Then: "The
acquisition
of karman is of four types.
Abhidharmakosabhasyam-Vol-1-Vasubandhu-Poussin-Pruden-1991
He brought together the diverse teachings dealing with disparate subjects and composed a Miscellaneous Skandhaka out of them; the teachings relative to the samyojanas, to the jndnas, to karman, to the mahdbhutas, to the indriyas, to samddhi and to the drspis constitute the Samyojanaskandhaka, etc.
In this same way all the Uddnagdthds were spoken by the Buddha: the Buddha Bhagavat spoke them, in diverse places for the benefit of different persons, in accord with circumstances.
After the Buddha left the world the Bhadanta Dharmatrata, who knew them from tradition, compiled them together and gave [to the groups] the name of varga.
He brought together the gathas relative to impermanence and made the Anitya-varga out of them, and so forth.
"The Abhidharma was originally the word of the Buddha; it is also a compilation of the Aryan KatyayanTputra.
"Whether the Buddha spoke [the Abhidharma], or whether the disciple spoke it does not contradict Dharmata, for all the Buddhas want the bhiksus to uphold the Abhidharma. Thus this Aryan, whether he knew the Abhidharma from tradition, or whether he sees and examines it by the light of his pranidhijndna, composed this treatise in order that the Good Law should remain a long time in the world . . . "
b. The Prakarana of Vasumitra.
This Is also called the Prakaranagrantha, or the Prakaranapadafastra: it is an important work, but little systematized (for many things have been brought together in the chapter of "The One Thousand Questions"); frequently quoted in the Kosa (for example, 17,9, ii. 4l, 51, 54 . . . ).
On one important point it differs from classical Vaibhas,ikavada: it ignores the akufalamahdbhumikas (iii. 32). Sometimes it expresses itself in terms which one must interpret with some violence to make them correct (ii. 46, 52, ii. 4, 41). It differs from the Jndnaprasthana, v. 10.
Ignorance of the akufalamahdbhumika category seems to prove that the Prakarana is earlier than the Jndnaprasthana.
Sometimes the authors of the Vibhdsd (p. 231c3) are unsure:
"Why does this treatise (the Jndnaprasthana) say prthagjanatva and not
? pphagjanadharma, whereas the Prakaranapada says prthagjanadharma and not prthagjanatva} . . . This Treatise having said prthagjanatva, the Prakaranapada does not repeat it; this Treatise not having said prthagjanadharma, the Prakarana- pada says prthagjanadharma. This indicates that this treatise was composed after that one. There are some persons who say: that Treatise having said prthagjana- dharma,thisTreatisedoesnotrepeatit. . . ;thisindicatesthatthatTreatisehas been composed after this one. "
The Prakarana does not enumerate the indriyas in the same order as the Sutra, the Jndnaprasthdna, or early Pali scholasticism (Kosa, i. 48).
c. The Vijndnakdya.
This is a work that some Chinese sources (quoted in Takakusu) place one hundred years after the Parinirvana; attributed to Devasarman or to lha-skyid (Devaksema? ). Concerning its author, who has the title of arhat in Hsuan-tsang, see Wassiliew in Taranatha, 296, Hiouen-thsang [=Hsuan-tsang], Vie, 123, Watters, i. 373.
The interest of this book, though small from the point of view of doctrine, is
notable from the point of view of history. The first chapter, Aiaudgalydyanaskan-
dhaka, and the second, Pudgalaskandhaka, are related to two great contro- 40
versies, the existence of the past and the future, and the existence of the pudgala. Devasarman refutes the doctrine of Mu-lien or Maudgalyayana: this latter denies the existence of the past and future, exaaly as does Tissa Moggaliputta in
the Pali language ecclesiastical histories.
Here we have, from the Sarvastivadin side, the controversy which gave rise to
the council of Asoka. According to the legend that Buddhaghosa has spread to Ceylon and to London, the king was assured that the Buddha was "a follower of distinction" (vibhajyavddin)--that is to say, probably, not totally accepting "the existence of all" (sarvdstivdda); he then charges Tissa Moggaliputta, that is to say, I believe, our Mu-lien, to preside over a council where only the opponents of the
41
existence of the past and the future were admitted
There is not a very close relationship between the Maudgalydyanaskandhaka
and the work of Tissa (Kathdvatthu, i. 6 and following). We should not be surprised at this, since the two works represent and bring about the triumph of two opposing doctrines.
On the contrary, the Pudgalaskandhaka presents, together with Kathavat- thu, i. l, some close analogies to this text even down to an identity of phrases.
4,2 Devasarman speaks of two masters--a follower of pudgala (pudgalavddin),
who admits a vital principle, a type of soul or self (pudgala), and a follower of
Poussin 21
? 22 Introduction
emptiness {sunyatavadin), that is to say a negator of the soul {dtman), an orthodox Buddhist who does not recognize any permanent principle.
1. The thesis of the pudgalavddin is formulated in terms which are partially
43
identical to those that the puggakvddin of the Kathavatthu employs.
2. The arguments are in part the same:
a) Argument taken from the passing from one realm of rebirth to another
(compare Kathavatthu, 11, 158-161).
b) Argument taken from the passing from one degree of holiness into another
{ibid. , LI, 221).
c) Connection between the doer of the action and the "partaker of its results"
{ibid. , i. l, 200).
d) Is suffering "done by oneself or "done by another"? {ibid. , i. l, 212).
e) Is the pudgala conditioned {samskrta) or unconditioned? {ibid. , i. l, 127). f) The pudgala is not perceived by any of the six consciousnesses; the
consciousnesses arise from well-known causes, without the intervention of a
pudgala {Kathavatthu, passim).
3- The method of argumentation is the same in the Sanskrit source and in the Pali source. The negator of the pudgala puts the follower of the pudgala in contradiction to the sutras, that is to say in contradiction with himself--for the follower of pudgala recognizes that the Buddha has well said all that he has said.
"The pudgalavddin says: There is a self {dtman), a living being {sattvd), a living principle (jiva), a being that arises (jantu), a being that nourishes itself
44
{posa), a person {purusa), a pudgala. Because there is a pudgala, he does actions
which should bear an agreeable result {sukhavedaniya), a disagreeable result, or a result neither disagreeable nor agreeable. Having done these three types of actions, he experiences, accordingly, sensations which are agreeable, disagreeable, neither disagreeable nor agreeable.
"The sunyatavadin asks him: Yes or no, it is the same person who does the action and who experiences the sensation?
"The pudgalavddin answers: No.
45
"Recognize the contradiction into which you fall! If there is a pudgala, and
if, because there is a pudgala, he does actions and experiences their proper retribution, then one should say that it is the same person who does the action and who experiences the sensation: hence your answer is illogical. If you now deny that it is the same person who does the action and who experiences the sensation, then one should not say that there is a self, a living being et cetera. To say this is illogical.
"If the pudgalavddin answers: It is the same person who does the action and
? who experiences the sensation', then he should be asked: Yes or no, is what the Bhagavat says in the Sutra well said, well defined, well declared, namely, 'Oh Brahmin, to say that it is the same person who does the action and who
46
experiences the sensation, is to fall into the extreme opinion of permanence? " "The pudgalavddin answers: Yes, this is well said.
"Recognize the contradiction into which you fall. . . "
The relation between the Pali and the Sanskrit Abhidharma treatises is close.
The comparison between the Prakarana and the Dhdtukdya with the Dham- masangani brings out, as does that of the Vijndnakdya with the Kathdvatthu, numerous evidences of the unity of this scholasticism. The controversy of the pudgala is, without doubt, one of the kathdvatthus, one of the oldest subjects of discussion. Presented in Pali and in Sanskrit according to the same principles, with, often, the same arguments and striking coincidences of phraseology--- clearer in Devasarman, but more archaic, it appears to me, in Tissa--it cannot fail to clarify to a certain degree the history of the gravest conflict to agitate early Buddhism. We may be surprised that Devasarman's pudgalavddin does not make anything of the sutra on the bearer of the burden, a sutra which is one of the principal authorities of Vasubandhu's pudgalavddin {Kosa, be).
As for the Kathdvatthu, it is not imprudent to think that this book is made up of bits and pieces. Certain parts are old, other parts are suspect.
d. The Dharmaskandha.
Takakusa asks if the compilation of this name is the work of Sariputra (Yasomitra) or of Maudgalyayana (Chinese title); but this is quite a useless concern.
This is a collection of sutras, promulgated in Jetavana, addressed to the bhiksus, preceded by two stanzas: "Homage to the Buddha . . . The Abhidharma is like the ocean, a great mountain, the great earth, the great sky. I wish to make an effort to present in summary the riches of Dharma which are found in it. "
The author, in fact, comments most frequently on the sutras which he quotes by quoting other sutras: "Among these four, what is stealing? The Bhagavat says . . . "
Without any doubt, the author was a scholarly man and well informed concerning the most subtle doctrines of the Sarvastivada: "The Bhagavat, in Jetavana, said to the bhiksus: 'There are four srdmanyaphalas, results of the religious life. What are these four? The result of srotadpanna . . . What is the
result of srotadpanna} It is twofold, conditioned and unconditioned (samskrta, asamskrta). Conditioned, that is to say the acquisition of this result and that which
Ponssin 23
? 24 Introduction
isacquiredthroughthisacquisition,thepreceptsoftheSaiksa. . . allthedharmas oftheSaiksa. Unconditioned,thatistosaythecuttingoffofthethreebonds . . . " (compare Kosa, vi. 51, 76).
e. The Prajnaptis*astro.
a. The Tibetan Prajndptifdstra is made up of three parts: Lokaprajndpti, Karanaprajndpti, and Karmaprajndpti.
The first two are described and analyzed in Cosmologie bouddhique, pp. 295-350.
The third is in the same style. The text is divided into chapters preceded by a summary. Here is the beginning:
"Summary: Intention, volition, past, good, object, sphere of desire, stanza, resume of aaions.
"1. Thus have I heard. The Bhagavat resided in Jetavana in the park of Anathapindada; he-said to the bhik? us, "I teach the retribution of intentional aaions, done and certain, retribution in this life . . . " Thus spoke the Bhagavat.
"2. There are two actions: volition aaion (cetana karman), and aaion after having willed (cetayitva karman). What is the first? It is called: cetana, abhisamcetand, cintand, cetayitatva, cittdbhisamskdra, mdnasa karman\ this is called volitional aaion . . .
"3. Volitional action is past, future, present. What is past volitional action? That which is jdta, utpanna, abhinirvrtta. . . abhyatita, ksina, niruddha, vipari- nata, atttasamgrhtia, atrtddhvasamgrhita . . .
"4. Volitional aaion is good, bad, or neutral. . .
"5. Is the objea of volitional aaion good?
"6. Volitional aaion is of the three Dhatus. What is of the sphere of desire and
of kamal
"7. A stanza in honor of the Buddha who teaches the different types of aaion. "8. One action: all actions done (literally: ekahetund karmandm samgrahah
karmeti). Two aaions; volition and aaion after having been willed. Three aaions: bodily, vocal, mental. Four aaions: of Kamadhatu, of Rupadhatu, of Arupyadhatu, and not belonging to the Dhatus. Five Aaions: definite (=either good or bad) and neutral aaions which are abandoned through Meditation, those not to be abandoned (aheya) . . . And so forth up to twelve. "
Almost all of the theories presented in Kosa, iv, are treated, with long quotations from the sutras.
Many details deserve to be mentioned. For example, "Lying arisen from ignorance (Kosa, iv. 68). " Asked by a hunter if he had seen the deer, one thinks,
? "It is not fitting that the hunter should kill the deer," and he answers that he has not seen it (compare the story of Ksantivadin, Chavannes, Cing cents contes, i. 161). Asked by the king's army if he has seen the bandits . . . Asked by the bandits if he has seen the king's troops . . . And, above all, in the case of frivolous speech arisen from desire, "or further, through attachment to examining the word of the Buddha. "
Chapter xi is interesting from another point of view. In relation to the definition of death from exhaustion of life or merit (paragraph copied by Vasubandhu, ii. 107), the story of Kasyapa the Nude (Samyutta> ii. 19-22) is cited, with some long developments: "A short time after he left the Bhagavat, he was killed by a bull. At the moment of his death, his organs became very clear; the color of his face became very pure; the color of his skin became very brilliant. " Of
note also is the fact that Kasyapa was received as an updsaka: "Master, I go to the Bhagavat: Master; I go to the Sugata; Master, I take refuge in the Bhagavat, I take refuge in the Dharma and in the Sangha. May the Bhagavat recognize (dhdretu) me as an updsaka having renounced killing . . . " (compare Samyutta, ii. 22 and Dtgha, i. 178).
Then: "The acquisition of karman is of four types. They are enumerated as in the Sangitiparydya"Then follow three paragraphs on giving: "Four gifts: it happens that the giver is pure and the recipient is impure . . . and so forth as in the Sangitiparydya. Eight gifts: the asadya gift (Kos'a, iv. 117), and so forth as in Sangitiparydya. Eight gifts: it happens that a person of little faith gives little, to immoral persons, for a short period of time . . . "
b. The Chinese Prajndptisdstra is incomplete. This edition, from its first part, gives only the title, "In the great Abhidharmasdstray the Lokaprajndpti, or first part. " And a gloss says that the Indian original is missing. There follows immediately the title of the second part: Karanaprajndpti.
The text begins as follows: "In the sdstra, the question is posed: For what reason does the Cakravartin have the jewel of a woman . . . ? " In comparing the Tibetan Karanaprajndpti, we see that the Chinese text omits the First Chapter on the laksanas and on the Bodhisattva; and that the Second Chapter omits the enumeration of the jewels and discussions on the wheel, the elephant, the horse and the jewel.
The third chapter, in Chinese as in Tibetan, is made up of stanzas on the Buddha, a king like the Cakravartin, and the jewel of the Buddha: the Dharma is a wheel; the rddhipadas are an elephant. The Tibetan tells us that these stanzas are the Sailagdthds. This refers to an edition that departs from the Suttanipdtay where the single stanza 554 has two pddas corresponding to the Tibetan: "Saila, I
Poussin 25
? 26 Introduction
am king, sublime king of the Dharma. In the circle of the earth, I set in motion the Wheel of the Dharma; like a Cakravartin king, consider the Tathagata as compassionate, full of pity, a Muni beneficial to the world. "
The Chinese text has fourteen chapters; the last, which is meteorological (rain, etc. ), corresponds closely, like the others, to the Tibetan text. This latter has four supplementary chapters: the four gatis, the five yonis, to which womb do beings of the different realms of rebirth belong, etc. It is likely that Vasubandhu had read this chapter, for his version has, like the Tibetan Prajndpti, the story of the preti who eats her ten children every day, the story of Saila, of Kapotamalim, etc. (Kola, iii. 9; Vibhdsd, TD 27, p626c).
Takakusu has ingeniously supposed that the Lokaprajndpti, omitted either by mistake or on purpose in the Chinese Prajndptisdstra, of which it should be the first "gate," is found in fact in the sdstra (Taisho 1644) entitled Li-shih: "Nanjio translates Lokasthiti (7)-abhidharmasdstra. But //' signifies constructing, establish- ing, and is practically equivalent to shih-she or prajnapti. " Thus Takakusu translates Lokaprajndpty-abhidharrnasdstra.
Taisho 1644 exhibits the characteristics of a sutra. Some editions precede the title with the words, "Spoken by the Buddha. "
The text begins: "As the Buddha Bhagavat, the Arhat, said, 'Thus have I heard. ' The Buddha resided in SravastI, in the monastery of the upasika Mrgaramatar Visakha, with many bhiksus, all arhats . . . with the exception of Ananda. Then the earth shook. And Purna Maitrayanlputra asks . . . " The chapters begin normally, "Then the Buddha said," "The Buddha said to the bhiksu Puma," "The Buddha said to the bhiksus," and end, "This is what the Buddha said; thus have I heard. "
The contents of the chapters, established by Takakusu, show that, even though it treats of subjects that the Tibetan Lokaprajndpti treats, Taisho 1644 has nothing in common with this Lokaprajndpti. In this latter there is nothing that corresponds to the chapter on the yaksas and notably to the conversion of "Satagira" and "Hemavata" (the stanzas of the Hemavatasutta of the Suttanipdta, Uragavagga, Taisho, p. 177). The Lokaprajndpti has only a summary indication of the heavenly gardens, concerning which Taisho 1644 has some long develop- ments. But in both works there is the battle of the suras and the asuras, the movements of the sun and the moon, the length of life, the hells, the three small and the three great calamities. Their order, however, differs.
The title of the chapters of Taisho 1644 do not give, sometimes, a precise idea of the contents of the book.
For example, in the First Chapter, we have 1. the two causes of the shaking of
? the earth (movement of wind, water; and the magical power of the saint who "considers the earth as small, the water as great"). 2. After two stanzas on the shaking of the earth, the Buddha says to Purna, "There are some winds named Vairambhas . . . " (Kosa, vi. 12). In this circle of wind, there is the water and the earth whose thickness and height are fixed as in the Kosa (iii. 45). 3. The Buddha explains the great hell called "Black Obscurity" which is found between the universes (and which is not mentioned in the Kosa), and the ten cold hells (Kosa, iii. 59a-c, second note) . . . A little later, Ananda manifests his admiration for the Buddha and his power. Udayin reprimands him and is, in his turn, reprimanded by the Master. This is an edition of the celebrated Suttanta
(Anguttara, i. 228), which differs from the Pali by the prophesy, "Aquatic beings are many, terrestial beings are few . . . Samayavimukta arhats {Kosa, vi. 56), are many, asamayavimukta arhats are few, and are difficult to encounter in this world: and I declare that Ananda will become an asamayavimukta Arhat. "
f. The Dhdtukdya.
1. At the beginning of this work, one finds the enumeration and the definition of the dharmas of the Sarvastivadins: 10 mahdbhumikas, 10 klesamahdbhumikas, 10 parittaklesas, 5 klesas, 5 drstis, 5 dharmas . . .
The kusalamahdbhumikas are missing, as are the akusalas.
The five klesas make a strange list: kdmardga, rilpardga, drupyardga, pratigha and vicikitsd.
More curious is the list of the five dharmas: vitarka, vicara, vijndna (understood as the six consciousnesses, eye, etc. ), dhrikya and anapatrdpya.
We can imagine that this book is from the early Sarvastivada. 2. The second part treats of samprayoga, association, and samgraha, inclusion.
"Vedand, which forms part of the mahdbhumikas, is associated with how many of the six vedanendriyas (pleasure, etc. )? With how many is it not associated? . . . " and so forth until: "Affection arisen from mental contact is associated with how many of the vedanendriyas? With how many is it not associated?
"That which is associated with vedand is included (samgrhita) in what? In the mind and mental states, eight dhdtus, two dyatanas, three skandhas. What is it that is left over? Vedand, rupa, asamskrta, the viprayuktasamskdras; that is to say, eleven dhdtus . . . "
These are precisely the type of questions that the Dhdtukathdpakarana examines: sukhindriyam . . . kehici sampayuttam katihi vippayuttam . . . ? These are the same questions: Vedandkkahandhena ye dhammd sampayuttdte dhammd
Poussin 27
? 28 Introduction
katihi khandhehi katihdyatanehi katihi dhdtuhi samgahitd? te dhammd tlhi khandhehi dvihdyatanehi atthahi dhdtuhi samgahitd (Section xii).
3. We can thus affirm the close relationship between the Dhdtukdya and the Dhdtukathdpakarana. The first, in its second part, is only a Sarvastivadin recension (theory of the mahdbhumikas, of the viprayuktasamskdras . . . ) of an earlier volume of scholastic exercises on the dharmas.
g. The Sangitiparyaya
The Sangitiparyaya is a recension of the Sangitisuttanta which forms part of the Dighanikdya.
Same niddna: The Buddha at Pava; the death of the Nirgrantha; Sariputra invites the monks to chant together the Dharma and the Vinaya so that, after the Nirvana of the Tathagata, his sons will not dispute them. Then follow chapters on the single dharmas, the pairs of dharmas. . . the tenfold dharmas. Finally the eulogy of Sariputra: sadhu sadhu, by the Bhagavat, "You have well collected and recited with the bhiksus the Ekottaradharmaparydya taught by the Tathagata . . . "
The close relationship of the Pali and the Sanskrit texts do not exclude some variants. It is thus that, among the eight-fold dharmas, the Abhidharma omits the eight mithydtvas (number one of the Pali list) and adds the eight vimoksas (omitted in the Pali list, but which figure in the Dasa-uttara). The order also differs. On the one hand mdrgdnga, pudgala, ddna, kausidyavastu, drabhayavastu, punyotpatti, parsad, lokadharma, vimoksa, abhibhvayatana; and oil the other hand micchatta, sammatta, puggala, kusitavatthu, drabbhavatthu, ddna, ddnuppatti, parisd, lokadhamma, abhibhdyatana. Note that punyotpatti is better than
1 ddnuppatti. *
Yasdmitra and Bu-ston attribute the Sangitiparyaya to Mahakausthila; the Chinese sources attribute it to Sariputra. Should we believe that in one recension, that known by Yasomitra, Mahakausthila had the role that fit the Paji and the Chinese texts assigned to Sariputra?
Takakusu says that the Sangitiparyaya, in volumes 15 and 18, quotes the Dharmaskandhasdstra. I have not encountered these quotations. The Prajndpti- sastra refers its reader to the Sangitiparyaya.
v. Some Masters of the Vibhdsa.
The Vibhdsa frequently quotes the divergent opinions of masters and different schools. This presentation is often followed by the opinion of P'ing or of the P'ing-chia: "The P'ing-chia says that the first opinion is the best one. "
? Elsewhere, as the commentators remark, "there is no Ping-chia" {Kosa, iii. 14,20, 41, Siddhi, 552,690).
A good specimen of the methods of the Vibhdsd: "If there is a pure prajnd outside of the sixteen dkdras" {Kosa, vii. 12, Vibhdsd, p. 529), why does the Vijnanakdya not say this. . . ? If not, why do the Prakarana and the Samgitiparyaya, and even this treatise, the Vibhdsd, say that. . . ? And how does one explain such a sutra? One should say that there is no pure prajnd outside of the sixteen dkdras. In this case, one understands the Vijndnakdya, but how does one explain the Prakarana . . . ? There are five reasons which justify this text. . . "
Among the masters of the Vibhdsd, of special note are Parsva, frequently quoted, and who, along with many anonymous commentators, comments on the Brahmajdla (Vibh. 98, p. 508, but see also Vibh. 175, p. 881, on the &uddhavasikas and 177, p. 889, on the number of the laksanas); Purnasa {Ko/a Hi. 28, Vibh. 23, p. 118b: Samadatta (? ), iii. 45, Vibh. 118c); Samghavasu {Vibh. 19, p. 97a, 106, p. 547a; 142, p. 732a) who recognizes only six indriyas in the absolute sense, the
jivita and the eight, eye, etc. , because these six are the root of being, sattvamula (Kola, ii. 5); and, with respect to this, Kusavarman, who only admits one indriya, the manas, a doctrine which leans towards the Vijnanavada.
48 a. Vasumitra.
1. Vasumitra is one of the great masters of the Vibhdsd, and one of the leaders of the Sarvastivadin school. His theory on "the existence of all" Is, Vasubandhu says, preferable to that of the three other masters, Dharmatrata, Ghosaka, and Buddhadeva {Kosa, v. 26).
One searches in vain in the two Abhidharmas (of the collection of seven) attributed to Vasumitra, the Prakaranapdda and the Dhdtukdya, for an allusion to this theory. Taranatha says, moreover, that the author of the Prakarana has nothing in common with the Vasumitra of the Vibhdsd (p. 68).
2. The Aryan Vasumitra Bodhisattva gives his name to a treatise (Taisho 1549). According to the preface, this was the Vasumitra who, after Maitreya, will be the Simhatathagata; the Vasumitra to whom the fathers refused entrance to the Council because he was not an arhat, and who later became the president of the Council {Hiuan-tsang, Watters, i. 271). Watters does not think that he is the great master of the Vibhdsd', indeed, the thirteenth chapter of Taisho 1549, entitled "Sarvdsti-akhanda," does not contain any references to the system of avasthdnya- thdtva of the Bhadanta Vasumitra. This is all that I dare say about this very complicated chapter.
Poussin 29
? 30 Introduction
The theory of the time periods is encountered in the second volume (p.
"The Abhidharma was originally the word of the Buddha; it is also a compilation of the Aryan KatyayanTputra.
"Whether the Buddha spoke [the Abhidharma], or whether the disciple spoke it does not contradict Dharmata, for all the Buddhas want the bhiksus to uphold the Abhidharma. Thus this Aryan, whether he knew the Abhidharma from tradition, or whether he sees and examines it by the light of his pranidhijndna, composed this treatise in order that the Good Law should remain a long time in the world . . . "
b. The Prakarana of Vasumitra.
This Is also called the Prakaranagrantha, or the Prakaranapadafastra: it is an important work, but little systematized (for many things have been brought together in the chapter of "The One Thousand Questions"); frequently quoted in the Kosa (for example, 17,9, ii. 4l, 51, 54 . . . ).
On one important point it differs from classical Vaibhas,ikavada: it ignores the akufalamahdbhumikas (iii. 32). Sometimes it expresses itself in terms which one must interpret with some violence to make them correct (ii. 46, 52, ii. 4, 41). It differs from the Jndnaprasthana, v. 10.
Ignorance of the akufalamahdbhumika category seems to prove that the Prakarana is earlier than the Jndnaprasthana.
Sometimes the authors of the Vibhdsd (p. 231c3) are unsure:
"Why does this treatise (the Jndnaprasthana) say prthagjanatva and not
? pphagjanadharma, whereas the Prakaranapada says prthagjanadharma and not prthagjanatva} . . . This Treatise having said prthagjanatva, the Prakaranapada does not repeat it; this Treatise not having said prthagjanadharma, the Prakarana- pada says prthagjanadharma. This indicates that this treatise was composed after that one. There are some persons who say: that Treatise having said prthagjana- dharma,thisTreatisedoesnotrepeatit. . . ;thisindicatesthatthatTreatisehas been composed after this one. "
The Prakarana does not enumerate the indriyas in the same order as the Sutra, the Jndnaprasthdna, or early Pali scholasticism (Kosa, i. 48).
c. The Vijndnakdya.
This is a work that some Chinese sources (quoted in Takakusu) place one hundred years after the Parinirvana; attributed to Devasarman or to lha-skyid (Devaksema? ). Concerning its author, who has the title of arhat in Hsuan-tsang, see Wassiliew in Taranatha, 296, Hiouen-thsang [=Hsuan-tsang], Vie, 123, Watters, i. 373.
The interest of this book, though small from the point of view of doctrine, is
notable from the point of view of history. The first chapter, Aiaudgalydyanaskan-
dhaka, and the second, Pudgalaskandhaka, are related to two great contro- 40
versies, the existence of the past and the future, and the existence of the pudgala. Devasarman refutes the doctrine of Mu-lien or Maudgalyayana: this latter denies the existence of the past and future, exaaly as does Tissa Moggaliputta in
the Pali language ecclesiastical histories.
Here we have, from the Sarvastivadin side, the controversy which gave rise to
the council of Asoka. According to the legend that Buddhaghosa has spread to Ceylon and to London, the king was assured that the Buddha was "a follower of distinction" (vibhajyavddin)--that is to say, probably, not totally accepting "the existence of all" (sarvdstivdda); he then charges Tissa Moggaliputta, that is to say, I believe, our Mu-lien, to preside over a council where only the opponents of the
41
existence of the past and the future were admitted
There is not a very close relationship between the Maudgalydyanaskandhaka
and the work of Tissa (Kathdvatthu, i. 6 and following). We should not be surprised at this, since the two works represent and bring about the triumph of two opposing doctrines.
On the contrary, the Pudgalaskandhaka presents, together with Kathavat- thu, i. l, some close analogies to this text even down to an identity of phrases.
4,2 Devasarman speaks of two masters--a follower of pudgala (pudgalavddin),
who admits a vital principle, a type of soul or self (pudgala), and a follower of
Poussin 21
? 22 Introduction
emptiness {sunyatavadin), that is to say a negator of the soul {dtman), an orthodox Buddhist who does not recognize any permanent principle.
1. The thesis of the pudgalavddin is formulated in terms which are partially
43
identical to those that the puggakvddin of the Kathavatthu employs.
2. The arguments are in part the same:
a) Argument taken from the passing from one realm of rebirth to another
(compare Kathavatthu, 11, 158-161).
b) Argument taken from the passing from one degree of holiness into another
{ibid. , LI, 221).
c) Connection between the doer of the action and the "partaker of its results"
{ibid. , i. l, 200).
d) Is suffering "done by oneself or "done by another"? {ibid. , i. l, 212).
e) Is the pudgala conditioned {samskrta) or unconditioned? {ibid. , i. l, 127). f) The pudgala is not perceived by any of the six consciousnesses; the
consciousnesses arise from well-known causes, without the intervention of a
pudgala {Kathavatthu, passim).
3- The method of argumentation is the same in the Sanskrit source and in the Pali source. The negator of the pudgala puts the follower of the pudgala in contradiction to the sutras, that is to say in contradiction with himself--for the follower of pudgala recognizes that the Buddha has well said all that he has said.
"The pudgalavddin says: There is a self {dtman), a living being {sattvd), a living principle (jiva), a being that arises (jantu), a being that nourishes itself
44
{posa), a person {purusa), a pudgala. Because there is a pudgala, he does actions
which should bear an agreeable result {sukhavedaniya), a disagreeable result, or a result neither disagreeable nor agreeable. Having done these three types of actions, he experiences, accordingly, sensations which are agreeable, disagreeable, neither disagreeable nor agreeable.
"The sunyatavadin asks him: Yes or no, it is the same person who does the action and who experiences the sensation?
"The pudgalavddin answers: No.
45
"Recognize the contradiction into which you fall! If there is a pudgala, and
if, because there is a pudgala, he does actions and experiences their proper retribution, then one should say that it is the same person who does the action and who experiences the sensation: hence your answer is illogical. If you now deny that it is the same person who does the action and who experiences the sensation, then one should not say that there is a self, a living being et cetera. To say this is illogical.
"If the pudgalavddin answers: It is the same person who does the action and
? who experiences the sensation', then he should be asked: Yes or no, is what the Bhagavat says in the Sutra well said, well defined, well declared, namely, 'Oh Brahmin, to say that it is the same person who does the action and who
46
experiences the sensation, is to fall into the extreme opinion of permanence? " "The pudgalavddin answers: Yes, this is well said.
"Recognize the contradiction into which you fall. . . "
The relation between the Pali and the Sanskrit Abhidharma treatises is close.
The comparison between the Prakarana and the Dhdtukdya with the Dham- masangani brings out, as does that of the Vijndnakdya with the Kathdvatthu, numerous evidences of the unity of this scholasticism. The controversy of the pudgala is, without doubt, one of the kathdvatthus, one of the oldest subjects of discussion. Presented in Pali and in Sanskrit according to the same principles, with, often, the same arguments and striking coincidences of phraseology--- clearer in Devasarman, but more archaic, it appears to me, in Tissa--it cannot fail to clarify to a certain degree the history of the gravest conflict to agitate early Buddhism. We may be surprised that Devasarman's pudgalavddin does not make anything of the sutra on the bearer of the burden, a sutra which is one of the principal authorities of Vasubandhu's pudgalavddin {Kosa, be).
As for the Kathdvatthu, it is not imprudent to think that this book is made up of bits and pieces. Certain parts are old, other parts are suspect.
d. The Dharmaskandha.
Takakusa asks if the compilation of this name is the work of Sariputra (Yasomitra) or of Maudgalyayana (Chinese title); but this is quite a useless concern.
This is a collection of sutras, promulgated in Jetavana, addressed to the bhiksus, preceded by two stanzas: "Homage to the Buddha . . . The Abhidharma is like the ocean, a great mountain, the great earth, the great sky. I wish to make an effort to present in summary the riches of Dharma which are found in it. "
The author, in fact, comments most frequently on the sutras which he quotes by quoting other sutras: "Among these four, what is stealing? The Bhagavat says . . . "
Without any doubt, the author was a scholarly man and well informed concerning the most subtle doctrines of the Sarvastivada: "The Bhagavat, in Jetavana, said to the bhiksus: 'There are four srdmanyaphalas, results of the religious life. What are these four? The result of srotadpanna . . . What is the
result of srotadpanna} It is twofold, conditioned and unconditioned (samskrta, asamskrta). Conditioned, that is to say the acquisition of this result and that which
Ponssin 23
? 24 Introduction
isacquiredthroughthisacquisition,thepreceptsoftheSaiksa. . . allthedharmas oftheSaiksa. Unconditioned,thatistosaythecuttingoffofthethreebonds . . . " (compare Kosa, vi. 51, 76).
e. The Prajnaptis*astro.
a. The Tibetan Prajndptifdstra is made up of three parts: Lokaprajndpti, Karanaprajndpti, and Karmaprajndpti.
The first two are described and analyzed in Cosmologie bouddhique, pp. 295-350.
The third is in the same style. The text is divided into chapters preceded by a summary. Here is the beginning:
"Summary: Intention, volition, past, good, object, sphere of desire, stanza, resume of aaions.
"1. Thus have I heard. The Bhagavat resided in Jetavana in the park of Anathapindada; he-said to the bhik? us, "I teach the retribution of intentional aaions, done and certain, retribution in this life . . . " Thus spoke the Bhagavat.
"2. There are two actions: volition aaion (cetana karman), and aaion after having willed (cetayitva karman). What is the first? It is called: cetana, abhisamcetand, cintand, cetayitatva, cittdbhisamskdra, mdnasa karman\ this is called volitional aaion . . .
"3. Volitional action is past, future, present. What is past volitional action? That which is jdta, utpanna, abhinirvrtta. . . abhyatita, ksina, niruddha, vipari- nata, atttasamgrhtia, atrtddhvasamgrhita . . .
"4. Volitional aaion is good, bad, or neutral. . .
"5. Is the objea of volitional aaion good?
"6. Volitional aaion is of the three Dhatus. What is of the sphere of desire and
of kamal
"7. A stanza in honor of the Buddha who teaches the different types of aaion. "8. One action: all actions done (literally: ekahetund karmandm samgrahah
karmeti). Two aaions; volition and aaion after having been willed. Three aaions: bodily, vocal, mental. Four aaions: of Kamadhatu, of Rupadhatu, of Arupyadhatu, and not belonging to the Dhatus. Five Aaions: definite (=either good or bad) and neutral aaions which are abandoned through Meditation, those not to be abandoned (aheya) . . . And so forth up to twelve. "
Almost all of the theories presented in Kosa, iv, are treated, with long quotations from the sutras.
Many details deserve to be mentioned. For example, "Lying arisen from ignorance (Kosa, iv. 68). " Asked by a hunter if he had seen the deer, one thinks,
? "It is not fitting that the hunter should kill the deer," and he answers that he has not seen it (compare the story of Ksantivadin, Chavannes, Cing cents contes, i. 161). Asked by the king's army if he has seen the bandits . . . Asked by the bandits if he has seen the king's troops . . . And, above all, in the case of frivolous speech arisen from desire, "or further, through attachment to examining the word of the Buddha. "
Chapter xi is interesting from another point of view. In relation to the definition of death from exhaustion of life or merit (paragraph copied by Vasubandhu, ii. 107), the story of Kasyapa the Nude (Samyutta> ii. 19-22) is cited, with some long developments: "A short time after he left the Bhagavat, he was killed by a bull. At the moment of his death, his organs became very clear; the color of his face became very pure; the color of his skin became very brilliant. " Of
note also is the fact that Kasyapa was received as an updsaka: "Master, I go to the Bhagavat: Master; I go to the Sugata; Master, I take refuge in the Bhagavat, I take refuge in the Dharma and in the Sangha. May the Bhagavat recognize (dhdretu) me as an updsaka having renounced killing . . . " (compare Samyutta, ii. 22 and Dtgha, i. 178).
Then: "The acquisition of karman is of four types. They are enumerated as in the Sangitiparydya"Then follow three paragraphs on giving: "Four gifts: it happens that the giver is pure and the recipient is impure . . . and so forth as in the Sangitiparydya. Eight gifts: the asadya gift (Kos'a, iv. 117), and so forth as in Sangitiparydya. Eight gifts: it happens that a person of little faith gives little, to immoral persons, for a short period of time . . . "
b. The Chinese Prajndptisdstra is incomplete. This edition, from its first part, gives only the title, "In the great Abhidharmasdstray the Lokaprajndpti, or first part. " And a gloss says that the Indian original is missing. There follows immediately the title of the second part: Karanaprajndpti.
The text begins as follows: "In the sdstra, the question is posed: For what reason does the Cakravartin have the jewel of a woman . . . ? " In comparing the Tibetan Karanaprajndpti, we see that the Chinese text omits the First Chapter on the laksanas and on the Bodhisattva; and that the Second Chapter omits the enumeration of the jewels and discussions on the wheel, the elephant, the horse and the jewel.
The third chapter, in Chinese as in Tibetan, is made up of stanzas on the Buddha, a king like the Cakravartin, and the jewel of the Buddha: the Dharma is a wheel; the rddhipadas are an elephant. The Tibetan tells us that these stanzas are the Sailagdthds. This refers to an edition that departs from the Suttanipdtay where the single stanza 554 has two pddas corresponding to the Tibetan: "Saila, I
Poussin 25
? 26 Introduction
am king, sublime king of the Dharma. In the circle of the earth, I set in motion the Wheel of the Dharma; like a Cakravartin king, consider the Tathagata as compassionate, full of pity, a Muni beneficial to the world. "
The Chinese text has fourteen chapters; the last, which is meteorological (rain, etc. ), corresponds closely, like the others, to the Tibetan text. This latter has four supplementary chapters: the four gatis, the five yonis, to which womb do beings of the different realms of rebirth belong, etc. It is likely that Vasubandhu had read this chapter, for his version has, like the Tibetan Prajndpti, the story of the preti who eats her ten children every day, the story of Saila, of Kapotamalim, etc. (Kola, iii. 9; Vibhdsd, TD 27, p626c).
Takakusu has ingeniously supposed that the Lokaprajndpti, omitted either by mistake or on purpose in the Chinese Prajndptisdstra, of which it should be the first "gate," is found in fact in the sdstra (Taisho 1644) entitled Li-shih: "Nanjio translates Lokasthiti (7)-abhidharmasdstra. But //' signifies constructing, establish- ing, and is practically equivalent to shih-she or prajnapti. " Thus Takakusu translates Lokaprajndpty-abhidharrnasdstra.
Taisho 1644 exhibits the characteristics of a sutra. Some editions precede the title with the words, "Spoken by the Buddha. "
The text begins: "As the Buddha Bhagavat, the Arhat, said, 'Thus have I heard. ' The Buddha resided in SravastI, in the monastery of the upasika Mrgaramatar Visakha, with many bhiksus, all arhats . . . with the exception of Ananda. Then the earth shook. And Purna Maitrayanlputra asks . . . " The chapters begin normally, "Then the Buddha said," "The Buddha said to the bhiksu Puma," "The Buddha said to the bhiksus," and end, "This is what the Buddha said; thus have I heard. "
The contents of the chapters, established by Takakusu, show that, even though it treats of subjects that the Tibetan Lokaprajndpti treats, Taisho 1644 has nothing in common with this Lokaprajndpti. In this latter there is nothing that corresponds to the chapter on the yaksas and notably to the conversion of "Satagira" and "Hemavata" (the stanzas of the Hemavatasutta of the Suttanipdta, Uragavagga, Taisho, p. 177). The Lokaprajndpti has only a summary indication of the heavenly gardens, concerning which Taisho 1644 has some long develop- ments. But in both works there is the battle of the suras and the asuras, the movements of the sun and the moon, the length of life, the hells, the three small and the three great calamities. Their order, however, differs.
The title of the chapters of Taisho 1644 do not give, sometimes, a precise idea of the contents of the book.
For example, in the First Chapter, we have 1. the two causes of the shaking of
? the earth (movement of wind, water; and the magical power of the saint who "considers the earth as small, the water as great"). 2. After two stanzas on the shaking of the earth, the Buddha says to Purna, "There are some winds named Vairambhas . . . " (Kosa, vi. 12). In this circle of wind, there is the water and the earth whose thickness and height are fixed as in the Kosa (iii. 45). 3. The Buddha explains the great hell called "Black Obscurity" which is found between the universes (and which is not mentioned in the Kosa), and the ten cold hells (Kosa, iii. 59a-c, second note) . . . A little later, Ananda manifests his admiration for the Buddha and his power. Udayin reprimands him and is, in his turn, reprimanded by the Master. This is an edition of the celebrated Suttanta
(Anguttara, i. 228), which differs from the Pali by the prophesy, "Aquatic beings are many, terrestial beings are few . . . Samayavimukta arhats {Kosa, vi. 56), are many, asamayavimukta arhats are few, and are difficult to encounter in this world: and I declare that Ananda will become an asamayavimukta Arhat. "
f. The Dhdtukdya.
1. At the beginning of this work, one finds the enumeration and the definition of the dharmas of the Sarvastivadins: 10 mahdbhumikas, 10 klesamahdbhumikas, 10 parittaklesas, 5 klesas, 5 drstis, 5 dharmas . . .
The kusalamahdbhumikas are missing, as are the akusalas.
The five klesas make a strange list: kdmardga, rilpardga, drupyardga, pratigha and vicikitsd.
More curious is the list of the five dharmas: vitarka, vicara, vijndna (understood as the six consciousnesses, eye, etc. ), dhrikya and anapatrdpya.
We can imagine that this book is from the early Sarvastivada. 2. The second part treats of samprayoga, association, and samgraha, inclusion.
"Vedand, which forms part of the mahdbhumikas, is associated with how many of the six vedanendriyas (pleasure, etc. )? With how many is it not associated? . . . " and so forth until: "Affection arisen from mental contact is associated with how many of the vedanendriyas? With how many is it not associated?
"That which is associated with vedand is included (samgrhita) in what? In the mind and mental states, eight dhdtus, two dyatanas, three skandhas. What is it that is left over? Vedand, rupa, asamskrta, the viprayuktasamskdras; that is to say, eleven dhdtus . . . "
These are precisely the type of questions that the Dhdtukathdpakarana examines: sukhindriyam . . . kehici sampayuttam katihi vippayuttam . . . ? These are the same questions: Vedandkkahandhena ye dhammd sampayuttdte dhammd
Poussin 27
? 28 Introduction
katihi khandhehi katihdyatanehi katihi dhdtuhi samgahitd? te dhammd tlhi khandhehi dvihdyatanehi atthahi dhdtuhi samgahitd (Section xii).
3. We can thus affirm the close relationship between the Dhdtukdya and the Dhdtukathdpakarana. The first, in its second part, is only a Sarvastivadin recension (theory of the mahdbhumikas, of the viprayuktasamskdras . . . ) of an earlier volume of scholastic exercises on the dharmas.
g. The Sangitiparyaya
The Sangitiparyaya is a recension of the Sangitisuttanta which forms part of the Dighanikdya.
Same niddna: The Buddha at Pava; the death of the Nirgrantha; Sariputra invites the monks to chant together the Dharma and the Vinaya so that, after the Nirvana of the Tathagata, his sons will not dispute them. Then follow chapters on the single dharmas, the pairs of dharmas. . . the tenfold dharmas. Finally the eulogy of Sariputra: sadhu sadhu, by the Bhagavat, "You have well collected and recited with the bhiksus the Ekottaradharmaparydya taught by the Tathagata . . . "
The close relationship of the Pali and the Sanskrit texts do not exclude some variants. It is thus that, among the eight-fold dharmas, the Abhidharma omits the eight mithydtvas (number one of the Pali list) and adds the eight vimoksas (omitted in the Pali list, but which figure in the Dasa-uttara). The order also differs. On the one hand mdrgdnga, pudgala, ddna, kausidyavastu, drabhayavastu, punyotpatti, parsad, lokadharma, vimoksa, abhibhvayatana; and oil the other hand micchatta, sammatta, puggala, kusitavatthu, drabbhavatthu, ddna, ddnuppatti, parisd, lokadhamma, abhibhdyatana. Note that punyotpatti is better than
1 ddnuppatti. *
Yasdmitra and Bu-ston attribute the Sangitiparyaya to Mahakausthila; the Chinese sources attribute it to Sariputra. Should we believe that in one recension, that known by Yasomitra, Mahakausthila had the role that fit the Paji and the Chinese texts assigned to Sariputra?
Takakusu says that the Sangitiparyaya, in volumes 15 and 18, quotes the Dharmaskandhasdstra. I have not encountered these quotations. The Prajndpti- sastra refers its reader to the Sangitiparyaya.
v. Some Masters of the Vibhdsa.
The Vibhdsa frequently quotes the divergent opinions of masters and different schools. This presentation is often followed by the opinion of P'ing or of the P'ing-chia: "The P'ing-chia says that the first opinion is the best one. "
? Elsewhere, as the commentators remark, "there is no Ping-chia" {Kosa, iii. 14,20, 41, Siddhi, 552,690).
A good specimen of the methods of the Vibhdsd: "If there is a pure prajnd outside of the sixteen dkdras" {Kosa, vii. 12, Vibhdsd, p. 529), why does the Vijnanakdya not say this. . . ? If not, why do the Prakarana and the Samgitiparyaya, and even this treatise, the Vibhdsd, say that. . . ? And how does one explain such a sutra? One should say that there is no pure prajnd outside of the sixteen dkdras. In this case, one understands the Vijndnakdya, but how does one explain the Prakarana . . . ? There are five reasons which justify this text. . . "
Among the masters of the Vibhdsd, of special note are Parsva, frequently quoted, and who, along with many anonymous commentators, comments on the Brahmajdla (Vibh. 98, p. 508, but see also Vibh. 175, p. 881, on the &uddhavasikas and 177, p. 889, on the number of the laksanas); Purnasa {Ko/a Hi. 28, Vibh. 23, p. 118b: Samadatta (? ), iii. 45, Vibh. 118c); Samghavasu {Vibh. 19, p. 97a, 106, p. 547a; 142, p. 732a) who recognizes only six indriyas in the absolute sense, the
jivita and the eight, eye, etc. , because these six are the root of being, sattvamula (Kola, ii. 5); and, with respect to this, Kusavarman, who only admits one indriya, the manas, a doctrine which leans towards the Vijnanavada.
48 a. Vasumitra.
1. Vasumitra is one of the great masters of the Vibhdsd, and one of the leaders of the Sarvastivadin school. His theory on "the existence of all" Is, Vasubandhu says, preferable to that of the three other masters, Dharmatrata, Ghosaka, and Buddhadeva {Kosa, v. 26).
One searches in vain in the two Abhidharmas (of the collection of seven) attributed to Vasumitra, the Prakaranapdda and the Dhdtukdya, for an allusion to this theory. Taranatha says, moreover, that the author of the Prakarana has nothing in common with the Vasumitra of the Vibhdsd (p. 68).
2. The Aryan Vasumitra Bodhisattva gives his name to a treatise (Taisho 1549). According to the preface, this was the Vasumitra who, after Maitreya, will be the Simhatathagata; the Vasumitra to whom the fathers refused entrance to the Council because he was not an arhat, and who later became the president of the Council {Hiuan-tsang, Watters, i. 271). Watters does not think that he is the great master of the Vibhdsd', indeed, the thirteenth chapter of Taisho 1549, entitled "Sarvdsti-akhanda," does not contain any references to the system of avasthdnya- thdtva of the Bhadanta Vasumitra. This is all that I dare say about this very complicated chapter.
Poussin 29
? 30 Introduction
The theory of the time periods is encountered in the second volume (p.
