Pramāṇa · Search · About · API

Early Monastic Buddhism v2.c35.s3.t4: Language and Literature

Early Monastic Buddhism v2.c35.s3.t4

Language
English

Language and Literature

v2.c35.s3.t4.1

The Tibetan traditions corroborated by the recent finds of manuscripts in Eastern Turkestan leave no room for doubt about the fact that the Sarvāstivādins adopted grammatical Sanskrit (and not mixed Sanskrit) as the medium of their literature and that they possessed a complete canon of their own in three divisions Sūtra, Vinaya and Abhidharma. The sub-divisions of these three piṭakas were also substantially the same as those in Pāli.

v2.c35.s3.t4.2

Our main source of information regarding the literature of this school is the Chinese and occasionally Tibetan versions of the Tripiṭaka, supplemented by the find of manuscripts in Central Asia, Eastern Turkestan, Gilgit and Nepal, and by quotations found in works like the Lalitavistara, Mahāvastu, Madhyamaka-vṛtti, Sūtrālaṅkāra of Asaṅga, Divyāvadāna, Abhidharmakośa with its Bhāṣya and Vyākhyā. It may be questioned whether the informations available about the literature of this school are of the Sarvāstivādins or of the Mūla-Sarvāstivādins. For the present it is not possible to distinguish between the two.

v2.c35.s3.t4.3

but it seems that the Āgamas were common to both, so also were the Abhidharma texts. It is only in regard to Vinaya and a few Avadāna texts there might have some differences.

v2.c35.s3.t4.4

SŪTRAS: The Sūtra-Piṭaka of the Sarvāstivādins was divided into Āgamas corresponding to Nikāyas of the Pāli school. There were four Āgamas called Dīrgha, Madhyama, Saṃyukta and Ekottara. In the Kośa there are references to the Kṣudraka implying thereby the existence of a Kṣudrakāgama too. Prof. Akanuma has compared the Āgamas in Chinese with the Pāli Nikāyas' in detail and has come to the following conclusions: The Dīrghāgama contains 30 Sūtras as against 34 of the Dīgha Nikāya. Of the 13 suttas in the first volume of the Pāli Dīgha Nikāya, 3 only are omitted in the Dīrghāgama viz., Mahāli (no. 6), Jāliya (no. 7) and Subba (no. 10). All the suttas of the other two volumes are contained in this Āgama with a few in the Madhyamāgama. The order of arrangement of the sūtras in the Āgamas and Nikāyas differs widely, e.g., Mahāpadāna is the first sutta in the Āgama in place of Brahmajāla of the Nikāya. In the Āgama the order of sūtras is as follows: Mahāpadāna, Mahāparinibbāna, Mahāgovinda, Janavasabha, Aggañña, Cakkavatti, Sīhanāda, Pāyāsi, Udumbarika-Sīhanāda, Saṅgīti, Dasuttara, Mahānidāna, Sakka-pañha, Pāṭika, Siṅgālovāda, Pāsādika, Sampasādanīya, Mahāsamaya, Ambaṭṭha, Brahmajāla, Sonadaṇḍa, Kūṭadanta, Kevaṭṭa, Kassapa-sīhanāda, Tevijja, Sāmaññaphala, Poṭṭhapāda, Lohicca. The Āgama contains two other suttas.

v2.c35.s3.t4.5

Of these suttas, fragments of the Āṭānāṭiya and Saṅgīti have been discovered in Eastern Turkestan² and quotations from the Brahmajāla and Saṅgīti appear in the Abhidharmakośa.

v2.c35.s3.t4.6

¹ The Comparative Catalogue of Chinese Āgamas & Pāli Nikāyas. Japan. (1929).

v2.c35.s3.t4.7

² Hoernle, Manuscript Remains of Buddhist literature found in Eastern Turkestan.

v2.c35.s3.t4.8

The relation of the sūtras of the *Madhyamāgama* to those of the *Majjhima Nikāya* is similar to the *Dīgha*. Of the 152 suttas in the three volumes of the *Majjhima*, only 19 suttas are omitted in the Āgama, viz., *Cūḷasāropama* (no. 30), *Mahāsaccaka* (no. 36), *Sāleyyaka* (no. 41), *Verañjaka* (no. 42), *Kandaraka* (no. 51), *Jīvaka* (no. 55), *Kukkuravatika* (no. 57), *Abhayarājakumāra* (no. 58), *Apaṇṇaka* (no. 60), *Tevijja-Vacchagotta* (no. 71), *Ghoṭamukha* (no. 94), *Caṅkī* (no. 95), *Vāseṭṭha* (no. 98), *Saṅgārava* (no. 100), *Pañcattaya* (no. 102), *Kiṃtī* (no. 103), *Sunakkhatta* (no. 105), *Anupada* (no. 111) and *Bhaddekaratta* (no. 131). In the *Madhyamāgama*, there are in all 222 sūtras, 82 of which correspond to the suttas in the *Aṅguttara*, 10 to suttas in the *Saṃyutta*, 9 to those in the *Dīgha* and rest to suttas in the *Majjhima*. There are a few of these suttas in Pāli not found in the Āgama., while a few stray suttas correspond to passages in the *Suttanipāta*, *Thera-theri-gāthā* and Vinaya (*Mahāvagga*). In view of the mixture of the suttas from two or three Nikāyas in this Āgama we can hardly expect much agreement in the order of the arrangement of the sūtras.

v2.c35.s3.t4.9

Fragments of two sūtras of the *Madhyamāgama*, viz., *Upāli* and *Suka*, have been discovered in Eastern Turkestan.¹

v2.c35.s3.t4.10

¹ Hoernle, *op. cit.*

v2.c35.s3.t4.11

The agreement between the *Saṃyukta-Āgama* and *Saṃyutta Nikāya* is similar to that of the *Madhyamāgama* and *Majjhima Nikāya*. The *Sagāthavagga* (Sec. I) of the two piṭakas has much in common but not the *Nidānavagga* (Sec. II); the 8th and 9th chapters of *Nidāna*, viz., *Samaṇa-brāhmaṇa* and *Antarapeyyāla* are wanting in the Āgama, while the 1st and 5th chapters (*Buddha* and *Gahapati*) show marked differences. In the same section, *Abhisamaya*, *Dhātu* and other saṃyuttas are almost passed over in the Āgama, but there is much commonness in the following five Saṃyuttas: *Anamatagga*, *Kassapa*, *Lakkhana*, *Opamma* and *Bhikkhu*. In the *Khandha-vagga* (Section III) of the Āgama, the following saṃyuttas are wanting: *Okkan-tika*, *Uppāda*, *Kilesa*, *Sāriputta*, *Nāga*, *Gandhabbakāya*, *Valāha*, *Vacchagotta* and *Jhāna*. In the *Salāyatana-vagga* (Section IV), the following are absent: *Mātugāma*, *Moggallāna*, *Asaṅkhata*, *Sammappadhāna*, *Bāla* and *Iddhipāda*, while major portions of the *Magga*, *Indriya* and *Sacca* omitted.

v2.c35.s3.t4.12

The *Saṃyuktāgama*, as it exists in Chinese, is divided into 50 sections and incorporates a large number of suttas of the *Aṅguttara Nikāya* and a few of the other texts. There are also a few sūtras which have no parallels in Pāli.

v2.c35.s3.t4.13

A fragment of the *Śroṇasūtra* of this Āgama has been discovered in Eastern Turkestan while Prof. Sylvain Lévi traced a few quotations from this Āgama in the *Sūtrālaṃkāra* of Asaṅga,¹ and identified the following fragments from the collection of Grünwedel: *Kokanada-sūtra* (= *Aṅguttara*, V, pp. 196-198), *Anāthapiṇḍada-sūtra* (= *Aṅguttara*, V, pp. 185-189); *Dīrghana-kha sūtra* (= *Majjhima*, I, pp. 497-501), *Sarabha-sūtra* (= *Aṅguttara*, I, pp. 185-188); *Pravrājaka Sthavirasūtra* and *Brāhmaṇasatyāni sūtra* (= *Aṅguttara*, II, p. 185) — all included in the Chinese translation of the *Saṃyuktāgama*.²

v2.c35.s3.t4.14

¹ See Winternitz. p. 234 fn.

v2.c35.s3.t4.15

² *Toung Pao*, V. p 299.

v2.c35.s3.t4.16

The *Ekottarāgama* and the *Aṅguttara Nikāya* have very little in common. This is partly due to the fact that a large number of the suttas of the *Aṅguttara* is included in the *Madhyama* and *Saṃyukta Āgamas*. The Pāli text is much more extensive than the Sanskrit and it seems that the growth of this part of the Piṭaka took place independently of each other. From Akanuma’s comparative studies, the following may be pointed out as being more or less common in the two piṭakas: *Samacitta* (I, pp. 61-9), *Devadūta* (I, pp. 132-150), *Brāhmaṇa* to *Loṇaphala* (I, pp. 155-258), *Cakka* (II, pp. 32-44), *Muṇḍarāja* (III, pp. 45-62), *Nīvaraṇa* (III, pp. 63-79), *Āghāta* (III, pp. 185-202), *Devatā* to *Mahā* (III, pp. 329-420), *Avyākata* to *Mahā* (IV, pp. 67-139), *Gahapati* (IV, pp. 208-235), *Savitta* (V, pp. 92-112), *Upāsaka* (V, pp. 176-210), *Jānussoṇi* (V, pp. 249-273), and *Anussati* (V, pp. 328-358). This is not an exhaustive list, for there are stray agreements in other sections.

v2.c35.s3.t4.17

A fifth Āgama was not recognised by the schools other than the Theravāda. In the *Divyāvadāna* (pp. 17, 331, 333) and elsewhere the Āgamas are referred to as *Āgamacatuṣṭayam*. In the Nāgārjunikoṇḍa inscriptions also, four Nikāyas are spoken of and not five. The Pāli *Khuddaka Nikāya* is really a collection not of discourses, short or long, but of a number of independent works which could not be included in one of the four Nikāyas. By *Khuddaka*, the Pālists probably meant "other works" or "miscellaneous works." Though the Sarvāstivādins did not have a fifth Nikāya, they had a few texts like the *Udānavarga*, *Suttanipāta* (*Aṭṭhaka* and *Pārāyaṇa* vaggas), *Sthavira-gāthā*, *Dharmapada*, *Vimānavastu*, and *Buddhavamsa*, which came later on to be collectively called *Kṣudrakāgama* (see above, p. 125).

v2.c35.s3.t4.18

VINAYA: Our information about the Vinaya texts of the Sarvāstivādins is derived solely from the catalogues of Chinese canonical literature. In Nanjio's Catalogue, we came across the following titles:—

v2.c35.s3.t4.19

(i) *Sarvāstivāda-vinaya-mātṛkā*, translated by Saṅghavarman (445 A.D.): Ta'sho, xxii, 1441; Nanjio 1132.

v2.c35.s3.t4.20

(ii) *Sarvāstivāda-vinaya-vibhāṣā*, translator unknown (350-431 A.D.): Taisho xxiii, 1440; Nanjio 1135, 1136.

v2.c35.s3.t4.21

(iii) *Sarvāstivāda-vinaya-saṅgraha*, compiled by the venerable Jinamitra, translated by I-tsing (700 A.D.); Nanjio 1127.

v2.c35.s3.t4.22

(iv) *Daśādhyāya-vinaya-nidāna*, translated by Vimalākṣa (being the preface to the *Daśādhyāya-vinaya*), Nanjio 1144.

v2.c35.s3.t4.23

(v) *Daśādhyāya-vinaya-bhikṣu-prātimokṣa*, translated by Kumārajīva (404 A.D.): Taisho xxiii, 1436; Nanjio 1160.

v2.c35.s3.t4.24

(vi) *Daśādhyāya-vinaya-bhikṣuṇī-prātimokṣa*, compiled by Fa-yin (420-479 A.D.): Taisho xxiii, 1437; Nanjio 1161.

v2.c35.s3.t4.25

(vii) *Daśādhyāya-vinaya* or the Sarvāstivāda Vinaya, translated by Punyatara together with Kumārajīva (404 A.D.): Taisho xxiii, 1435; Nanjio 1115.

v2.c35.s3.t4.26

The principal text of the Sarvāstivādins was the *Daśādhyāya-vinaya*. Fa-hien writes that he came across a Sarvāstivāda-vinaya in verses, but the Chinese translation of the *Daśādhyāya-vinaya* attributed to the Sarvāstivādins is in prose. The *Daśādhyāya* (Taisho ed., xxiii, 1435) is divided into 14 sections. It opens with the eight sections of the Prātimokṣa-sūtra. The ninth section deals with “seven dharmas” viz., *śikṣāpada, poṣadha, pāpadeśanā, varṣāvāsa, carmavastu, bhaiṣajyavastu* and *cīvara* (precepts, fortnightly ceremonies, confession, dwelling in the rainy season, use of leather-shoes, use of medicines, and robes of monks). The tenth section contains “eight dharmas”, viz., *Kaṭhina, Kauśāmbi, Campā, Pāṇḍulohitaka, Saṅghāvaśeṣaparivāsa, Paṭicchādana, Śayanāsana* and *Āsamudācārika-dharma* (rules re. making of Kaṭhina-robes, dispute at Kauśāmbi, events at Campā, deeds of Pāṇḍulohitaka-monks, atonement for Saṅghāvaśeṣa-offences, concealment of irreligious acts, rules regarding bed and seat, and rules regarding proper conduct of monks).¹ The eleventh section entitled "saṃyutta", i.e. miscellaneous rules, deals with the dhūta and other extra-ordinary precepts observed by monks. The twelfth section is devoted to Bhikṣuṇī-prātimokṣa containing, as it does, 8 Pārājikā, 17 Saṅghāvaśeṣa, 30 Naiḥsargikā, 78 Pāyantika, 8 Prātideśanīyā and Aṣṭa-dharmā. The thirteenth section re-arranges the preceding rules in the Ekottara style, from one to eleven dharmas. The concluding section, the fourteenth, contains the Upāli-paripṛcchā, a well-known text on disciplinary rules.²

v2.c35.s3.t4.27

¹ The titles of vastus (chapters) in the Mūlasarvāstivāda-vinaya are as follows: Pravrajyā, Poṣadha, Pravāraṇā, Varṣā, Carma, Bhaiṣajya, Cīvara, Kaṭhina, Kośāmbaka, Karma, Pāṇḍulohitaka, Pudgala, Pārivāsika, Poṣadhasthāpana and Saṃghabheda.

v2.c35.s3.t4.28

² See Bodhisattva-prātimokṣa-sūtra, Intro, p. 3 (IHQ., VII. 2).

v2.c35.s3.t4.29

This text contains almost all the chapters of the vinaya of the Theravādins and the Mūlasarvāstivādins,³ and apears to be a much shorter version of the text of the latter. From the title, one expects ten chapters, but actually there are fourteen, and so we have to assume that four of the fourteen chapters were later additions or were originally treated as supplements. The 11th, 13th and 14th chapters are no doubt later additions, but it is difficult to ascertain the fourth additional chapter. A close study of the Chinese translation along with the Sanskrit text of the Mūlasarvāstivādins will reveal the actual position.

v2.c35.s3.t4.30

³ For further details, see introduction to the Mūlasarvāstivāda-vinaya, Gilgit Manuscripts, vol. III, pt. ii.

v2.c35.s3.t4.31

ABHIDHARMA: The Abhidharma literature of the Sarvāstivādins is fairly extensive. Apart from the well-known seven texts belonging to this school and the famous Vibhāṣā-śāstras of the Vaibhāṣikas, this school had to its credit a few other philosophical works written by Vasubandhu, Saṃghabhadra, Dharmatrāta and Dharmottara. None of these valuable works are available in original Sanskrit except the *Abhidharmakośa-vyākhyā* of Yaśomitra. The *Vyākhyā* is no doubt a mine of information and contains most of the philosophical topics discussed in the Abhidharma literature of the Sarvāstivādins. It may also be regarded as a quintessence of the seven Abhidharma texts. For a general idea of the several texts at the present moment, we shall have to depend on the valuable analysis of the Chinese translations of the texts made by Prof. Takakusu in the *J.P.T.S.*, 1904-5 and the notes left by Prof. Louis de la Vallée Poussin in his introduction to the French translation of the *Abhidharmakośa*. With the publication of the *Vyākhyā* it has become possible to comprehend the terms and nomenclatures suggested by Takakusu on the basis of the Chinese renderings and form a better idea of the contents of the texts. The seven texts claimed by the Sarvāstivādins as constituting their original Abhidharmapiṭaka are as follows¹:—

v2.c35.s3.t4.32

¹ *Kośa*, I, 9 & 11.

v2.c35.s3.t4.33

(i) *Jñānaprasthānasūtra* of Ārya Kātyāyaṇī-putra with its six supplements, (*ṣaṭ pādāḥ*), viz.,
(ii) *Prakaraṇapāda* of Sthavira Vasumitra
(iii) *Vijñānakāya* of Sthavira Devaśarmā
(iv) *Dharmaskandha* of Ārya Śāriputra
(v) *Prajñaptiśāstra* of Ārya Maudgalyāyana
(vi) *Dhātukāya* of Pūrṇa and
(vii) *Saṅgīti-paryāya* of Mahākauṣṭhila.

v2.c35.s3.t4.34

(1) The *Jñānaprasthāna-sūtra* is attributed to Ārya Kātyāyaṇīputra. In the *Kośa* it is stated that the actual author of the work is Buddha but the arrangements of chapters and topics had been made by Kātyāyaṇīputra and so its authorship is attributed to him. It was twice translated into Chinese, by Gotama Saṃghadeva of Kashmir and Chu Fo-nien, in the 4th century A.D. and by Hiuen-tsang in the 7th century. It is divided into eight sections. The first section contains exposition of laukikā-gradharmas, jñāna, pudgala, śraddhā, ahrīkatā, rūpa and its lakṣaṇa, anartha (?), and caitasikas (=supramundane topics,¹ knowledge, individuality, faith and reverence, lack of modesty, material constituents of the body and their characteristics, anartha (?) and mental states). The second section details the saṃyojanas or defilements which hinder the spiritual progress of an adept, and the causes of defilements. The third section is devoted to the acquisition of the knowledge (jñāna) (a) of doctrinal matters by which a sekha becomes an asekha, (b) of right and wrong views, (c) of the means of attaining six abhijñās, (d) of the four truths and of the acquisitions to be made in the four stages of sanctification. The fourth section details the evil words and acts with their consequences as also explains the vijñaptis and avijñapti². The fifth section gives an exposition of rūpaskandha, i.e., the four great elements and of things originating out of them, both internal and external. The sixth section analyses the 22 indriyas (predominant faculties) and the three spheres of existence viz., kāma, rūpa and arūpa, and explains in detail the sparśendriya, mūla-citta, etc. The seventh section is devoted to the mental states developed by an adept while he is in samādhi, and gradually advances from Sakadāgāmi to Anāgāmi stage. The last, the eighth section explains the four smṛtyupasthānas, the various wrong views, and such other matters.

v2.c35.s3.t4.35

¹ Kośa (Fr transl), Intro, p. xxx. See Infra. p. 153-4

v2.c35.s3.t4.36

² Kośa, I, 11.

v2.c35.s3.t4.37

Dr. Barua suggests that this work may be paralleled with the Pāli text *Paṭisambhidāmagga*.¹ There may be a verbal resemblance between the two texts, but the *Jñānaprasthāna* is written more on the lines of Buddhaghosa’s *Visuddhimagga* than on the *Paṭisambhidāmagga*. The title also suggests that the work is expected to contain topics leading to the highest knowledge, which, in other words, is purity or emancipation.

v2.c35.s3.t4.38

¹ Law, *History of Pāli Lit.*, I, p. 337.

v2.c35.s3.t4.39

The second book is entitled *Prakaraṇapāda*. Its authorship is attributed to Sthavira Vasumitra, who, according to the Chinese tradition, composed it in a monastery, at Puṣkarāvatī. It was translated into Chinese by Guṇabhadra and Bodhiyaśas of Central India (A.D. 435-443) and also by Hiuen-tsang (A.D. 659). The work is divided into eight chapters. The first defines *rūpa*, *citta*, *caitasikas*, *cittaviprayuktas* and *asaṃskṛtas*² (material constituents, mind, mental states, non-mental states, and the unconstituted). The second deals with the same topics as those discussed in the last two chapters of the third section of the *Jñānaprasthānasūtra*. The third explains the sense-organs and their spheres of action while the fourth defines several terms, such as *dhātu*, *āyatana*, *skandha*, *mahābhūmikas*² (cf. *Kośa*, II, 23, III, 32), etc. The fifth chapter analyses the *anuśayas* (dormant passions), while the sixth touches the *vijñeya*, *anumeya* and *anāsrava dharmas* (things to be known and inferred, and pure dharmas). The concluding chapter, the seventh, appears to be an index, containing all the technical terms with their meanings in short.

v2.c35.s3.t4.40

² See *Infra*, pp 135, 141-2

v2.c35.s3.t4.41

The third book, *Vijñāna-kāya* is attributed to Devaśarmā, who, according to Hiuen-tsang, compiled it at Viśoka near Śrāvastī, about a century after Buddha’s death. It was translated into Chinese by Hiuen-tsang (649 A.D.). It is divided into six chapters. It contains an exposition of *pudgala*, *indriya*, *citta*, *kleśa*, *vijñāna*, etc. as given by Maudgalyāyana, enumerates the different classes of beings, persons, etc., defines the function of mental states as *hetu* (cause) and *ālambana* (basis) of spiritual progress¹ and also of mental states of a perfect and an imperfect adept.² Prof. La Vallée Poussin remarks in his *Etudes Asiatiques* 1925 (i. 343-376) that the first two chapters contain the controversies relating to the existence of past and future, and of *pudgala* (soul).³

v2.c35.s3.t4.42

¹ *Kośa*, ii, 61-62

v2.c35.s3.t4.43

² *Kośa*, vii, 12

v2.c35.s3.t4.44

³ Cf. *Infra*, p. 137 f.

v2.c35.s3.t4.45

The fourth book is entitled *Dharmaskandha*. Its authorship is attributed to Śāriputra. It was translated into Chinese by Hiuen-tsang (659 A.D.). In the colophon of the Chinese translation this text is described as "the most important of the Abhidharma works, and the fountain-head of the Sarvāstivāda system". This book it seems appealed to the Chinese not for its subtleness and depth of philosophical discussions but for its comprehensiveness as outlining the general course of spiritual training prescribed for a Buddhist monk. This work also can be paralleled to the *Visuddhimagga* of Buddhaghosa. Its 21 sections are as follows: *Śikṣāpadas* or *Śīlas*; attainments leading to Srotaāpatti; development of faith in the Triratna; the fruits of four stages of sanctification, four *ariya-pudgalas*; *samyak-samkalpa* of the eight-fold path; attainment of *ṛddhipādas*; practice of *smṛtyupasthānas*; exposition of the *āryasatyas*; four *dhyānas*; four *apramāṇas*; four higher *samāpattis* (*ārūpyas*), practice of *bhāvanā*; exposition of *bodhyaṅgas*, and then an exposition of *indriyas*, *āyatanas*, *skandhas* and *dhātus*. Its concluding chapter explains the twelve terms of the formulæ of causation (*pratītyasamutpāda*).

v2.c35.s3.t4.46

The fifth book, *Dhātukāya*, is attributed to Pūrṇa in the Sanskrit and Tibetan tradition, and to Vasumitra by the Chinese writers. Prof. Takakusu remarks that the original Sanskrit had probably more than one recension. It was translated into Chinese by Hiuen-tsang (663 A.D.). The object of the treatise is to enumerate the dharmas, considered as ‘reals’ by the Sarvāstivādins. The dharmas are classified under the heads: 10 *mahābhūmikas*, 10 *kleśa-mahābhūmikas*, 10 *parittakleśas*, 5 *kleśas*, 5 *dṛṣṭis*, 5 *dharmas*, etc. This classification differs slightly from that found in Pāli texts or the *Abhidharmakośa*.¹ Prof. La Vallée Poussin thinks that this must be a very old text, which may be regarded as the source of the Pāli *Dhātukathā* also, as it discusses the *sampayutta* and *vippayutta* relations of the dharmas as has been done in the *Dhātukathā*.

v2.c35.s3.t4.47

¹ Cf. *Kośa* (Transl.), ii, p. 150 In *See Infra* p 141-2.

v2.c35.s3.t4.48

The sixth book *Prajñaptiśāstra* is attributed to Maudgalyāyana. It was translated into Chinese at a very late date (1004-1055 A.D.) by Fa-nu (=Dharmapāla) of Magadha. The Chinese text is incomplete. In Tibetan version this treatise is divided into three parts, viz., *lokaprajñapti*, *kāraṇaprajñapti* and *karmaprajñapti*. The *lokaprajñapti* appears in a well-digested form in the *Abhidharmakośa* (III). Prof. La Vallée Poussin has analysed the first two *prajñaptis* in the *Cosmologie bouddhique* (pp. 275-350).² In the *lokaprajñapti* the cosmological ideas of the Buddhists are given, in the *kāraṇaprajñapti* the characteristics that make a Bodhisattva are discussed while in the *karmaprajñapti* there is an enumeration and classification of different kinds of deeds.

v2.c35.s3.t4.49

² *Kośa*, Intro. p. xxxvii ff.

v2.c35.s3.t4.50

The seventh book *Saṅgītiparyāya* is attributed to Mahākauṣthila by Yaśomitra and Buston, and to Śāriputra by the Chinese writers. It was translated into Chinese by Hiuen-tsang (660-663 A.D.). This text was compiled, according to the introductory remarks, immediately after Buddha's death to avert disputes among the disciples regarding the Buddhist teachings and disciplinary rules. The scene of this text is laid at Pāvā, where the dissensions among the Nigaṇṭha Nāṭaputtas started after the death of their teacher. It arranges the dharmas both doctrinal and disciplinary, numerically in the Ekottara style, i.e., gradually increasing the number of dharmas from one to ten. The contents of this text agree to a large extent with those of the Saṅgīti and Dasuttara suttantas¹ of the Dīghanikāya.

v2.c35.s3.t4.51

¹ Cf Dasuttara-sūtra in *Abhidharmakośa-vyākhyā* (Jap ed.), p 590

v2.c35.s3.t4.52

Besides these seven recognised texts of the Sarvāstivādin Abhidharmapiṭaka, there were a few other digests and commentaries dealing with the topics of the Abhidharma. The exhaustive commentary on the Jñānaprasthāna-sūtra was, of course, the Mahāvibhāṣā, compiled according to Paramārtha, by the Kātyāyaniputra himself with the assistance of Aśvaghoṣa of Sāketa. Among the digests, the most important work is Vasubandhu's *Abhidharmakośa*, which has got a bhāṣya written by Vasubandhu himself and a vyākhyā written by Yaśomitra. Then there are two other texts viz., *Abhidharma-nyāyānusāra* and *Abhidharma-samaya-pradīpikā*, attributed to Saṃghabhadra an opponent of Vasubandhu. Saṃghabhadra wrote these works to refute some of the theses of Vasubandhu, specially those which were in support of Sautrāntika views.

v2.c35.s3.t4.53

There was an earlier digest called the *Abhidharmasāra* written by Dharmaśrī. It contained eight chapters, viz., dhātu, saṃskāra, anuśaya, ārya, jñāna, samādhi, miscellaneous, śāstra-varga or vāda-varga².

v2.c35.s3.t4.54

² For details see La Vallée Poussin's Intro to the *Kośa*, p lxiii

v2.c35.s3.t4.55

Among other works of note belonging to this school we may mention *Śāriputrābhidharma*, *Abhidharmāmṛtaśāstra* of Ghoṣa, *Abhidharmahr̥daya* of Dharmottara and *Lokaprajñapti-abhidharmaśāstra* of an unknown author.

Linked texts (77)