The sloka on the eight
aniyatas
(Kosa, ii.
Abhidharmakosabhasyam-Vol-1-Vasubandhu-Poussin-Pruden-1991
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. 780b), where the following text is discussed: "The past and the future are impermanent, and even more so, the present. " Why does the Bhagavat say "And even more so, the present"? Six explanations follow (among which the fifth: "In former times the length of life was 80,000 years; it will again become 80,000 years"); then: "The Bhadanta says, The present appears for a short period of time; the past and the future do not remain permanently, but they come and go reciprocally. This is what conforms to the sutra. '"
The paragraph devoted to avidya (p. 722) does not formulate the opinion of the author. There is only "It is said," notably the opinion of the Mahisasakas. Is this ajndna, the five nivaranas, ayonisomanaskdra, viparydsa, etc. ? (See Kosa, iii. 28). It appears, from the silence of Kyokuga Saeki, the editor of the Kosa, that the Vibhdsd does not treat this point.
The problem of alcohol is treated on p. 786 (Kosa, iv. 34, Vibhdsd, jp. 645).
The discussion on labha and bhdvand (Kosa, vii. 63), in which Vasumitra takes part (according to the Vydkhyd,) should be referred to Taisho 1549, for Vasumitra is not named in Vibhdsd, p. 554b. The same remark applies to the erroneous opinion of Vasumitra on the falling away from the nirvedhabhdgryas, Kosa, vi. 21.
The sloka on the eight aniyatas (Kosa, ii. 27) is not found in Taisho 1549.
For the discussion, "Does it happen that the dharma which is hetupratyaya of a dharma is not hetupratyaya of this dharma? '\ see Taisho 1549, p. 791a, and compare the Jndnaprasthdna in Kosa, ii. 52.
On living longer than a kalpa, see p. 782b; manodanda, schism, p. 785, classical doctrines.
9
3. Vasubandhu (ii. 44) quotes the Pariprcchd,* and has also written a
Pancavastuka (Vydkhyd). The Pariprcchd teaches a doctrine which is clearly
Sautrantika (that nirodhasamdpatti is accompanied by a subtle mind). Also, K'uei-
chi (Siddhi, 211) says that this Vasumitra is a divergent Sautrantika master. 50
As for the Pancavastuka, we possess a commentary, the Pancavastukavi-
bhdsa, from the hand of Dharmatrata (Taisho 1555). The five vastus are the
vastus explained in Kosa, ii. 55-56 (svabhdvavastu, dlambanavastu . . . ).
51
This work does not appear to contain Sautrantika opinions;
it is divided into
three chapters, Rupavibhdga, 'Cittavibhdga, and Caittavibhdga.
Vasubandhu adopts the demonstration of "seeing by the two eyes" through the argument of seeing the two moons (Kosa, i. 43, Pancavastukavibhdsd, p. 991c), and he probably adopts the theory (i. 38) that the five vijndnas are both of retribution and out-flowing and that the sixth is also ksanika (Pancavastuka,
p. 933c).
? The demonstration of the existence of sukha {Kosa, vi. 3) is very similar to the demonstration established in the Pancavastuka, p. 994c.
K'uei-chi, in his treatise on the sects (Sarvastivadin thesis, 28), mentions the opinion of the Pancavastuka on the nature of the svalaksana which is the object of the vijndnas (Kosa, i. 10).
b. Ghosaka and the Abhidharmdmrtasastra.
After the Council and the death of Kaniska, a Tho-gar or Tukhara was invited with Vasumitra to the country of A? maparanta--to the west of Kasmlr and close to Tukhara (TSranatha, 61)--by its ruler; he was the proponent of a theory that "all exists," and is frequently quoted in the Vibhdsd', he was also the author of the Abhidharmamamrtasdstra (Taisho 1553).
This treatise is a truly exquisite, small book, very readable (in spite of the early date of its translation, 220-265), very complete (for example, Chap, vi, on the doctrine of the laksanas and sub-faksanas, Kosa, ii. 45), but concise; however, we find some well-chosen details (for example, the enumeration of the fields of merit: father, mother, an old person, a sick person . . . ).
The list of the viprayuktas, p. 970, is related to that of the Prakarana (Kosa, ii. 35-36a): prdpti, jdti, sthiti, anityatd, asamjnisamdpatti, nirodhasamdpatti, asamjni-dyatana, ndnddesaprdpti {? ), vastuprdpti (? ), dyatanaprdpti (? ), the three kdyas, prthagjanatva.
The Ko? a (ii. 44) reproduces the essentials of a discussion between Ghosaka and the Sautrantika Vasumitra, the author of the Pariprcchd (see above p. 30), on the existence of the mind in the absorption of nirodha. The Vibhdsd, it appears, ignores the author of the Pariprcchd.
Among the opinions of Ghosaka mentioned in the Vibhdsd and mentioned again in the Kosa, the most notable is that "visibles are seen by prajnd associated with the visual consciousness," an opinion that departs from orthodoxy {Vibhdsd, p. 61c, Kosa, 142).
52
Elsewhere Ghosaka is very orthodox, or his divergences, which are minimal,
indicate a progress; for example, Kosa, vi. 19, 20, 78. The references iv. 4, 79 and \. 66 merit examination.
Vibhdsd, p. 397b, is interesting: "Ghosaka says: The five skandhas which form part of one's own series, of the series of another, which belong to living beings and which do not belong to living beings, are Suffering and the Truth of Suffering. The ascetic, upon understanding (abhisamaya), sees only that the five skandhas of his own series are suffering; he does not see that the others are
Poussin 31
? 32 Introduction
suffering. Why is this? Because one understands suffering under the aspect of torment : now the skandhas of another's series . . . do not torment his own series. "
c. Buddhadeva.
S. Levi asks (JA. 1896, 2, 450, compare Barnett, JRAS. 1913, 945) if the Buddhadeva of the Vibhdsd should be identified with the Aryan Buddhadeva, a Sarvastivadin, the Lion of Mathura. The Sarvastivadin Budhila mentions this same Lion who appears to be related to the Mahasarhghikas and who is perhaps
53
the Fo-t'i-lo of Hsiian-tsang,
cayasdstra) used by the Mahasarhghikas (? ) (Levi, ibid. ; Watters, i. 82).
We will find in Konow (Kharoshtht Inscriptions, 44-49) the most recent remarks on these difficult inscriptions. There is nothing wrong with Buddhadeva being very much earlier than the Vibhdsd. The Sarvastivadins owe their name to the theory that "all exists," which Buddhadeva was probably one of the first to have explained.
In addition to his theory that "all exists" {Kosa, v. 26), Buddhadeva is unique in
maintaining that derived matter (bhautika) is only a mode (avasthd) of primary
matter (mahdbhutas) (i. 64), and that the mental states {caittas, sensation, ideas,
volition) are only modes of the mind {citta, vijndna) (ii. 23, ix; Siddhi, 395,
Vibhdsd, p. 66lc, p. 730b), a doctrine which connects Buddhadeva to Dharmatrata
54
There are at least two Dharmatratas:
1. The Bodhisattva who compiled the Uddnavarga. The Vibhdsd, followed by the Kosa, quotes it to show that a work can be the "word of the Buddha" even though edited by a master. According to the preface to Taisho 212 (A. D. 399), this Dharmatrata was the maternal uncle of Vasumitra (Chavannes, Cinq cents contes, iii. 297).
2. The master quoted in the Vibhdsd, the proponent of a theory of "all exists" {Kosa, v. 26) which appeared to the Vaibhasikas to be too close to the systems of the heterodox.
3. This master of the Vibhdsd is also the author of an Abhidharmasdra which bears his name ? (Taisho 1552), a commentary and a new edition of the Abhidharmasdra of Dharmasri (Taisho 1550). In fact, the Dharmatrata of the Vibhdsd (p. 383b) denies the rupa which forms part of the dharmdyatana, that is
and to the Darstantikas-Sautrantikas.
d. Dharmatrata.
the author of the Chi-chin-lun (Tattvasamuc-
? to say, avijnatirupa; the same negation, somewhat more involved, but clear nevertheless, appears in the Sara (chapter on action, p. 888, see below).
4. There is nothing to prevent this same Dharmatrata from being
the commentator on the Pancavastuka of Vasumitra, Taisho 1555, the 55
e. The Bhadanta Dharmatrata.
The Vibhdsd, it appears, ignores Kumaralata and Srilata, who were the heads
of the Sautrantika school (Siddhi, 221, told to us by K'uei-chi). The Sautrantika
school, or, more exactly, the school which should take the name of Sautrantika, is
represented in the Vibhdsd by the Darstantikas and by two masters: Dharmatrata,
a divergent Sarvastivadin, and the master whom the Vibhdsd simply calls
"Bhadanta," whom the Vydkhyd of the Kosa calls "the Sautrantika Bhadanta"
(Kosa, viii. 9), who is at the "head of the list of Sautrantikas" (viii. 40), and who
56
adheres to or leans toward the Sautrantika system (i. 20).
Hsiian-tsang, P'u-kuang, and Bhagavadvisesa recognize on occasion, in the
"Bhadanta" of the Vibhdsd, the Sthavira or Bhadanta Dharmatrata (Kosa, i. 20, iv. 4). Y as'omitra declares that Bhagavadvisesa is wrong: "Bhadanta is the philosopher that the Vibhdsd calls simply by the name of Bhadanta, a philosopher who adheres to the Sautrantika system or leans toward this system; whereas Dharmatrata, whom the Vibhdsd calls by his name, is a Sarvastivadin, the author of one of the four theories of the Sarvastivada (Kosa, i. 20). " On the other hand, "the first version of the Vibhdsd gives the name of this master [Bhadanta] in transcription and precedes this, like those of the other masters, with the title of venerable" (note of N. Peri, in Cosmologie bouddhique, 276).
Dharmatrata expresses some opinions on important points which clearly depart from the system of the Vibhdsd and from the orthodox Sarvastivadin system; the same may be said of Buddhadeva.
1. The Bhadanta does not admit that the eye sees: it is the visual consciousness that sees (i. 42); he has a particular theory on the non-contact of atoms (Wassiliew,
57
Pancavastukavibhdsd.
279), which Vasubandhu accepts and which Samghabhadra discusses (i. 43); admits three caittas (ii. 23, Add. ), a position which distinguishes him from the Darstantikas; like the Sarvastivadins, he denies that rupa is samanantarapratyaya (ii. 62); he admits the prolonged existence of antardbhava, against the Sarvasti- vadins (iii. 14); he has a particular opinion on pratityasamutpdda and samutpanna (iii. 28); he denies avijnapti, which is clearly anti-Sarvastivadin (iv. 4); on the four modes of kusala, he has a very orthodox doctrine (iv. 8); he holds to the "mortal" sin of intelligent animals (iv. 97): he gives an explanation of the word vimoksa
Poussin 33
he
? 34 Introduction
(viii. 33); he mixes the dhydnas and the suddhdvdsikas (Vibhdsd, p. 881c); he treats
the last thought of the arhat (p. 954a); on uccheda and faivata (p. 1003c); on the
meaning of alpa, sulabha, anavadya, and on the praise of his disciples by the 58
Buddha (p. 909a, and p. 900b, where he differs from Vasumitra).
59
The Bhadanta is very clear on vicdra-vitarka {Vibhdsd, p. 744b, and p. 269,
Kofa, ii. 33 and viii. 23): the author of the Jndnaprasthdna wants to refute what the Darstantika says. The latter says: "There is vitarka-vicddra from Kamadhatu up to Bhavagra. Why is this? Because the Sutra says that grossness of mind is vitarka and that subtlety of mind is vicdra: now grossness and subtlety of mind exist up to Bhavagra. " The Bhadanta says: "The masters of the Abhidharma say that vitarka- vicdra are grossness and subtlety of mind Now grossness and subtlety are relative things and exist up to Bhavagra. However, these masters only admit vitarka and vicdra in Kamadhatu and in Brahmaloka. This is poorly said, this is not well said. " The masters of the Abhidharma say, "What we say is well said, not poorly said. In fact. . . "
2. Vasubandhu {Kosa, vii. 31) attributes to the Bhadanta Dharmatrata (see the correction ad vii. 31) an opinion on the power of the Bhagavat which is an opinion of the Bhadanta according to Vibhdsd, p. 155c.
In the Vibhdsd (p. 61c) Dharmatrata says that visible matter is seen by the visual consciousness {caksurvijndna): an opinion that the Kosa attributes to a Vijnanavadin (i. 42), and which differs from that of the Bhadanta {Vibhdsd, p. 62b) who says that the eye sees by reason of light and the manas knows by reason of the act of attention.
3. Vibhdsd) p. 66lcl6: the Jndnaprasthdna wants to refute what the other masters say. In this school there are two masters, the first Buddhadeva and the second Dharmatrata.
Buddhadeva says that rupa is only the four mahdbhutas, that the caittas are citta. Derived rupa (updddyarupa) is only mahdbhutavi? esa\ the caittas are only cittavisesa. Sutras quoted in support of this theory: 1. "What is in the eye is solid . . . " (Kola, i. 35); 2. "Sarrufc/^* is cittaikdgrya . . . " (viii. 2). How does Buddhadeva prove the existence of dhdtus, dyatanas, skandhas? . . . The Abhi- dharmacaryas say, "The quoted sutras do not have this meaning . . . "
Dharmatrata admits the existence of derived rupa apart from primary rupa, and of the caittas apart from citta. But he holds that derived tangibles and the rupa which forms part of the dharmdyatana do not exist. He thus attempts to prove the existence of the dhdtus, dyatanas, skandhas, as does the Abhidharma system. But the derived tangibles exist separately, as do the other derived rupas; but if the rupa of the dharmdyatana does not exist, avijnapti does not exist {Kosa, i. 35, iv. 3).
? 4. Vibhdsd, p. 383b: The Abhidharma says, "What is rupaskandba?
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. 780b), where the following text is discussed: "The past and the future are impermanent, and even more so, the present. " Why does the Bhagavat say "And even more so, the present"? Six explanations follow (among which the fifth: "In former times the length of life was 80,000 years; it will again become 80,000 years"); then: "The Bhadanta says, The present appears for a short period of time; the past and the future do not remain permanently, but they come and go reciprocally. This is what conforms to the sutra. '"
The paragraph devoted to avidya (p. 722) does not formulate the opinion of the author. There is only "It is said," notably the opinion of the Mahisasakas. Is this ajndna, the five nivaranas, ayonisomanaskdra, viparydsa, etc. ? (See Kosa, iii. 28). It appears, from the silence of Kyokuga Saeki, the editor of the Kosa, that the Vibhdsd does not treat this point.
The problem of alcohol is treated on p. 786 (Kosa, iv. 34, Vibhdsd, jp. 645).
The discussion on labha and bhdvand (Kosa, vii. 63), in which Vasumitra takes part (according to the Vydkhyd,) should be referred to Taisho 1549, for Vasumitra is not named in Vibhdsd, p. 554b. The same remark applies to the erroneous opinion of Vasumitra on the falling away from the nirvedhabhdgryas, Kosa, vi. 21.
The sloka on the eight aniyatas (Kosa, ii. 27) is not found in Taisho 1549.
For the discussion, "Does it happen that the dharma which is hetupratyaya of a dharma is not hetupratyaya of this dharma? '\ see Taisho 1549, p. 791a, and compare the Jndnaprasthdna in Kosa, ii. 52.
On living longer than a kalpa, see p. 782b; manodanda, schism, p. 785, classical doctrines.
9
3. Vasubandhu (ii. 44) quotes the Pariprcchd,* and has also written a
Pancavastuka (Vydkhyd). The Pariprcchd teaches a doctrine which is clearly
Sautrantika (that nirodhasamdpatti is accompanied by a subtle mind). Also, K'uei-
chi (Siddhi, 211) says that this Vasumitra is a divergent Sautrantika master. 50
As for the Pancavastuka, we possess a commentary, the Pancavastukavi-
bhdsa, from the hand of Dharmatrata (Taisho 1555). The five vastus are the
vastus explained in Kosa, ii. 55-56 (svabhdvavastu, dlambanavastu . . . ).
51
This work does not appear to contain Sautrantika opinions;
it is divided into
three chapters, Rupavibhdga, 'Cittavibhdga, and Caittavibhdga.
Vasubandhu adopts the demonstration of "seeing by the two eyes" through the argument of seeing the two moons (Kosa, i. 43, Pancavastukavibhdsd, p. 991c), and he probably adopts the theory (i. 38) that the five vijndnas are both of retribution and out-flowing and that the sixth is also ksanika (Pancavastuka,
p. 933c).
? The demonstration of the existence of sukha {Kosa, vi. 3) is very similar to the demonstration established in the Pancavastuka, p. 994c.
K'uei-chi, in his treatise on the sects (Sarvastivadin thesis, 28), mentions the opinion of the Pancavastuka on the nature of the svalaksana which is the object of the vijndnas (Kosa, i. 10).
b. Ghosaka and the Abhidharmdmrtasastra.
After the Council and the death of Kaniska, a Tho-gar or Tukhara was invited with Vasumitra to the country of A? maparanta--to the west of Kasmlr and close to Tukhara (TSranatha, 61)--by its ruler; he was the proponent of a theory that "all exists," and is frequently quoted in the Vibhdsd', he was also the author of the Abhidharmamamrtasdstra (Taisho 1553).
This treatise is a truly exquisite, small book, very readable (in spite of the early date of its translation, 220-265), very complete (for example, Chap, vi, on the doctrine of the laksanas and sub-faksanas, Kosa, ii. 45), but concise; however, we find some well-chosen details (for example, the enumeration of the fields of merit: father, mother, an old person, a sick person . . . ).
The list of the viprayuktas, p. 970, is related to that of the Prakarana (Kosa, ii. 35-36a): prdpti, jdti, sthiti, anityatd, asamjnisamdpatti, nirodhasamdpatti, asamjni-dyatana, ndnddesaprdpti {? ), vastuprdpti (? ), dyatanaprdpti (? ), the three kdyas, prthagjanatva.
The Ko? a (ii. 44) reproduces the essentials of a discussion between Ghosaka and the Sautrantika Vasumitra, the author of the Pariprcchd (see above p. 30), on the existence of the mind in the absorption of nirodha. The Vibhdsd, it appears, ignores the author of the Pariprcchd.
Among the opinions of Ghosaka mentioned in the Vibhdsd and mentioned again in the Kosa, the most notable is that "visibles are seen by prajnd associated with the visual consciousness," an opinion that departs from orthodoxy {Vibhdsd, p. 61c, Kosa, 142).
52
Elsewhere Ghosaka is very orthodox, or his divergences, which are minimal,
indicate a progress; for example, Kosa, vi. 19, 20, 78. The references iv. 4, 79 and \. 66 merit examination.
Vibhdsd, p. 397b, is interesting: "Ghosaka says: The five skandhas which form part of one's own series, of the series of another, which belong to living beings and which do not belong to living beings, are Suffering and the Truth of Suffering. The ascetic, upon understanding (abhisamaya), sees only that the five skandhas of his own series are suffering; he does not see that the others are
Poussin 31
? 32 Introduction
suffering. Why is this? Because one understands suffering under the aspect of torment : now the skandhas of another's series . . . do not torment his own series. "
c. Buddhadeva.
S. Levi asks (JA. 1896, 2, 450, compare Barnett, JRAS. 1913, 945) if the Buddhadeva of the Vibhdsd should be identified with the Aryan Buddhadeva, a Sarvastivadin, the Lion of Mathura. The Sarvastivadin Budhila mentions this same Lion who appears to be related to the Mahasarhghikas and who is perhaps
53
the Fo-t'i-lo of Hsiian-tsang,
cayasdstra) used by the Mahasarhghikas (? ) (Levi, ibid. ; Watters, i. 82).
We will find in Konow (Kharoshtht Inscriptions, 44-49) the most recent remarks on these difficult inscriptions. There is nothing wrong with Buddhadeva being very much earlier than the Vibhdsd. The Sarvastivadins owe their name to the theory that "all exists," which Buddhadeva was probably one of the first to have explained.
In addition to his theory that "all exists" {Kosa, v. 26), Buddhadeva is unique in
maintaining that derived matter (bhautika) is only a mode (avasthd) of primary
matter (mahdbhutas) (i. 64), and that the mental states {caittas, sensation, ideas,
volition) are only modes of the mind {citta, vijndna) (ii. 23, ix; Siddhi, 395,
Vibhdsd, p. 66lc, p. 730b), a doctrine which connects Buddhadeva to Dharmatrata
54
There are at least two Dharmatratas:
1. The Bodhisattva who compiled the Uddnavarga. The Vibhdsd, followed by the Kosa, quotes it to show that a work can be the "word of the Buddha" even though edited by a master. According to the preface to Taisho 212 (A. D. 399), this Dharmatrata was the maternal uncle of Vasumitra (Chavannes, Cinq cents contes, iii. 297).
2. The master quoted in the Vibhdsd, the proponent of a theory of "all exists" {Kosa, v. 26) which appeared to the Vaibhasikas to be too close to the systems of the heterodox.
3. This master of the Vibhdsd is also the author of an Abhidharmasdra which bears his name ? (Taisho 1552), a commentary and a new edition of the Abhidharmasdra of Dharmasri (Taisho 1550). In fact, the Dharmatrata of the Vibhdsd (p. 383b) denies the rupa which forms part of the dharmdyatana, that is
and to the Darstantikas-Sautrantikas.
d. Dharmatrata.
the author of the Chi-chin-lun (Tattvasamuc-
? to say, avijnatirupa; the same negation, somewhat more involved, but clear nevertheless, appears in the Sara (chapter on action, p. 888, see below).
4. There is nothing to prevent this same Dharmatrata from being
the commentator on the Pancavastuka of Vasumitra, Taisho 1555, the 55
e. The Bhadanta Dharmatrata.
The Vibhdsd, it appears, ignores Kumaralata and Srilata, who were the heads
of the Sautrantika school (Siddhi, 221, told to us by K'uei-chi). The Sautrantika
school, or, more exactly, the school which should take the name of Sautrantika, is
represented in the Vibhdsd by the Darstantikas and by two masters: Dharmatrata,
a divergent Sarvastivadin, and the master whom the Vibhdsd simply calls
"Bhadanta," whom the Vydkhyd of the Kosa calls "the Sautrantika Bhadanta"
(Kosa, viii. 9), who is at the "head of the list of Sautrantikas" (viii. 40), and who
56
adheres to or leans toward the Sautrantika system (i. 20).
Hsiian-tsang, P'u-kuang, and Bhagavadvisesa recognize on occasion, in the
"Bhadanta" of the Vibhdsd, the Sthavira or Bhadanta Dharmatrata (Kosa, i. 20, iv. 4). Y as'omitra declares that Bhagavadvisesa is wrong: "Bhadanta is the philosopher that the Vibhdsd calls simply by the name of Bhadanta, a philosopher who adheres to the Sautrantika system or leans toward this system; whereas Dharmatrata, whom the Vibhdsd calls by his name, is a Sarvastivadin, the author of one of the four theories of the Sarvastivada (Kosa, i. 20). " On the other hand, "the first version of the Vibhdsd gives the name of this master [Bhadanta] in transcription and precedes this, like those of the other masters, with the title of venerable" (note of N. Peri, in Cosmologie bouddhique, 276).
Dharmatrata expresses some opinions on important points which clearly depart from the system of the Vibhdsd and from the orthodox Sarvastivadin system; the same may be said of Buddhadeva.
1. The Bhadanta does not admit that the eye sees: it is the visual consciousness that sees (i. 42); he has a particular theory on the non-contact of atoms (Wassiliew,
57
Pancavastukavibhdsd.
279), which Vasubandhu accepts and which Samghabhadra discusses (i. 43); admits three caittas (ii. 23, Add. ), a position which distinguishes him from the Darstantikas; like the Sarvastivadins, he denies that rupa is samanantarapratyaya (ii. 62); he admits the prolonged existence of antardbhava, against the Sarvasti- vadins (iii. 14); he has a particular opinion on pratityasamutpdda and samutpanna (iii. 28); he denies avijnapti, which is clearly anti-Sarvastivadin (iv. 4); on the four modes of kusala, he has a very orthodox doctrine (iv. 8); he holds to the "mortal" sin of intelligent animals (iv. 97): he gives an explanation of the word vimoksa
Poussin 33
he
? 34 Introduction
(viii. 33); he mixes the dhydnas and the suddhdvdsikas (Vibhdsd, p. 881c); he treats
the last thought of the arhat (p. 954a); on uccheda and faivata (p. 1003c); on the
meaning of alpa, sulabha, anavadya, and on the praise of his disciples by the 58
Buddha (p. 909a, and p. 900b, where he differs from Vasumitra).
59
The Bhadanta is very clear on vicdra-vitarka {Vibhdsd, p. 744b, and p. 269,
Kofa, ii. 33 and viii. 23): the author of the Jndnaprasthdna wants to refute what the Darstantika says. The latter says: "There is vitarka-vicddra from Kamadhatu up to Bhavagra. Why is this? Because the Sutra says that grossness of mind is vitarka and that subtlety of mind is vicdra: now grossness and subtlety of mind exist up to Bhavagra. " The Bhadanta says: "The masters of the Abhidharma say that vitarka- vicdra are grossness and subtlety of mind Now grossness and subtlety are relative things and exist up to Bhavagra. However, these masters only admit vitarka and vicdra in Kamadhatu and in Brahmaloka. This is poorly said, this is not well said. " The masters of the Abhidharma say, "What we say is well said, not poorly said. In fact. . . "
2. Vasubandhu {Kosa, vii. 31) attributes to the Bhadanta Dharmatrata (see the correction ad vii. 31) an opinion on the power of the Bhagavat which is an opinion of the Bhadanta according to Vibhdsd, p. 155c.
In the Vibhdsd (p. 61c) Dharmatrata says that visible matter is seen by the visual consciousness {caksurvijndna): an opinion that the Kosa attributes to a Vijnanavadin (i. 42), and which differs from that of the Bhadanta {Vibhdsd, p. 62b) who says that the eye sees by reason of light and the manas knows by reason of the act of attention.
3. Vibhdsd) p. 66lcl6: the Jndnaprasthdna wants to refute what the other masters say. In this school there are two masters, the first Buddhadeva and the second Dharmatrata.
Buddhadeva says that rupa is only the four mahdbhutas, that the caittas are citta. Derived rupa (updddyarupa) is only mahdbhutavi? esa\ the caittas are only cittavisesa. Sutras quoted in support of this theory: 1. "What is in the eye is solid . . . " (Kola, i. 35); 2. "Sarrufc/^* is cittaikdgrya . . . " (viii. 2). How does Buddhadeva prove the existence of dhdtus, dyatanas, skandhas? . . . The Abhi- dharmacaryas say, "The quoted sutras do not have this meaning . . . "
Dharmatrata admits the existence of derived rupa apart from primary rupa, and of the caittas apart from citta. But he holds that derived tangibles and the rupa which forms part of the dharmdyatana do not exist. He thus attempts to prove the existence of the dhdtus, dyatanas, skandhas, as does the Abhidharma system. But the derived tangibles exist separately, as do the other derived rupas; but if the rupa of the dharmdyatana does not exist, avijnapti does not exist {Kosa, i. 35, iv. 3).
? 4. Vibhdsd, p. 383b: The Abhidharma says, "What is rupaskandba?
