tale of Leaves: On Sanskrit Manuscripts in Tibet, Their Past and Their Future (= 2003 gonda lecture). Amsterdam: royal Netherlands Academy of Arts and sciences, 2004. study group on Buddhist sanskrit literature, ed. Introduction to Vimalakīrtinirdeśa and Jñānālokāla{mkāra. Tokyo: Taisho university Press, 2004.