Chil-she lun sung shih), is
preserved
in TD volume 41, number 1823.
Abhidharmakosabhasyam-Vol-1-Vasubandhu-Poussin-Pruden-1991
35.
] "Let it, then, be clearly understood," says Mrs.
Rhys Davids, "that our present knowledge of such philosophy as is revealed in the Buddhist Pali canon would be practically undiminished if the whole of the Abhidhammapitaka were non-existent .
.
.
The burden, then, of Abhidhamma is not any positive contribution to the philosophy of early Buddhism (?
), but analytic and logical and methodological elaboration of what is already given .
.
.
The chief methods of that (=Abhidhamma) training were: first, the definition and determination of all names or terms entering into the Buddhist scheme of culture; secondly, the enunciation of all doctrines, theoretical and practical, as formulas, with coordination of all such as were logically interrelated; and finally, practice in reducing all possible heterodox
7
positions to an absurdity . . . " Nevertheless, the word Abhidhamma takes on a
higher scope, which we can understand by example. The prohibition against drinking alcohol is a precept of the Vinaya; but to examine the transgression of alcohol as a transgression of nature or a transgression of disobedience is to bring pure theory to play upon the Vinaya, to "refined the Vinaya, and this is what is called Abhivinaya. In the same way, the Abhidharma did not remain a stranger to scientific research and philosophy; it concerns itself with questions whose relationship with the Dharma properly so-called are quite loose. This tendency is very much accentuated in the latest of the Paji Abhidhamma treatises, the
8
Kathdvatthu, which tradition dates from the Council of Asoka. This work is an
account of heresies, and fixes their positions very clearly with respect to a mass of purely speculative points: in this work one can verify the long work of exegesis of which the Sutra had been the object.
The Paji Abhidhamma does not form part of the ancient patrimony common
to all the sects--which is not to say that it is Singhalese! Whereas all the
soundings carried out in the canonical literature of purely Indian Buddhism
reveals to us some Vinayas and sutras that have developed out of the Pali
literature, or which have a close connection with this literature, no one has yet
9 discovered the presence of any "prototypes" of the Pali Abhidhammas.
In any case, according to the tradition itself, the Kathavatthu belongs in its own right to a certain philosophic school, the Vibhajjavadins, "the followers of
distinction. "
10 11 To the old question, discussed in the sutras, "Does all exist? ",
these philosophers answered by distinguishing (vibhajya): "The present, and the
past which has not yet brought forth its result exist; the future and the past which
u
have brought forth their result do not exist.
To this school there is opposed--from ancient times, we may believe--the
school of "all exists," Sarvastivada, (Sarvastivadinas, Sabbatthivadino). This
? school--which also formed a sect, which had a special Vinaya and its own canon, 13
and which was Sanskritized --"carved out" the Dharma. In addition to "casuists," vinayadharas, they had "philosophers," dbhidhdrmikas. [Their Devasarman, the proponent of the existence of the past and the future, was opposed to Mu-lien or
Moggaliputta. ] A long work, with regard to which we are little informed,
14
led to
the redaction of numerous works among which are the seven books of the
Abhidharma, Treatises (sdstra) or Works (prakarana), the Jnanaprasthdna and its
six "feet" (pdda), the Dharmaskandha, etc. There were philosophies which came
15
out of this first level of wisdom literature.
and, towards the end of the first century of our era (Council of Kaniska), commentary was written on the Jnanaprasthdna: the Vibhdsd, a collective work which gives its name to all the masters who adopted it. The Vaibhasikas are the philosophers who refer to the Vibhdsd (Watters, i. 276). The center of the school appears to have been KaSmlr, even though there were Sarvastivadins outside of Kasmlr,--Bahirdesakas, "masters from foreign lands"; Pascattyas, "Westerners [relative to Kasmlr]"; Aparantakas, "masters from the western borders"--and some Kasmiris who were not Vaibhasikas.
The Sarvastivadins and the Vaibhasikas believed that the Abhidharmas were
the word of the Buddha. But there were masters who did not recognize the
authenticity of these books. When they were obliged to observe that there is no
"basket of the Abhidharma" outside of the Abhidharmas of the Sarvastivadins but
that each one of them knew that the word of the Buddha was embraced within
three "baskets," they answered that the Buddha taught the Abhidharma in the
Siltra itself--which is quite true. They recognized only the authority of the Sutra,
17
and took the name of Sautrantikas.
But we should not be mistaken with respect to their attitude. Even though
formally opposed to some of the theses of the Vibhdsd and of the Vaibhasikas, the Sautrantikas had a modern enough speculation and perhaps a Buddhology. They did not systematically combat their opponents, who were, without doubt, their predecessors. They admitted everything from the system of the Vaibhasikas which they had no formal reason to deny.
Such is, at least--to speak with greater prudence,--the attitude of our author,
18 V asubandhu.
His work, the Abhidharmakosa, a collection of approximately six hundred verses, describes itself as "a presentation of the Abhidharma as taught by the Vaibhasikas of Kasmlr. " This is not to say that Vasubandhu is a Vaibhasika; neither is he a Sarvastivadin. He has evident sympathies for the Sautrantikas, and utilizes the opinions of the "early masters"--namely "the Yogacarins, the chief
But the speculative work continued
Poussin 3
16
a
? 4 Introduction
among them being Asariga"--but without doubt, in his own mind, the system of
the Vaibhasikas is indispensable: the Vaibhasikas are "the School. " One does not
find anywhere else a body of doctrine as organized or as complete as theirs.
Nevertheless they are sometimes in error, and on important points too.
Vasubandhu completes his collection of technical verses, an impartial presentation
of the Vaibhasika system, with a prose commentary, the Abhidharmakosa-
bhdsyam, wherein his personal opinions, objections, and the opinions of diverse
schools and masters are found presented among numerous theses rejected by the
19
School. We know that Vasubandhu was, in his turn, combated and refuted by
orthodox Vaibhasikas.
But it matters little to us whether he is always right! The essential thing, for
us as for the masters who followed him, is that his book and his bhdsyam are truly a treasure (kosa).
###
From the point of view of dogmatics the Abhidharmakosa, with the Bhdsyam, is perhaps the most instructive book of early Buddhism (the Hinayana). It renders a great service in the study of canonical philosophy and in the study of scholasticism properly so-called.
It would be very wrong to say that we do not know the philosophy of canonical Buddhism: we know its essentials, its principle teachings, its major affiliations, and many of its details. But the history of this philosophy, its origins and development, is less clear: even though we can imagine that Buddhism, like the Buddha himself, took many steps at its birth, and these in all directions. But it is fair to say (and encouraging to repeat) that if the history of the canonical philosophy has not yet been done, the image that scholars such as Rhys Davids and Oldenberg have given of this philosophy either remains definitive or calls for but light retouching. We may believe, however, that we do not fully know any part, because we so imperfectly know the scholasticism which certainly enriched it and perhaps deformed it, but which certainly unfolded within it; which moreover should be, by its methods and its tendencies, completely parallel to the early speculation from whence the canonical philosophy itself arose. This philosophy is made up of the earlier strata of a speculation which continues within scholasticism proper, Pali as well as Sanskrit.
The impression of ignorance is very strong when we attempt to read early works such as the Dhammasangani or the Kathdvatthu; or when, with some rigor, we attempt to determine the sense of the sutras themselves, word for word
? (avayavdrtha). How many terms the exact significance of which escape us! It is easy and often correct to observe that these terms originally did not have a precise meaning; that the general orientation of Buddhist thought alone merits our interest; that, if we were to ignore precisely the four dhydnas and the four drupyasamdpattisdy vitarka and vicdra, rupa, the "fruits" and the "candidates for these fruits," we nevertheless have a sufficient idea of the major purport of and the methods leading to holiness within Buddhism; and that it is the candidates for these fruits who should preoccupy themselves with the details of the Eightfold Path rather than Western historians. Some think that scholasticism is not interesting; that, throughout Buddhist history, it remains alien to religion proper, as with the early doctrine. This is wrong: iti cen na sMravirodhatah, "If you think thus, no, for this is in contradiction with the Sutral" Buddhism was born complicated and verbose; its scholastic classifications are often pre-Buddhist; it is our good fortune to be able to examine them up close, in sources more ancient than Buddhaghosa; and the Abhidharmakosa bestows this good fortune upon us
in the measure in which we have the courage to be worthy of it.
An example of this is given by the Buddhists themselves. The Abhidharma-
kosa has had a great destiny: "This work . . . had an enormous influence. From
the time of its appearance, it became indespensable to all, friend and foe, we are
told; and there is reason to believe this, for the same fortune followed it
everywhere, first in China with Paramartha, and Hsiian-tsang and his disciples,
and then in Japan, where to this day specialized Buddhist studies begin with the
20 Kosas'astra. "
The author assures us that we will find in his book a correct summary of the
doctrine of the Vaibhasikas; but, however close may be his dependence on earlier
Abhidharma masters, we may believe that he improves upon what they have said.
When the Kos*a has been read, the earlier works of the Sarvastivadins, the
Abhidharmas and the Vibhdsa, undoubtedly lose part of their practical interest.
Though the Chinese have translated these works, the Tibetan Lotsavas did not
think it proper to put these works into Tibetan (with the sole exception of the
21
Prajnapti ), doubtless because the Abhidharmakosa, in accord with the resolution
of Vasubandhu, constitutes a veritable summa, embracing all problems--ontology, psychology, cosmology, discipline and the doctrine of aaion, the theory of results, mysticism and sanctity--and treating them with sobriety and in clear language, with all the method of which the Indians are capable. After Vasubandhu, the Northern Buddhists--whichever school they belonged to, and whether or not they adhered to the Mahayana--learned the elements of Buddhism from the Kosa. All schools, in fact, are in agreement with respect to a great number of
Poussm 5
? 6 Introduction
fundamental items, the same admitted by Pali orthodoxy, and the same, we may add, which are often subjacent to the sutras themselves. These items, which the Vaibhasikas have elucidated, are nowhere so wisely presented as in the Abhi- dharmakosa. This sufficiently explains the reputation of the author and the popularity of the book.
If Vasubandhu is an excellent professor of Buddhism, of Buddhism without
epithet of sect or school, he furthermore renders us a precious service by initiating
us into the systematic philosophy of these schools. He constructs before us the
spacious edifice of Vaibhasika dogma; he shows us its flaws; he explains what the
Sautrantika says, what the Vaibhasika answers, and what he himself thinks. Like
many philosophical treatises, and like the best of them, the Abhidharmakosa is a
creature of circumstances, written sub specie aeternitatis. We find in it many 22
proper names, and many allusions to contemporary debates. This is not a dull book.
We also find in it a great number of quotations which are shortened
of the earlier literature. Its quotations add to the numerous fragments of the Sanskrit canon which the sands of Turkestan have given us or which have been discovered under the modernist prose of the Divydvadana and the sutras of the Great Vehicle. These bear most often on texts of a doctrinal order, and we become clear with respect to the doctrinal, if not the historical, relationships of the canons.
***
For a long time the importance of the Abhidharma has been recognized by European scholars, initially by Burnouf. Let us see why the study of this work has been deferred for such a long period of time.
The work of Vasubandhu is made up of two distinct parts: the Abhidharma- kosa or the kdrikds, a collection of approximately six hundred verses; and their commentary or bhdsyam.
And of the vast exegetical literature that fills eight volumes of the Tibetan canon, the Nepalese scribes have preserved only a single document for us, a commentary on the Bhasya by Yasomitra, the Abhidharmakosavydkhyd, which bears the name of Sphutdrthd, "of clear meaning. "
This commentary by Yasomitra is not a complete commentary. It occasionally quotes the stanzas of Vasubandhu, and it elucidates such and such a passage of the Bhdsyam, indicating the passage in question by the first words of that passage,
elsewhere. Because of this, the Kosabhdsyam is a precious testament for the study 23
? following the general usage of commentators. 'The subject itself," says Burnouf, "is difficult to follow because of the form of the commentary, which detaches each word from the text, and develops it or argues with it in a gloss which ordinarily is very long. It is only very rarely possible to distinguish the text from among those commentaries in the midst of which it is lost. " Let us add that Yasomitra passes over in silence everything that appears easy to him or without interest, and he plunges the reader ex abrupto into discussions of items and "positions" which are not indicated. In the First Chapter, he explains nearly every word of the text. Elsewhere he applies himself only to the points with respect to which there is something important to say.
The commentary of Yasomitra is thus, as Burnouf says, "an inexhaustible mine of precious teachings" (Introduction, p. 447); we read thousands of interesting things in it; but it is, by itself, a very ineffective instrument for the study of the Abhidharmakosa.
This is why this work has been neglected for such a long time. Or, better, why, even though it solicited the attention of many seekers, no one has yet set his hand to work on it. A knowledge of Sanskrit is insufficient; one must join a knowledge of Tibetan and Chinese to this, for until recently it was solely in its Tibetan and Chinese versions that there existed, integrally, the book of Vasubandhu, Kdrikd and Bhdsyam.
ii. Bibliography of the Kosa.
1. Burnouf, Introduction, 34, 46, 447 (its importance), 563; Wassiliew, Buddhismus, 77,78,108,130,220; S. Levi, La science des religions et les religions de llnde (Iicole des Hautes-Etudes, Syllabus 1892), Hastings' Encyclopedia, 1. 20 (1908); Minayew, Recherches et MatSriaux, 1887, trans. 1894.
J. Takakusu, "On the Abhidharma Literature," JPTS, 1905.
Noel Pe'ri, "A propos de la date de Vasubandhu," BEPEO, 1911.
De La Vallee Poussin, Cosmologie Bouddhique, Troisieme chapitre de
I'Abhidharmakoca, kdrikd, bhdsya et vydkhyd, avec [uneintroduction et] une analyse de la Lokaprajnapti et de la Karanaprajndpti de Maudgalydyana, 1914-1919; Paul DemieVille, "Review of the Kosa i-ii," Bulletin, 1924, 463; 0. Rosenberg, Probleme der buddhistischen philosophie, 1924, trans, of the work published in Russian in 1918 (the appendix contains a rich bibliography of Abhidharma literature, Chinese sources and Japanese works); Th. Stcherbatsky, 1.
The Central Conception of Buddhism and the Meaning of the Word "Dharma,"
1923 (the first appendix is a translation of Kosa, v, p. 48-65 of the French
Poussin 7
? 8 Introduction
translation; the second is a list of the 75 dharmas with substantial notes); 2. an English translation of the Pudgalapratisedhaprakarana or the ninth chapter of the Kos'a, Ac de Petrograd, 1918.
Sogen Yamakami, Systems of Buddhistic Thought, Calcutta, 1912, Chap, iii, "Sarvastivadins. " Bibliography of contemporary Japanese articles and works in Pe*ri, Demieville, Rosenberg, and notably in Suisai Funabashi, Kusha Tetsugaku, Tokyo, 1906.
2. The Kosa and its commentaries, Sanskrit, Tibetan, and Chinese sources.
a. Abhidharmakosavydkhyd, Bibliotheca Buddhica, Sphutdrtha Abhidharma- kocavydkhyd, the work of Yacpmitra, first Kocasthdna, edited by Prof. S. Levi and Prof. Th. Stcherbatsky, 1st fasc, Petrograd, 1918; 2nd fasc by Wogihara, Stcherbatsky and Obermiller, (part of the second chapter), Leningrad, 1931.
Text of the third chapter, kdrikds and vydkhyd, in Bouddhisme, Cos- mologie . . . L. de La Vallee Poussin [with the collaboration of Dr. P. Cordier], Brussels, 1914-1919.
b. Tibetan translation of the Abhidharmakocakdrikdh and of the Abhi- dharmakocabhdsya of Vasubandhu, edited by Th. I. Stcherbatsky, 1st fasc. 1917,2nd fasc. 1930.
3. Tibetan sources, Palmyr Cordier, Catalogue de fonds tibetain de la Biblioteque Nationale, third part, Paris 1914, p. 394 and 499:
a. Abhidharmakosakadrikd and Bhdsya of Vasubandhu, Mdo 63, fol. 1-27, and fol. 28---Mdo 64, fol. 109.
b. Sutrdnurupd noma abhidharmakosavrttih of Vinltabhadra, 64, fol. 109-304.
c. Sphutdrtha ndrna abhidharmakosavydkhyd of Ya^omitra, 65 and 66. This is the commentary preserved in Sanskrit.
d. Laksandnusdrini ndma abhidharmakosattkd of Purnavardhana, a student of Sthiramati and master of Jinamitra and Silendrabodhi, 67 and 68.
e. Updyikd ndma abhidharmakosattkd of Samathadeva, 69 and 60, fol. 1-144. f. Marmapradipo ndma abhidharmakosavrttih of Dignaga, 70, fol. 144-286.
g. Laksandnusdrini ndma abhidharmakosattkd, an abridged recension of the
"Brhattika," above item d, 70, fol. 286-316.
h. Sdrasamuccayo ndma abhidharmavataratikd, anonymous, 70, fol. 315-393. i. Abhidharmdvatdraprakarana, anonymous, 70, fol. 393-417.
j. Tattvdrtho ndma abhidharmakosabhdsyatikd of Sthiramati, 129 and 130.
4. Abhidharmakofasdstra, of Vasubandhu, trans, by Paramartha in the period
564-567, Taisho volume 29, number 1559, p. 161-309; trans, by Hsiian-tsang, 651-654, Taisho volume 29, number 1558, p. 1-160.
The references in our translation are to the edition of Kyokuga Saeki, the
? Kando Abidatsuma Kusharon (Kyoto, 1891), the pages of which correspond to those of the Ming edition, a remarkable work which notably contains, in addition to interesting notes of the editor, copious extracts 1. from the two major Chinese commentators, 2. from the Vibhdsd, 3. from the commentary of Samghabhadra, and 4. from the work of K'uei-chi on the Trimsikd.
5. Among the Chinese commentaries on the Kosa:
a. Shen-t'ai, the author of a Shu: the Chil-she lun shu, originally in twenty Chinese volumes, today only volumes 1, 2, 4, 5, 6, 7, and 17 are extant; Manji Zoku-zokyo-1. 83. 3-4.
b. P'u-kuang, the author of the thirty-volume Chil-she lun Chi; TD 41, number 1821.
c. Fa-pao, the author of a thirty-volume Chil-she lun Shu; TD 41, number 1822.
Two other disciples of Hsiian-tsang, Huai-su and K'uei-chi, have written commentaries on the Kosa which are lost. P'u-kuang has also written a short treatise on the teachings of the Ko/a.
d. Yuan-hui wrote a thirty-volume Shu on the Kdrikds of the Kosa, a work with a preface written by Chia-ts'eng and dated before 727; this work, the Chil-she lun sung Shu (var.
Chil-she lun sung shih), is preserved in TD volume 41, number 1823. This Shu "was commented upon many times in China and very widely disseminated in Japan; it is from this intermediary text that Mahayanists in general draw their knowledge of the Kola. But from the point of view of Indology, it does not offer the same interest as the three preceding com- mentaries. "
Hsiian-tsang dictated his version of Samghabhadra to Yuan-yu. There are some fragments of a commentary written by him.
6. Gunamati and the Laksandnsdra.
Gunamati is known through his commentary on the Vydkhydyukti\ many fragments of this commentary are quoted in the Chos-'byun of Bu-ston, trans. Obermiller, 1931. It is mentioned four times by Yasomitra in his Abhidharmakofavydkhyd.
a. Introductory stanzas: Gunamati comments on the Kosa, as has Vasumitra; Yasomitra follows this commentary when it is correct.
b. "Gunamati and his disciple Vasumitra say that the word nutrias is declined in the fourth case. But when the word namas is not independent, we have the accusative. This is why this master (Vasubandhu), in the Vydkhydyukti, says, 'Saluting the Muni with my head' . . . " (Kosa, Vydkhyd, i p. 7).
c. Gunamati holds that the Kosa wrongly teaches that "Conditioned things,
Poussin 9
? 10 Introduction
with the exception of the Path, are sasrava (Kosa, i. 4b)," for all of the dharmas, without exception, can be taken as an object by the dsravas(Vydkhyd i, p. 13).
d. On the subject of the continuity of the mental series, "the master Gunamati, with his disciple the master Vasumitra, through affection for the doctrine of his own nikaya, instead of confining themselves to explaining the Kosa, refute it" (Kosa,iiL Ua-b, note).
N. Peri (Date, 41) recalls that Burnouf mentioned (Introduction, 566), according to Yasomitra, the commentary of Gunamati. He adds: "An author very rarely quoted. His Laksananusarasastra (Taisho 1641) forms part of the Canon, where it is classified among the HInayana works. It summarizes the ideas of the Kosa, and then presents his own opinions on several points. The Hsi-yu-chi, after having listed him among the celebrated monks of Nalanda (TD 51, p. 924a2), tells us that he left the monastery where he had been living in order to move to V alabhF(p. 936c2). "
Taisho 1641 is only an extract of the treatise of Gunamati, the chapter which examines the sixteen aspects of the truths (Kosa, vii. 13): "Do we have sixteen things or sixteen names? The masters of the Vibhasa say that sixteen names are posited because there are sixteen things. But the siltra-upadesa masters say that there are sixteen names, but only seven things; four things for the first Truth, one thing for each of the three others. In the beginning the Buddha promulgated the Upadesasutra. After the disappearance of the Buddha, Ananda, Katyayana, etc. , recited that which they had heard. In order to explain the meaning of the Sutra, as disciples do, they composed a sastra explaining the Sutra, which is thus called a siltra-upadesa. Then the Vibhasa extracted an upadesa from that which was to be found [in this upadesa']; since it only indirectly comes from the Sutra, it is not called a sutra-upadefa. "
Gunamati continues as in the Kosa, vii. 13a, "According to the first explana- tion, anitya, impermanent, because it arises dependent on causes (pratyayd- dhinatvdt)" And he comments, "Conditioned things, without force, do not arise in and of themselves . . . "
The first volume ends, "The thesis of Vasubandhu is similar to the meaning of the siitra-upadesa masters. "
The second begins, "The author says, 1 am now going to give the explanation of what I believe. Anitya, impermanent, because, having arising, it has extinction. Conditioned things, having arising, and extinction, are not permanent. Arising is existence . . . "
The treatise touches on diverse points of philosophy, the absence of dtman, etc. In this work we encounter some very interesting notes, for example (Taisho,
? page 168b9), "In the Hinayana, the pretas are superior to animals; in the Mahayana, the opposite. In fact, the pretas are enveloped in flames . . . "
It is curious that the title of the work of Gunamati, literally Laksandnusd- rasdstra, is exactly identical to that of the book attributed to Purnavardhana in the Tanjur. We have Gunamati, a teacher of Sthiramati, and Purnavarudhana, a
25
student of Sthiramati.
7. Sthiramati, a student of Gunamati, defended the Kosa against Samgha-
bhadra. "His commentary on the Kosa is mentioned many times by Shen-t'ai, P'u-kuang, and Fa-pao in their work on the same text. The precise manner in which they-quote it, in which they note and discuss its opinions, causes us to believe that Hsiian-tsang may have brought it to China, and perhaps they themselves had also read it" (N. Pe'ri, Date, 41). Sthiramati, the author of the Tsa-chi, is one of the great masters of the Vijnaptimitrata system.
There exists (Taisho 1561) a small treatise by Sthiramati (transcription and translation) entitled Kosatattvdrthatikd or Abhidharmakosasdstratattvdrathaptkd, which is doubtless an extract of a voluminous work of the same name and by the same author preserved in Tibetan (Cordier, 499).
We observed, at the beginning, the commentary on the seven points indicated in the introductory stanza of the Kosa.
On the wisdom of the Buddha, superior to that of the saints, the author quotes the Kalpanamanditika stanza (Huber, Sutrdlamkara), Kosa, i. l, vii. 30; and recalls the ignorance that Maudgalyayana had of the place where his mother had been reborn, Kosa, i. 1.
In order to demonstrate the thesis of the Kosa that sraddhendriya can be impure, ii. 9, the author quotes at length the sutra on the request of Brahma to the Buddha (setting into motion the Wheel of the Law), a sutra briefly indicated by V asubhandu.
The work ends with some remarks on the duration of life: The stanza says: "Among the Kurus life is always 1,000 years in length; half of this to the west and the east. In this continent, it is not set: atits end, some ten years; in the beginning, without measure" {Kosa, iii. 75-77), "There are, in fact, in this world, some beings who have extra meritorious actions and who make the resolution, 'May I have a long life! , without desiring more precisely, 'May I live one hundred years, ninety years, eighty years! ' Or rather some venerable persons, parents and friends, say, 'May you live long! ' without saying more precisely how long a time. If one makes similar vows, it is because the actions done by persons of this continent are associated with thoughts of desire. The Sutra says, 'Know, oh Bhiksus that the length of life was over 80,000 years under Vipasyin, 20,000 years under Kasyapa;
Ponssin 11
? 12 Introduction
the length of life is now 100 years; few will go beyond this, and many will have less. ' If the length of life is not set, why does the Blessed One express himself in this way? . . . " The treatise concludes with the well-known stanza: sucirna- brahmacarye'smin . . . (Kosa, vi. 60a).
8. Samghabhadra has written two works.
The first (TD 29, number 1562), the title of which is transcribed into Chinese as Abhidharmanydyanusdrasdstra--or perhaps better as Nydydnusdro ndma Abhidharmaidstram--is a commentary which reproduces without any changes the Kdrikds of the Abhidharmakosa. But this eighty-volume commentary criticizes the Kdrikds, which present the Vaibhasika doctrine by noting them with the word kila, which means "in the words of the School"; it refutes the Bhdsyam, the auto-commentary of Vasubandhu, when this work presents views opposed to those of the Vaibhasikas,and it corrects them when it attributes to the Vaibhasikas views which are not theirs.
The title of the second treatise (TD 29, number 1563) is not completely transcribed: Abhidharmasamaya-hsien-sdstra or Abhidharmasamaya-kuang-sastra. J. Takakusu proposes Abhidharmasamayapradipikds'dstra, which is not bad;
however pradipa, "lamp," is always teng, and we have for hsien the equivalents prakdsa and dyotana.
This is a forty-volume extract from the Nydydnusdra, from which all polemic
is excluded and which is thus a simple presentation of the system (samaya) of the
Abhidharma. It differs from the Nydydnusdra by the presence of a rather long
introduction, in seven stanzas and prose, and also by the manner in which it treats
the Kdrikds of Vasubandhu: these Kdrikds are either omitted (ii. 2-3) or corrected
( i l l , 14) when they express false doctrines or when they cast suspicion on true
26 doctrines by the addition of the word kila.
Samghabhadra is an innovator, and K'uei-chi distinguishes the earlier and the later Sarvastivadins, Siddhi, 45 (theory of atoms), 65 (laksanas of "conditioned things"), 71 (the viprayukta called ho-ho)y 147 (vedand? ), and 311 (divergent Sarvastivadins, on adhimoksa).
***
(Additions to the Bibliography, by Hubert Durt. )
The following titles are editions of texts and works related to the Abhi- dharmakosabhasyam, which have appeared in print since the first appearance of de La Valle'e Poussin's French translation (1923-1931).
? Sanskrit:
Gokhale, V. V. , The Text of the Abhidharmakosakdrikd of VasubandhuJournal
of the Bombay Branch, Royal Asiatic Society, n. s. , vol. 22,1946, p. 73-102. Pradhan, P. , Abhidharm-Koshabha/ya of Vasubandhu, Tibetan Sanskrit Works
Series, vol. vrn, K. P. Jayaswal Research Institute, Patna, 1967; 2nd edition, revised with introduction and indices, by Dr. Aruna Haldar, 1975.
Shastri, Swami Dwarikadas, Abhidharmakosa & Bhdsya of Acharya Vasu- bandhu with Sphutdrthd Commentary of Acdrya Yasomitra, Part I (I and II Kosasthdna), critically edited, Bauddha Bharati, Varanasi, 1970.
Wogihara, Unrai, Sphutdrthd Abhidharmakosavyakhya, 2 vols. , Tokyo 1933-1936; photomechanical reprint edition, Tokyo, 1971.
Tibetan:
Otani University, The Tibetan Tripitaka, Peking Edition, vol. 115 (number
5590), to Vol. 119 (number 5597), Suzuki Research Foundation, Tokyo, 1962.
Qiinese:
Takakusu, J. K. Watanabe, The Tripitaka in Chinese, vol. 29 (numbers 1558 to
1563), The Taisho Issai-kyo kanko kwai, Tokyo 1926.
Otani University, Index to the Taisho Tripitaka, no. 16, Bidon-bu III (vol. 29), Research Association for the Terminology of the Taisho Tripitaka, Tokyo, 1962.
Funabashi, Suisai and Issai Funabashi, Kando Abidatsuma Kusharon Sakuin, Kyoto, 1956. This index is based on the Chinese version of Kyokuga Saeki--the Qiinese version used by de La Valine Poussin--, the Kand6-bon Kusharon, in thirty volumesJCyoto, 1887.
Hi. The Date of Vasubandhu. The Former Vasubandhu.
We shall not undertake here a bibliography of Vasubandhu. But his treatise,
the Pratityasamutpddavydkhyd (Cordier, iii. 365), calls for the attention of the
reader of the Kosa. G. Tucci has published some fragments of this work (JRAS.
1930, 611-623) where the twelve links in the chain are explained in detail, with
numerous quotations from scriptures. G. Tucci also proposes to publish the
21
Trisvabhdvakarika and some parts of the commentary to the Madhydntavibhdga.
Concerning the "definition of pratyaksa by Vasubandhu," vdsubdndhava pratyaksalaksana, known through the Tdtaparyatikd, 99, and the Vddavidhi attributed to Vasubandhu, see the articles by G. Tucci, A. B. Keith, R Iyengar, JRAS. 1929, 473; Ind. Hist. Quarterly, 1928, 221; 1929, 81; Stcherbatski, Logic, ii. 161,382; G. Tucci, Maitreya [ndtha] et Asanga, 70-71, and finally Pramdna-
Poussin 13
? 14 Introduction
samuccaya, chap, i, by R. Iyengar, pp. 31-35. It appears that Dignaga denies the authorship of the Vddavidhi to Vasubandhu,in spite of universal opinion, and the Ttka quotes Kosa ii. 64, which contradicts the above-mentioned definition of pratyaksa. There are also numerous passages of the Vydkhyayukti in the Chos- 'byun of Bu-ston (above p. 16 ).
Wassiliew, Buddhismus, 235 (1860): "Life of Vasubandhu. "
Kern, Geschiedenis, trans. Huet, ii. 450.
S. Levi, JA. , 1890, 2. 252; Theatre indien, 1890, i. 165, ii. 35; "Donations
religieuses des rois de ValabhT' {Htes Etudes, vii, p. 97); "Date de Candragomin," BEFEO. ,1903,47;Sutrdkmkdra, trans,preface,2-3,1911.
Biihler, Alter der indischen Kunst-Poesie, p. 97,1890.
J. Takakusu, "Life of Vasubandhu," Voung-pao, 1904; "A Study of Para-
metria's Life of Vasubandhu and the date of Vasubandhu," JRAS. , 1905; "Sdmkhyakdrikd," BEFEO, 1904.
f
Wogihara, Asanga s Bodhisattvabhumi, 14, Strasbourg thesis, Leipzig, 1908.
Noel Peri, "A propos de la date of Vasubandhu," BEFEO. , 1911, 339-392.
Pathak, Bhandarkar, Indian Antiquare, 1911 - 1912 (V Smith, History, 3rd edition 328,4th edition, 346).
B. Shiiwo [Benkyo Shiio], Dr. Takakusu and Mr. Peri on the date of Vasubandhu (270-350), Tetsugaku Zasshi, Nov. -Dec. 1912.
Winternitz, ]eshichte, ii. 256 (1913), iii. 693 (1922).
H. . Ui, "On the Author of the Mahdydnasutrdlamkdra" Z. fur Ind. und Iranistik, vi. 1928,216-225.
7
positions to an absurdity . . . " Nevertheless, the word Abhidhamma takes on a
higher scope, which we can understand by example. The prohibition against drinking alcohol is a precept of the Vinaya; but to examine the transgression of alcohol as a transgression of nature or a transgression of disobedience is to bring pure theory to play upon the Vinaya, to "refined the Vinaya, and this is what is called Abhivinaya. In the same way, the Abhidharma did not remain a stranger to scientific research and philosophy; it concerns itself with questions whose relationship with the Dharma properly so-called are quite loose. This tendency is very much accentuated in the latest of the Paji Abhidhamma treatises, the
8
Kathdvatthu, which tradition dates from the Council of Asoka. This work is an
account of heresies, and fixes their positions very clearly with respect to a mass of purely speculative points: in this work one can verify the long work of exegesis of which the Sutra had been the object.
The Paji Abhidhamma does not form part of the ancient patrimony common
to all the sects--which is not to say that it is Singhalese! Whereas all the
soundings carried out in the canonical literature of purely Indian Buddhism
reveals to us some Vinayas and sutras that have developed out of the Pali
literature, or which have a close connection with this literature, no one has yet
9 discovered the presence of any "prototypes" of the Pali Abhidhammas.
In any case, according to the tradition itself, the Kathavatthu belongs in its own right to a certain philosophic school, the Vibhajjavadins, "the followers of
distinction. "
10 11 To the old question, discussed in the sutras, "Does all exist? ",
these philosophers answered by distinguishing (vibhajya): "The present, and the
past which has not yet brought forth its result exist; the future and the past which
u
have brought forth their result do not exist.
To this school there is opposed--from ancient times, we may believe--the
school of "all exists," Sarvastivada, (Sarvastivadinas, Sabbatthivadino). This
? school--which also formed a sect, which had a special Vinaya and its own canon, 13
and which was Sanskritized --"carved out" the Dharma. In addition to "casuists," vinayadharas, they had "philosophers," dbhidhdrmikas. [Their Devasarman, the proponent of the existence of the past and the future, was opposed to Mu-lien or
Moggaliputta. ] A long work, with regard to which we are little informed,
14
led to
the redaction of numerous works among which are the seven books of the
Abhidharma, Treatises (sdstra) or Works (prakarana), the Jnanaprasthdna and its
six "feet" (pdda), the Dharmaskandha, etc. There were philosophies which came
15
out of this first level of wisdom literature.
and, towards the end of the first century of our era (Council of Kaniska), commentary was written on the Jnanaprasthdna: the Vibhdsd, a collective work which gives its name to all the masters who adopted it. The Vaibhasikas are the philosophers who refer to the Vibhdsd (Watters, i. 276). The center of the school appears to have been KaSmlr, even though there were Sarvastivadins outside of Kasmlr,--Bahirdesakas, "masters from foreign lands"; Pascattyas, "Westerners [relative to Kasmlr]"; Aparantakas, "masters from the western borders"--and some Kasmiris who were not Vaibhasikas.
The Sarvastivadins and the Vaibhasikas believed that the Abhidharmas were
the word of the Buddha. But there were masters who did not recognize the
authenticity of these books. When they were obliged to observe that there is no
"basket of the Abhidharma" outside of the Abhidharmas of the Sarvastivadins but
that each one of them knew that the word of the Buddha was embraced within
three "baskets," they answered that the Buddha taught the Abhidharma in the
Siltra itself--which is quite true. They recognized only the authority of the Sutra,
17
and took the name of Sautrantikas.
But we should not be mistaken with respect to their attitude. Even though
formally opposed to some of the theses of the Vibhdsd and of the Vaibhasikas, the Sautrantikas had a modern enough speculation and perhaps a Buddhology. They did not systematically combat their opponents, who were, without doubt, their predecessors. They admitted everything from the system of the Vaibhasikas which they had no formal reason to deny.
Such is, at least--to speak with greater prudence,--the attitude of our author,
18 V asubandhu.
His work, the Abhidharmakosa, a collection of approximately six hundred verses, describes itself as "a presentation of the Abhidharma as taught by the Vaibhasikas of Kasmlr. " This is not to say that Vasubandhu is a Vaibhasika; neither is he a Sarvastivadin. He has evident sympathies for the Sautrantikas, and utilizes the opinions of the "early masters"--namely "the Yogacarins, the chief
But the speculative work continued
Poussin 3
16
a
? 4 Introduction
among them being Asariga"--but without doubt, in his own mind, the system of
the Vaibhasikas is indispensable: the Vaibhasikas are "the School. " One does not
find anywhere else a body of doctrine as organized or as complete as theirs.
Nevertheless they are sometimes in error, and on important points too.
Vasubandhu completes his collection of technical verses, an impartial presentation
of the Vaibhasika system, with a prose commentary, the Abhidharmakosa-
bhdsyam, wherein his personal opinions, objections, and the opinions of diverse
schools and masters are found presented among numerous theses rejected by the
19
School. We know that Vasubandhu was, in his turn, combated and refuted by
orthodox Vaibhasikas.
But it matters little to us whether he is always right! The essential thing, for
us as for the masters who followed him, is that his book and his bhdsyam are truly a treasure (kosa).
###
From the point of view of dogmatics the Abhidharmakosa, with the Bhdsyam, is perhaps the most instructive book of early Buddhism (the Hinayana). It renders a great service in the study of canonical philosophy and in the study of scholasticism properly so-called.
It would be very wrong to say that we do not know the philosophy of canonical Buddhism: we know its essentials, its principle teachings, its major affiliations, and many of its details. But the history of this philosophy, its origins and development, is less clear: even though we can imagine that Buddhism, like the Buddha himself, took many steps at its birth, and these in all directions. But it is fair to say (and encouraging to repeat) that if the history of the canonical philosophy has not yet been done, the image that scholars such as Rhys Davids and Oldenberg have given of this philosophy either remains definitive or calls for but light retouching. We may believe, however, that we do not fully know any part, because we so imperfectly know the scholasticism which certainly enriched it and perhaps deformed it, but which certainly unfolded within it; which moreover should be, by its methods and its tendencies, completely parallel to the early speculation from whence the canonical philosophy itself arose. This philosophy is made up of the earlier strata of a speculation which continues within scholasticism proper, Pali as well as Sanskrit.
The impression of ignorance is very strong when we attempt to read early works such as the Dhammasangani or the Kathdvatthu; or when, with some rigor, we attempt to determine the sense of the sutras themselves, word for word
? (avayavdrtha). How many terms the exact significance of which escape us! It is easy and often correct to observe that these terms originally did not have a precise meaning; that the general orientation of Buddhist thought alone merits our interest; that, if we were to ignore precisely the four dhydnas and the four drupyasamdpattisdy vitarka and vicdra, rupa, the "fruits" and the "candidates for these fruits," we nevertheless have a sufficient idea of the major purport of and the methods leading to holiness within Buddhism; and that it is the candidates for these fruits who should preoccupy themselves with the details of the Eightfold Path rather than Western historians. Some think that scholasticism is not interesting; that, throughout Buddhist history, it remains alien to religion proper, as with the early doctrine. This is wrong: iti cen na sMravirodhatah, "If you think thus, no, for this is in contradiction with the Sutral" Buddhism was born complicated and verbose; its scholastic classifications are often pre-Buddhist; it is our good fortune to be able to examine them up close, in sources more ancient than Buddhaghosa; and the Abhidharmakosa bestows this good fortune upon us
in the measure in which we have the courage to be worthy of it.
An example of this is given by the Buddhists themselves. The Abhidharma-
kosa has had a great destiny: "This work . . . had an enormous influence. From
the time of its appearance, it became indespensable to all, friend and foe, we are
told; and there is reason to believe this, for the same fortune followed it
everywhere, first in China with Paramartha, and Hsiian-tsang and his disciples,
and then in Japan, where to this day specialized Buddhist studies begin with the
20 Kosas'astra. "
The author assures us that we will find in his book a correct summary of the
doctrine of the Vaibhasikas; but, however close may be his dependence on earlier
Abhidharma masters, we may believe that he improves upon what they have said.
When the Kos*a has been read, the earlier works of the Sarvastivadins, the
Abhidharmas and the Vibhdsa, undoubtedly lose part of their practical interest.
Though the Chinese have translated these works, the Tibetan Lotsavas did not
think it proper to put these works into Tibetan (with the sole exception of the
21
Prajnapti ), doubtless because the Abhidharmakosa, in accord with the resolution
of Vasubandhu, constitutes a veritable summa, embracing all problems--ontology, psychology, cosmology, discipline and the doctrine of aaion, the theory of results, mysticism and sanctity--and treating them with sobriety and in clear language, with all the method of which the Indians are capable. After Vasubandhu, the Northern Buddhists--whichever school they belonged to, and whether or not they adhered to the Mahayana--learned the elements of Buddhism from the Kosa. All schools, in fact, are in agreement with respect to a great number of
Poussm 5
? 6 Introduction
fundamental items, the same admitted by Pali orthodoxy, and the same, we may add, which are often subjacent to the sutras themselves. These items, which the Vaibhasikas have elucidated, are nowhere so wisely presented as in the Abhi- dharmakosa. This sufficiently explains the reputation of the author and the popularity of the book.
If Vasubandhu is an excellent professor of Buddhism, of Buddhism without
epithet of sect or school, he furthermore renders us a precious service by initiating
us into the systematic philosophy of these schools. He constructs before us the
spacious edifice of Vaibhasika dogma; he shows us its flaws; he explains what the
Sautrantika says, what the Vaibhasika answers, and what he himself thinks. Like
many philosophical treatises, and like the best of them, the Abhidharmakosa is a
creature of circumstances, written sub specie aeternitatis. We find in it many 22
proper names, and many allusions to contemporary debates. This is not a dull book.
We also find in it a great number of quotations which are shortened
of the earlier literature. Its quotations add to the numerous fragments of the Sanskrit canon which the sands of Turkestan have given us or which have been discovered under the modernist prose of the Divydvadana and the sutras of the Great Vehicle. These bear most often on texts of a doctrinal order, and we become clear with respect to the doctrinal, if not the historical, relationships of the canons.
***
For a long time the importance of the Abhidharma has been recognized by European scholars, initially by Burnouf. Let us see why the study of this work has been deferred for such a long period of time.
The work of Vasubandhu is made up of two distinct parts: the Abhidharma- kosa or the kdrikds, a collection of approximately six hundred verses; and their commentary or bhdsyam.
And of the vast exegetical literature that fills eight volumes of the Tibetan canon, the Nepalese scribes have preserved only a single document for us, a commentary on the Bhasya by Yasomitra, the Abhidharmakosavydkhyd, which bears the name of Sphutdrthd, "of clear meaning. "
This commentary by Yasomitra is not a complete commentary. It occasionally quotes the stanzas of Vasubandhu, and it elucidates such and such a passage of the Bhdsyam, indicating the passage in question by the first words of that passage,
elsewhere. Because of this, the Kosabhdsyam is a precious testament for the study 23
? following the general usage of commentators. 'The subject itself," says Burnouf, "is difficult to follow because of the form of the commentary, which detaches each word from the text, and develops it or argues with it in a gloss which ordinarily is very long. It is only very rarely possible to distinguish the text from among those commentaries in the midst of which it is lost. " Let us add that Yasomitra passes over in silence everything that appears easy to him or without interest, and he plunges the reader ex abrupto into discussions of items and "positions" which are not indicated. In the First Chapter, he explains nearly every word of the text. Elsewhere he applies himself only to the points with respect to which there is something important to say.
The commentary of Yasomitra is thus, as Burnouf says, "an inexhaustible mine of precious teachings" (Introduction, p. 447); we read thousands of interesting things in it; but it is, by itself, a very ineffective instrument for the study of the Abhidharmakosa.
This is why this work has been neglected for such a long time. Or, better, why, even though it solicited the attention of many seekers, no one has yet set his hand to work on it. A knowledge of Sanskrit is insufficient; one must join a knowledge of Tibetan and Chinese to this, for until recently it was solely in its Tibetan and Chinese versions that there existed, integrally, the book of Vasubandhu, Kdrikd and Bhdsyam.
ii. Bibliography of the Kosa.
1. Burnouf, Introduction, 34, 46, 447 (its importance), 563; Wassiliew, Buddhismus, 77,78,108,130,220; S. Levi, La science des religions et les religions de llnde (Iicole des Hautes-Etudes, Syllabus 1892), Hastings' Encyclopedia, 1. 20 (1908); Minayew, Recherches et MatSriaux, 1887, trans. 1894.
J. Takakusu, "On the Abhidharma Literature," JPTS, 1905.
Noel Pe'ri, "A propos de la date de Vasubandhu," BEPEO, 1911.
De La Vallee Poussin, Cosmologie Bouddhique, Troisieme chapitre de
I'Abhidharmakoca, kdrikd, bhdsya et vydkhyd, avec [uneintroduction et] une analyse de la Lokaprajnapti et de la Karanaprajndpti de Maudgalydyana, 1914-1919; Paul DemieVille, "Review of the Kosa i-ii," Bulletin, 1924, 463; 0. Rosenberg, Probleme der buddhistischen philosophie, 1924, trans, of the work published in Russian in 1918 (the appendix contains a rich bibliography of Abhidharma literature, Chinese sources and Japanese works); Th. Stcherbatsky, 1.
The Central Conception of Buddhism and the Meaning of the Word "Dharma,"
1923 (the first appendix is a translation of Kosa, v, p. 48-65 of the French
Poussin 7
? 8 Introduction
translation; the second is a list of the 75 dharmas with substantial notes); 2. an English translation of the Pudgalapratisedhaprakarana or the ninth chapter of the Kos'a, Ac de Petrograd, 1918.
Sogen Yamakami, Systems of Buddhistic Thought, Calcutta, 1912, Chap, iii, "Sarvastivadins. " Bibliography of contemporary Japanese articles and works in Pe*ri, Demieville, Rosenberg, and notably in Suisai Funabashi, Kusha Tetsugaku, Tokyo, 1906.
2. The Kosa and its commentaries, Sanskrit, Tibetan, and Chinese sources.
a. Abhidharmakosavydkhyd, Bibliotheca Buddhica, Sphutdrtha Abhidharma- kocavydkhyd, the work of Yacpmitra, first Kocasthdna, edited by Prof. S. Levi and Prof. Th. Stcherbatsky, 1st fasc, Petrograd, 1918; 2nd fasc by Wogihara, Stcherbatsky and Obermiller, (part of the second chapter), Leningrad, 1931.
Text of the third chapter, kdrikds and vydkhyd, in Bouddhisme, Cos- mologie . . . L. de La Vallee Poussin [with the collaboration of Dr. P. Cordier], Brussels, 1914-1919.
b. Tibetan translation of the Abhidharmakocakdrikdh and of the Abhi- dharmakocabhdsya of Vasubandhu, edited by Th. I. Stcherbatsky, 1st fasc. 1917,2nd fasc. 1930.
3. Tibetan sources, Palmyr Cordier, Catalogue de fonds tibetain de la Biblioteque Nationale, third part, Paris 1914, p. 394 and 499:
a. Abhidharmakosakadrikd and Bhdsya of Vasubandhu, Mdo 63, fol. 1-27, and fol. 28---Mdo 64, fol. 109.
b. Sutrdnurupd noma abhidharmakosavrttih of Vinltabhadra, 64, fol. 109-304.
c. Sphutdrtha ndrna abhidharmakosavydkhyd of Ya^omitra, 65 and 66. This is the commentary preserved in Sanskrit.
d. Laksandnusdrini ndma abhidharmakosattkd of Purnavardhana, a student of Sthiramati and master of Jinamitra and Silendrabodhi, 67 and 68.
e. Updyikd ndma abhidharmakosattkd of Samathadeva, 69 and 60, fol. 1-144. f. Marmapradipo ndma abhidharmakosavrttih of Dignaga, 70, fol. 144-286.
g. Laksandnusdrini ndma abhidharmakosattkd, an abridged recension of the
"Brhattika," above item d, 70, fol. 286-316.
h. Sdrasamuccayo ndma abhidharmavataratikd, anonymous, 70, fol. 315-393. i. Abhidharmdvatdraprakarana, anonymous, 70, fol. 393-417.
j. Tattvdrtho ndma abhidharmakosabhdsyatikd of Sthiramati, 129 and 130.
4. Abhidharmakofasdstra, of Vasubandhu, trans, by Paramartha in the period
564-567, Taisho volume 29, number 1559, p. 161-309; trans, by Hsiian-tsang, 651-654, Taisho volume 29, number 1558, p. 1-160.
The references in our translation are to the edition of Kyokuga Saeki, the
? Kando Abidatsuma Kusharon (Kyoto, 1891), the pages of which correspond to those of the Ming edition, a remarkable work which notably contains, in addition to interesting notes of the editor, copious extracts 1. from the two major Chinese commentators, 2. from the Vibhdsd, 3. from the commentary of Samghabhadra, and 4. from the work of K'uei-chi on the Trimsikd.
5. Among the Chinese commentaries on the Kosa:
a. Shen-t'ai, the author of a Shu: the Chil-she lun shu, originally in twenty Chinese volumes, today only volumes 1, 2, 4, 5, 6, 7, and 17 are extant; Manji Zoku-zokyo-1. 83. 3-4.
b. P'u-kuang, the author of the thirty-volume Chil-she lun Chi; TD 41, number 1821.
c. Fa-pao, the author of a thirty-volume Chil-she lun Shu; TD 41, number 1822.
Two other disciples of Hsiian-tsang, Huai-su and K'uei-chi, have written commentaries on the Kosa which are lost. P'u-kuang has also written a short treatise on the teachings of the Ko/a.
d. Yuan-hui wrote a thirty-volume Shu on the Kdrikds of the Kosa, a work with a preface written by Chia-ts'eng and dated before 727; this work, the Chil-she lun sung Shu (var.
Chil-she lun sung shih), is preserved in TD volume 41, number 1823. This Shu "was commented upon many times in China and very widely disseminated in Japan; it is from this intermediary text that Mahayanists in general draw their knowledge of the Kola. But from the point of view of Indology, it does not offer the same interest as the three preceding com- mentaries. "
Hsiian-tsang dictated his version of Samghabhadra to Yuan-yu. There are some fragments of a commentary written by him.
6. Gunamati and the Laksandnsdra.
Gunamati is known through his commentary on the Vydkhydyukti\ many fragments of this commentary are quoted in the Chos-'byun of Bu-ston, trans. Obermiller, 1931. It is mentioned four times by Yasomitra in his Abhidharmakofavydkhyd.
a. Introductory stanzas: Gunamati comments on the Kosa, as has Vasumitra; Yasomitra follows this commentary when it is correct.
b. "Gunamati and his disciple Vasumitra say that the word nutrias is declined in the fourth case. But when the word namas is not independent, we have the accusative. This is why this master (Vasubandhu), in the Vydkhydyukti, says, 'Saluting the Muni with my head' . . . " (Kosa, Vydkhyd, i p. 7).
c. Gunamati holds that the Kosa wrongly teaches that "Conditioned things,
Poussin 9
? 10 Introduction
with the exception of the Path, are sasrava (Kosa, i. 4b)," for all of the dharmas, without exception, can be taken as an object by the dsravas(Vydkhyd i, p. 13).
d. On the subject of the continuity of the mental series, "the master Gunamati, with his disciple the master Vasumitra, through affection for the doctrine of his own nikaya, instead of confining themselves to explaining the Kosa, refute it" (Kosa,iiL Ua-b, note).
N. Peri (Date, 41) recalls that Burnouf mentioned (Introduction, 566), according to Yasomitra, the commentary of Gunamati. He adds: "An author very rarely quoted. His Laksananusarasastra (Taisho 1641) forms part of the Canon, where it is classified among the HInayana works. It summarizes the ideas of the Kosa, and then presents his own opinions on several points. The Hsi-yu-chi, after having listed him among the celebrated monks of Nalanda (TD 51, p. 924a2), tells us that he left the monastery where he had been living in order to move to V alabhF(p. 936c2). "
Taisho 1641 is only an extract of the treatise of Gunamati, the chapter which examines the sixteen aspects of the truths (Kosa, vii. 13): "Do we have sixteen things or sixteen names? The masters of the Vibhasa say that sixteen names are posited because there are sixteen things. But the siltra-upadesa masters say that there are sixteen names, but only seven things; four things for the first Truth, one thing for each of the three others. In the beginning the Buddha promulgated the Upadesasutra. After the disappearance of the Buddha, Ananda, Katyayana, etc. , recited that which they had heard. In order to explain the meaning of the Sutra, as disciples do, they composed a sastra explaining the Sutra, which is thus called a siltra-upadesa. Then the Vibhasa extracted an upadesa from that which was to be found [in this upadesa']; since it only indirectly comes from the Sutra, it is not called a sutra-upadefa. "
Gunamati continues as in the Kosa, vii. 13a, "According to the first explana- tion, anitya, impermanent, because it arises dependent on causes (pratyayd- dhinatvdt)" And he comments, "Conditioned things, without force, do not arise in and of themselves . . . "
The first volume ends, "The thesis of Vasubandhu is similar to the meaning of the siitra-upadesa masters. "
The second begins, "The author says, 1 am now going to give the explanation of what I believe. Anitya, impermanent, because, having arising, it has extinction. Conditioned things, having arising, and extinction, are not permanent. Arising is existence . . . "
The treatise touches on diverse points of philosophy, the absence of dtman, etc. In this work we encounter some very interesting notes, for example (Taisho,
? page 168b9), "In the Hinayana, the pretas are superior to animals; in the Mahayana, the opposite. In fact, the pretas are enveloped in flames . . . "
It is curious that the title of the work of Gunamati, literally Laksandnusd- rasdstra, is exactly identical to that of the book attributed to Purnavardhana in the Tanjur. We have Gunamati, a teacher of Sthiramati, and Purnavarudhana, a
25
student of Sthiramati.
7. Sthiramati, a student of Gunamati, defended the Kosa against Samgha-
bhadra. "His commentary on the Kosa is mentioned many times by Shen-t'ai, P'u-kuang, and Fa-pao in their work on the same text. The precise manner in which they-quote it, in which they note and discuss its opinions, causes us to believe that Hsiian-tsang may have brought it to China, and perhaps they themselves had also read it" (N. Pe'ri, Date, 41). Sthiramati, the author of the Tsa-chi, is one of the great masters of the Vijnaptimitrata system.
There exists (Taisho 1561) a small treatise by Sthiramati (transcription and translation) entitled Kosatattvdrthatikd or Abhidharmakosasdstratattvdrathaptkd, which is doubtless an extract of a voluminous work of the same name and by the same author preserved in Tibetan (Cordier, 499).
We observed, at the beginning, the commentary on the seven points indicated in the introductory stanza of the Kosa.
On the wisdom of the Buddha, superior to that of the saints, the author quotes the Kalpanamanditika stanza (Huber, Sutrdlamkara), Kosa, i. l, vii. 30; and recalls the ignorance that Maudgalyayana had of the place where his mother had been reborn, Kosa, i. 1.
In order to demonstrate the thesis of the Kosa that sraddhendriya can be impure, ii. 9, the author quotes at length the sutra on the request of Brahma to the Buddha (setting into motion the Wheel of the Law), a sutra briefly indicated by V asubhandu.
The work ends with some remarks on the duration of life: The stanza says: "Among the Kurus life is always 1,000 years in length; half of this to the west and the east. In this continent, it is not set: atits end, some ten years; in the beginning, without measure" {Kosa, iii. 75-77), "There are, in fact, in this world, some beings who have extra meritorious actions and who make the resolution, 'May I have a long life! , without desiring more precisely, 'May I live one hundred years, ninety years, eighty years! ' Or rather some venerable persons, parents and friends, say, 'May you live long! ' without saying more precisely how long a time. If one makes similar vows, it is because the actions done by persons of this continent are associated with thoughts of desire. The Sutra says, 'Know, oh Bhiksus that the length of life was over 80,000 years under Vipasyin, 20,000 years under Kasyapa;
Ponssin 11
? 12 Introduction
the length of life is now 100 years; few will go beyond this, and many will have less. ' If the length of life is not set, why does the Blessed One express himself in this way? . . . " The treatise concludes with the well-known stanza: sucirna- brahmacarye'smin . . . (Kosa, vi. 60a).
8. Samghabhadra has written two works.
The first (TD 29, number 1562), the title of which is transcribed into Chinese as Abhidharmanydyanusdrasdstra--or perhaps better as Nydydnusdro ndma Abhidharmaidstram--is a commentary which reproduces without any changes the Kdrikds of the Abhidharmakosa. But this eighty-volume commentary criticizes the Kdrikds, which present the Vaibhasika doctrine by noting them with the word kila, which means "in the words of the School"; it refutes the Bhdsyam, the auto-commentary of Vasubandhu, when this work presents views opposed to those of the Vaibhasikas,and it corrects them when it attributes to the Vaibhasikas views which are not theirs.
The title of the second treatise (TD 29, number 1563) is not completely transcribed: Abhidharmasamaya-hsien-sdstra or Abhidharmasamaya-kuang-sastra. J. Takakusu proposes Abhidharmasamayapradipikds'dstra, which is not bad;
however pradipa, "lamp," is always teng, and we have for hsien the equivalents prakdsa and dyotana.
This is a forty-volume extract from the Nydydnusdra, from which all polemic
is excluded and which is thus a simple presentation of the system (samaya) of the
Abhidharma. It differs from the Nydydnusdra by the presence of a rather long
introduction, in seven stanzas and prose, and also by the manner in which it treats
the Kdrikds of Vasubandhu: these Kdrikds are either omitted (ii. 2-3) or corrected
( i l l , 14) when they express false doctrines or when they cast suspicion on true
26 doctrines by the addition of the word kila.
Samghabhadra is an innovator, and K'uei-chi distinguishes the earlier and the later Sarvastivadins, Siddhi, 45 (theory of atoms), 65 (laksanas of "conditioned things"), 71 (the viprayukta called ho-ho)y 147 (vedand? ), and 311 (divergent Sarvastivadins, on adhimoksa).
***
(Additions to the Bibliography, by Hubert Durt. )
The following titles are editions of texts and works related to the Abhi- dharmakosabhasyam, which have appeared in print since the first appearance of de La Valle'e Poussin's French translation (1923-1931).
? Sanskrit:
Gokhale, V. V. , The Text of the Abhidharmakosakdrikd of VasubandhuJournal
of the Bombay Branch, Royal Asiatic Society, n. s. , vol. 22,1946, p. 73-102. Pradhan, P. , Abhidharm-Koshabha/ya of Vasubandhu, Tibetan Sanskrit Works
Series, vol. vrn, K. P. Jayaswal Research Institute, Patna, 1967; 2nd edition, revised with introduction and indices, by Dr. Aruna Haldar, 1975.
Shastri, Swami Dwarikadas, Abhidharmakosa & Bhdsya of Acharya Vasu- bandhu with Sphutdrthd Commentary of Acdrya Yasomitra, Part I (I and II Kosasthdna), critically edited, Bauddha Bharati, Varanasi, 1970.
Wogihara, Unrai, Sphutdrthd Abhidharmakosavyakhya, 2 vols. , Tokyo 1933-1936; photomechanical reprint edition, Tokyo, 1971.
Tibetan:
Otani University, The Tibetan Tripitaka, Peking Edition, vol. 115 (number
5590), to Vol. 119 (number 5597), Suzuki Research Foundation, Tokyo, 1962.
Qiinese:
Takakusu, J. K. Watanabe, The Tripitaka in Chinese, vol. 29 (numbers 1558 to
1563), The Taisho Issai-kyo kanko kwai, Tokyo 1926.
Otani University, Index to the Taisho Tripitaka, no. 16, Bidon-bu III (vol. 29), Research Association for the Terminology of the Taisho Tripitaka, Tokyo, 1962.
Funabashi, Suisai and Issai Funabashi, Kando Abidatsuma Kusharon Sakuin, Kyoto, 1956. This index is based on the Chinese version of Kyokuga Saeki--the Qiinese version used by de La Valine Poussin--, the Kand6-bon Kusharon, in thirty volumesJCyoto, 1887.
Hi. The Date of Vasubandhu. The Former Vasubandhu.
We shall not undertake here a bibliography of Vasubandhu. But his treatise,
the Pratityasamutpddavydkhyd (Cordier, iii. 365), calls for the attention of the
reader of the Kosa. G. Tucci has published some fragments of this work (JRAS.
1930, 611-623) where the twelve links in the chain are explained in detail, with
numerous quotations from scriptures. G. Tucci also proposes to publish the
21
Trisvabhdvakarika and some parts of the commentary to the Madhydntavibhdga.
Concerning the "definition of pratyaksa by Vasubandhu," vdsubdndhava pratyaksalaksana, known through the Tdtaparyatikd, 99, and the Vddavidhi attributed to Vasubandhu, see the articles by G. Tucci, A. B. Keith, R Iyengar, JRAS. 1929, 473; Ind. Hist. Quarterly, 1928, 221; 1929, 81; Stcherbatski, Logic, ii. 161,382; G. Tucci, Maitreya [ndtha] et Asanga, 70-71, and finally Pramdna-
Poussin 13
? 14 Introduction
samuccaya, chap, i, by R. Iyengar, pp. 31-35. It appears that Dignaga denies the authorship of the Vddavidhi to Vasubandhu,in spite of universal opinion, and the Ttka quotes Kosa ii. 64, which contradicts the above-mentioned definition of pratyaksa. There are also numerous passages of the Vydkhyayukti in the Chos- 'byun of Bu-ston (above p. 16 ).
Wassiliew, Buddhismus, 235 (1860): "Life of Vasubandhu. "
Kern, Geschiedenis, trans. Huet, ii. 450.
S. Levi, JA. , 1890, 2. 252; Theatre indien, 1890, i. 165, ii. 35; "Donations
religieuses des rois de ValabhT' {Htes Etudes, vii, p. 97); "Date de Candragomin," BEFEO. ,1903,47;Sutrdkmkdra, trans,preface,2-3,1911.
Biihler, Alter der indischen Kunst-Poesie, p. 97,1890.
J. Takakusu, "Life of Vasubandhu," Voung-pao, 1904; "A Study of Para-
metria's Life of Vasubandhu and the date of Vasubandhu," JRAS. , 1905; "Sdmkhyakdrikd," BEFEO, 1904.
f
Wogihara, Asanga s Bodhisattvabhumi, 14, Strasbourg thesis, Leipzig, 1908.
Noel Peri, "A propos de la date of Vasubandhu," BEFEO. , 1911, 339-392.
Pathak, Bhandarkar, Indian Antiquare, 1911 - 1912 (V Smith, History, 3rd edition 328,4th edition, 346).
B. Shiiwo [Benkyo Shiio], Dr. Takakusu and Mr. Peri on the date of Vasubandhu (270-350), Tetsugaku Zasshi, Nov. -Dec. 1912.
Winternitz, ]eshichte, ii. 256 (1913), iii. 693 (1922).
H. . Ui, "On the Author of the Mahdydnasutrdlamkdra" Z. fur Ind. und Iranistik, vi. 1928,216-225.
