Pramāṇa · Search · About · API

Studies in the Origins of Buddhism c7: Chapter III. The Stratification of Sutta-Nipāta Itivuttaka and Udāna

Studies in the Origins of Buddhism c7

Language
English

Chapter III. The Stratification of Sutta-Nipāta Itivuttaka and Udāna

c7.2

Suttanipata (Sn.) and the Chinese Agamas.—According to Anesaki¹ no Chinese version of Sn. as a whole can be traced, but the "Aṭṭhakavagga" may be identified with No. 674 of Nanjio: "Sūtra spoken by Buddha on the fulness of meaning"—which was translated about the beginning of the 3rd century A.D. (According to Nanjio between 222–280 A.D.). It contains sixteen texts or stories in which verses are incorporated; and the agreement of the verse portion with Pali is perfect.

c7.3

¹ JPTS 1906-'7, pp. 50-51.

c7.4

Now "it is just the Aṭṭhavagga which happened to be preserved in the fragments"² discovered from E. Turkestan by Stein; and what is more, the outstanding difference between the Pali and the Sanskrit version is the existence of prose narratives prefixed to the verses of the several vargas of the latter.³ And Hoernle comes to the "unavoidable" conclusion "that the Sanskrit text is a translation from some vernacular....original; and that the translator, observing the absence of an introductory narrative, himself supplied that narrative, and pointed out the exact place where he came to the translation of the verses of his original text."⁴ May not the prose of the Chinese translation be similar in character and thus a later, almost commentarial, addition to the verse?

c7.5

² Hoernle JRAS. 1916, p. 709.

c7.6

³ Ib., p. 718. In the text the only discrepancies of any importance are these: For Pali verses 839-840 there must have been a much shorter text in the original for the Sanskrit translation; and there could have been only one verse for Pali 841-842 (Ib., p. 720).

c7.7

⁴ Ib., p. 719.

c7.8

Prose and Verse.—In this connection it may be noticed that Franke regards the prose of Sn. as representing a later stratum than the Gāthās.⁵ And Fausboll holds all the prose passages to be later additions.⁶ It appears, however, that in regard to this problem one must be a “Vibhajjavādin” and not an “Ekamsavādin”. In sutta 4, the prose sequel at any rate contains what is undoubtedly a late stock-finish—it recurs in suttas 7, 10, 19, 30, 31, 32 (with relevant additions) and 35, and is equally frequent in the other Nikāyas. In sutta 10 the verses are in the ‘riddle-dialogue’ form and do not stand in need of the prose preface. In sutta 17 the verses, in spite of different contents, have a somewhat similar form and hence receive a prose preface closely resembling that in suta 10. The verses in sutta 35 contain within the framework of the dialogue the necessary narrative also, and it is possible that the prose in the beginning is a later addition. The suttas 6, 7, 16, 19, 24, 26 and 31 can do very well without the brief prose with which they commence. Nor are the prose portions of sutta 32 at all necessary for comprehending the essential verse. And the possibility of the prose being a merely conventional addition in these cases is strong, although it is hard to clinch the argument. In sutta 29, on the other hand, the prose has as much claim to be regarded as authentic as the verse, which in the next sutta seems to take the preceding prose for granted as is also the case in sutta 33. In suttas 36 and 38 both the prose and verse can stand independently. In fact in these suttas the verse is always introduced thus “Idam avoca Bhagavā, idam vatvā Sugato athāparam etad avoca satthā.” But it is impossible to say which is earlier.

c7.9

⁵ ZDMG. 1909, 1.

c7.10

⁶ Winternitz, II, p. 98.

c7.11

The antiquity of the Atthaka and Parayana vaggas. —It appears that Pārāyaṇa vagga also occurs in the Chinese Tripiṭaka,⁷ the Saṃyuktāgama of which contains, besides, the Megha and the Kokāliya suttas.⁸ Further the Khaggavisāṇa, the Pabbajjā, the Padhāna, the Nāḷaka and the Sabhiya suttas find their counterparts, in some cases almost word for word, in the Mahāvastu and the Lalitavisatara.⁹

c7.12

⁷ Ib., 92 fn. 3.

c7.13

⁸ Bapat's ed. of Sn., p. xxx.

c7.14

⁹ Ib. Since the names of the sutta in Sn. do not seem to have been rigidly fixed, and many of them are known to have alternative titles (Ib. XVIII) may it not be that the Chinese counterparts of many other suttas of Sn. may yet be discovered?

c7.15

It is, however, only with reference to the Aṭṭhaka and Pārāyaṇa vaggas that we can, on external evidence alone, assert their extreme antiquity. Mahāvagga V, 13 speaks of the recitation of the Aṭṭhakavaggikas, and the same story recurs in Ud. V. 6 which has “soḷasa aṭṭhakavaggikas.”¹⁰ Divyāvadāna (really a part of the Mūlasarvāstivādavinaya as translated in Chinese and Tibetan) has developed the episode into a “petit roman” and relates the recitation of Ud. Pārāyaṇa, Satyadṛṣṭi, Śailagāthā, Munigāthā, (Dulva adds Sthaviragāthā and °āgāthā) and the Arthavargīya sūtra.¹¹ The Sarvāstivāda Vinaya, preserved only in Chinese (the Chinese title=Daśādhyāyavinaya) agrees closely with it, but has the recitation of Pārāyaṇa and Satyadarśa only.¹² The Mahīśāsakavinaya in Chinese (entitled Pañcādhyāya-vinaya) recites the Koṭikarṇa story after the manner of the Pali Vinaya. It speaks of the "sūtra des Seize Section de Sens" (The Chinese term—arthavarga).¹³ The Dharmaguptavinaya, available in Chinese alone, and closely allied to the above mentions the recitation of the "Seize Phrases de Sens" (Arthapada). The Mahāsaṅghikavinaya differs most; it has Pūrṇa and Śroṇaparánta for Kātyāyana and Avantī, but speaks of the recitation of "Aṣṭavarga."¹⁴

c7.16

¹⁰ Lévi—JA. 1915—Mai-Juin, p. 403.

c7.17

¹¹ Ib., 405.

c7.18

¹² Ib., 407.

c7.19

¹³ Ib., 408.

c7.20

¹⁴ Ib., 411.

c7.21

The Koṭikarṇa history, thus, must have early formed a part of the canon, and it already presupposes the Aṭṭhakavagga. Again, the Dharmagupta enumeration of the contents of the Kṣudrakāgama in the Rājagṛha saṅgīti places "Arthapada" and "Dharmapada" side by side.¹⁵ It may be noted that SN. III. 12 refers to the Aṭṭhaka by name.

c7.22

¹⁵ Ib., 415.

c7.23

Pārāyaṇa is mentioned twice in AN. II. 45, III. 399 and 401. So in S. II. 49. Sarvāstivāda vinaya mentions it among the "grands sūtras", in pursuance of a theme which is identical with Mahāvagga (Pali III. 5. 9); and occurs also in the Varṣāvastu of the Mahīśāsakas, Dharmaguptas, Mūlasarvāstivādins and Mahāsaṅghikas. The actual list, however, is given only in the Sarvāstivāda vinaya.¹⁶ But in one passage the mahāsaṅghikavinaya speaks of the Pārāyaṇa and Arthavarga together,¹⁷ and so does the Abhidharmamahāvibhāṣā.¹⁸

c7.24

¹⁶ Ib., 421-422.

c7.25

¹⁷ Ib., 422.

c7.26

¹⁸ Ib., 424.

c7.27

The internal evidence too points towards the antiquity of the Aṭṭhaka and Pārāyaṇa and has been in a general way indicated by Fausböll.¹⁹ He points to the archaic and relatively obscure character of the language²⁰ of the Aṭṭhakavagga, and, on the ideological side, states “we see here a picture not of life in monasteries but of the life of hermits in its first stage. We have before us not the systematizing of the later Buddhist church, but the first germs of a system.”²¹

c7.28

¹⁹ Int. to the Tr. of Sn. in SBE X, pp. xi-xii.

c7.29

²⁰ See supra, Chap. II. It may be mentioned that there is a notable increase in the v. ls in the last two vaggas of Sn.

c7.30

²¹ loc. cit.

c7.31

The strata in Sn.—But whereas the Aṭṭhakavagga and the Pārāyaṇa undoubtedly represent the earliest stratum in Sn., it appears, in contradiction to Fausböll, that the Mahāvagga as a whole belongs to the latest stratum, while the suttas of the first two Vaggas have, in general, an uncertain status.

c7.32

The title of the “Aṭṭhavagga” presents a difficult problem. In Lévi’s survey whereas the Mahīśāsaka and Dharmagupta Vinayas have Arthavarga and Arthapada respectively, the heterodox sect par excellence—the Mahāsanghikas—have Aṣṭavarga. Sanskrit sources like the Divyāvadāna, the Ab. k. or the East Turkestan fragments already referred to, have Arthavarga in favour of which Lévi decides. That looks very unlikely, for while it is easy to see how Aṭṭhakavagga could have been mis-sanskritized as Arthavarga—there are other instances of wrong Sanskrit renderings—the reverse is difficult to explain. “Artha” becomes in Pali regularly “Attha” except in the compounds Aṭṭhapatti and Aṭṭhakathā where it means “meaning” (See PTSD). Indeed, that is how the Mahīśāsaka and Dharmagupta Vinayas seem to understand it, but with little cogency, since Artha in Arthapada or Arthavarga is best taken in the same sense in which we have Atthābhisamaya, Atthapatti, Paramattha etc. Besides, how otherwise explain the entry of ‘Ka’ in Aṭṭhaka? Finally, and this clinches, the vagga actually contains a group of four Aṭṭhakas (so called in the text) each consisting of eight verses.

c7.33

This is of great significance. The Aṭṭhakavagga reveals a definite plan of arranging the suttas. We begin with one containing six verses; then follow octades, after which the sequence of the total number of verses in the suttas is as follows:--

c7.34

10, 10, 11, 13, 14, 16, 17, 20, 20,²² 20 and 21. This striking progressive order shows clearly that the present Aṭṭhakavagga is the product of careful editorial activity. (It is also noteworthy that both the first and the last suttas of this vagga speak of the Parissayas). It appears that like the Family Books of the Ṛgveda, the four Aṭṭhakas form a closely connected nuclear group to which the other suttas were editorially added.

c7.35

²² Metrical variation ought to be noted.

c7.36

The nuclear Atthakavagga.—The “Aṭṭhaka-group” has a distinguishing unity of metre, language, subject-matter and style of expression. Each of the four Aṭṭhakas consists of eight verses in a mixture of Indravajrā and Upendravajrā.²³ Considering the extent of these Aṭṭhakas, they show, even within the Aṭṭhakavagga, a greater use of archaic forms than the other suttas. In the Guhaṭṭhaka occur cutāse (verse 3), avītataṇhāse (v. 5) Vineyya (v. 7) and Pariññā (v. 8); the next has Pāvā which occurs twice (v. 3), ātumānaṃ (Ib), and niccheyya (v. 6); the Suddhaṭṭhaka has Pāvā (v. 2), vadānaṃ (Ib), sitāse (v. 4), and gahāya (Ib);²⁴ and the Paramaṭṭhaka has paribbasāno (v. 1),²⁵ samuggahāya (v. 2) and Paṭicchitāse (v. 8). On the other hand, the following Jarāsutta of 10 verses has hardly any marked ancient form; Tissametteyya of similar extent has, however, sikkhisāmase (v. 1) sutvāna (Ib), caritvāna (v. 3), and vippahātave (v. 4) Pasū° has vadānā (twice-v. 1. & 2), sagghasi and sampayātave (both in v. 11) in 11 verses; Māgandiya has disvāna (v. 1), niccheyya (v. 3), vinicchayā (v. 4) and anuggahāya (v. 5) in 13 verses. After this though the length of the suttas increases much, only the Kalahavivāda°, Cūlaviyūha° and Mahāviyūha° have marked archaic forms to show.

c7.37

²³ The metrical scheme of the sixteen suttas in the Aṭṭhakavagga is as follows:--
1. Anuṣṭubh; 2-5. Upajāti; 6. Vaitālīya (loosely); 7. Anuṣṭubh; 8-9. Upajāti; 10. Anuṣṭubh; 11-13. Upajāti; 14. Mixture of Anuṣṭubh and Vaitālīya; 15. Anuṣṭubh; 16. First half Anuṣṭubh, second Upajāti.

c7.38

²⁴ Is the Nirassajanti of v. 4 in this Aṭṭhaka the result of a confusion between Nirassanti and Nissajjanti?

c7.39

²⁵ Is it connected with “Paryavasyati”? Cf. the expression Diṭṭhiparibbasāno°, in Cūḷa° and Mahāviyūha°?

c7.40

In expression the Aṭṭhakas are more involved and obscure than the other suttas.²⁶ The very frequent occurrence of Nara and Jantu in them is also significant, and peculiar. The use of Guhā in the Guhaṭṭhaka is reminiscent of Kaṭha up.

c7.41

²⁶ Cf. Fausböll. loc. cit.

c7.42

The state of ideas, again, is perhaps the earliest to be found in the canon. Man-Nara Jantu—is bound and tormented in the world (Bhavo, loko) by pleasures (Kāma) and desire (Taṇhā, chando). Instead of the later formula of Paṭiccasamuppāda we have such simple statements as “(Kāmā....icchānidānā bhavasātabaddhā....” (Guhaṭṭhaka-v. 2), or the implication in v. 778 that desire (Chanda) springs from contact (Phassa). Egoism (Mamāyita) and Saññā—which is the wordly counterpart of Paññā (see vv. 847, 779, 792),—must be abandoned. So must all speculation (Diṭṭhi) and encumbrance (upadhi). Avoiding the extremes (ubhesu antesu vineyya chandaṃ) and grasping nothing, not even individuality (Attaṃ pahāya anupādiyāno—800) a man attains to a state which is neither the self nor the non-self (Attaṃ nirattaṃ na hi tassa atthi—787), which is beyond all limitations (sīmātigo—795) and all mental and merely moral processes (Diṭṭha, suta, muta, ñāṇa, and sīlavata; where there is not the slightest mental construction: “Pakappitā natthi aṇū pi saññā”—802), and which means a peace (Santi, Abhinibbuti—783) that has neither the taste of pleasure nor is distasteful (na rāgarāgo na virāgaratto—795).

c7.43

The rest of the Atthakavagga. Sutta no. 44, which immediately succeeds the four Aṭṭhaka, goes out of its way to incorporate in the last verse the conclusions of suttas 41 and 42. Sutta 49 develops the theme of v. 773 and we are here nearer the formula of Paṭiccasamuppāda. The Mahāviyūha appears to be a slightly more elaborate version of the Cūḷaviyūha.

c7.44

Conclusion about the Atthakavagga.—On the whole, however, the ideological difference between the various suttas in the Aṭṭhakavagga is too small to permit a definite stratification within it, but it must be remembered that from the literary standpoint the Aṭṭhakavagga is not homogeneous.

c7.45

The Parayana vagga.—It has already been mentioned that tradition closely connects the Aṭṭhaka and Pārāyaṇa vaggas. It is noticeable in this connection that just as the Aṭṭhaka has sixteen “suttas” so the Pārāyaṇa has sixteen “Pucchās”. Its language shows occasionally old forms like Saṅkhātadhammāse (56.7), Sutvāna (twice in 60.1 & 2 also in 66.7), vajju (61.8), samaṇabrāhmaṇāse (four times in 62) Daṭṭhu (66.3); Disvāna (77.2).

c7.46

In 58.6 occurs the rare, and possibly old, word Parovarāni. It is similarly rare for Buddha to be addressed as "Sakka" and even "Brahmā" (60, 3 & 5; 61.1) and "Samantakkhu" is here suggested by "Sakka"—it has not as yet become dissociated and independent.

c7.47

Sutta 63 of four verses is composed in the "gāyatrī" metre, which is rare if not unique. Though on a much more metaphysical level than the Aṭṭhakavagga, its doctrinal position is, on the whole, not much different. Its clearest exposition is found in sutta 61, which, taken in conjunction with the other suttas in the vagga, yields a picture somewhat as follows:—

c7.48

Viññāṇa or consciousness as manifested in the individual, ceases in the highest experience (Paramavimutti cf. 61.4) but of what remains neither being nor non-being can be predicated. We have here a transcendentalism which is in danger of being misunderstood as Nihilism.

c7.49

It may be noted that Nāmarūpa (at one place Nāma kāya 61.6) is a frequent expression in this text, but there is not the faintest suggestion of its interpretation in terms of the "khandhas" which find no mention. Its psychological analysis is exhausted in the tetrad "Diṭṭhasutamutaviññāta" (63.3) "Viññāṇaṭṭhiti" is yet used in its general and unformulized meaning (68.3). In the description of "Aññāvimokkha" in 68.2-3 we have the materials for the later formula of the fourth Jhāna.

c7.50

The sixteen "Queries" may thus be placed in the same stratum as the Aṭṭhakavagga, though they are, perhaps, slightly later.

c7.51

Conclusion about the Parayana.—The Vatthugāthās are, however, a definitely later addition. They show a knowledge of the south which is exceptional in the Nikāyas, are written in a fluent narrative style reminiscent of the Mbh-ballads (mark, for instance the verses 5–8), seem to use Abhiññā in the sense of miraculous powers, assume a fully developed theory of the Thirty-two Marks, and serve the merely commentarial purpose of furnishing a narrative setting to the “Queries” which in no wise depend upon it.²⁷ It is possible that the epilogue entitled “Pārāyaṇasutta” is also a later addition.

c7.52

²⁷ Cf. “The Mahāpurisalakkhaṇas....are in Sn. 1022 attributed to Bāvari (i.e., the “Babylonian”), and clearly point to the late origin of the vatthugāthās as well as to Babylonian influence” (W. Stede-Afterword to PTSD. p. 202. fn. 3).

c7.53

Mahavagga.—Though it is true that the Pabajjā- and Padhāna suttas find very close parallels in Mvu. II. 198ff and Ib. II 238ff (also LV.XVIII) respectively, and present the Buddha-legend in a form much less developed than many suttas of DN and MN, still as Winternitz has stated, it is impossible to place them in the earliest Nikāya stratum. The incident in the Pabbajjāsutta can hardly be regarded as historical and seems to have been invented “for the greater glory” of Buddha. The Padhanasutta is almost a continuation of it, but seems to stand in need of intermediate prose narrative to explain the change of speakers. Besides, it glories in an excess of asceticism, which is quite in contravention to Buddha’s own mandate (sāsana). Much of the simplicity of these poems is to be attributed to their being good poetry.

c7.54

The long and straggling Sabhiyasutta (cf. Mvu III. 394ff), which gives fanciful etymologies for Samaṇa and Brāhmaṇa, belongs to a still later stage in as much as it refers to sixty-three Sramaṇic dogmas.

c7.55

The Selasuttas (M. sutta 92-II. 146ff) can hardly be earlier since the theory of "Thirty-two Marks" plays a central rôle in it, and Buddha is described as the supreme Dhammarāja, with Sāriputta as his commander.

c7.56

The Kokāliyasutta (occurs in the Chinese Saṃyuktāgama—see supra) belongs to the same or even later stratum. For the relatively light sin of criticizing Sāriputta-Moggalāna, the monk Kokāliya dies miserably and, what is more, is condemned to a hell, the duration of which is simply inconceivable. We are given a series of ten hells, which are described in gruesome detail. Here is, in fact, a type-sutta for the extreme Purāṇic style.

c7.57

Of the Nālakasutta (parallels in Mvu. & Lal. Vist), the first half consisting of the Vatthugāthās is, at any rate, from a late stratum. Buddha’s birth is a great event for the gods, and Buddha himself is called "Bodhisatta". Besides, Asita is described as "Lakkhaṇamantapāragu" which may refer to his knowledge of the Thirty-two Marks. These, it may be remembered, are often described as "Mantesu āgatāni."

c7.58

Finally, the Dvayatānupassanā-sutta is doubtless late. It introduces a dichotomy within the formula for the Four Truths, uses "Upādisesa" and "Anāgāmitā" technically, seems to distinguish between Cetovimutti and Paññāvimutti, and describes eleven Paccayas of Dukkha which include all those that occur in the Paṭiccasamuppāda except Nāmarūpa, Saḷāyatana, Bhava, Jāti and Jarāmaraṇa. In short, it evinces considerable dogmatic and analytical development.

c7.59

Conclusion about the Mv.—Thus of the twelve suttas of the Mahāvagga seven appear definitely late, the term late comprising two strata; the first consisting of suttas 27 and 28, and the second of the other five.

c7.60

The Rest.—Some of the remaining suttas of the Mahāvagga and the first two Vaggas embody the Buddhist protest against the social and sacrificial ideas of the Brāhmaṇas. They portray what may be termed the Brāhmaṇa-Śramaṇa conflict. To this class belong the Vasala°, Brāhmaṇadhammika°, Sundarikabhāradvāja°, Māgha° and Vāseṭṭha suttas. Except for the last, these suttas appear similar and are less dialectical than similar ones in DN and MN and the exception mentioned is actually found in the latter collection. They do not aim so much at confuting logically the theory of the Brāhmaṇas as at presenting a truer conception of the “Vasala”, the “Brāhmaṇa” etc. They partake more of the nature of ethical sermons than of polemics. The Brāhmaṇadhammikasutta is partly an attempt to convince the Brāhmaṇas of the Buddhist ideal by representing it as having been followed by the ancestors of the Brāhmaṇas themselves, and partly an outcome of the tendency to imagine ideal visions of the future as having been actually realized in the past (since to go back to the past appears easier). This latter tendency is also followed in the Aggaññasutta of DN but receives there a much more developed and mythological treatment.

c7.61

These suttas thus represent the earlier stages of a distinct genre in Buddhist writing; but it seems impossible to correlate, with any degree of certainty, the steps in this particular trend with those in the general ideological evolution of the Nikāyas.

c7.62

Sutta 3 is apparently inconsistent with developed communal life among the Buddhist monks. Its use of the terms Attha and Parissaya, which later on practically disappeared in favour of “āsava” and “anusaya”, also suggests the possibility of it being early. So does the use of Sneha more than Sineha (see PTSD). It describes the state of the fourth Jhāna without mentioning it by name. The five “āvaraṇas” (nīvaraṇas) are, however, presupposed as a technical expression. Also, its use of “ñāṇa” is different from that in the Aṭṭhaka- and Pārāyaṇa-vaggas.

c7.63

Suttas 1 & 2 are not far removed in point of ideology and diction. The former uses samūhatāse and paccayāse (1.14 & 15) and the latter carāmase and bhavāmase (2.15). Besides, both these are written in the same metre, and have the same number of verses. It is likely that they have common authorship.

c7.64

As already suggested the central five verses of sutta 4 are possibly nuclear in character. At any rate, they expound a lively parable of spiritual cultivation, which seems to belong to the same stage of development as the parable of the spiritual car (Dhammayāna) in SN. It is quite innocent of any formulae, and may be placed in the earliest stratum. For the goal it uses the expressions Yogakkhema, Amata, and Sabbadukkhā (Pamutti).

c7.65

Sutta 5 has much to say on the “maggo” but nowhere speaks of it as eightfold or tenfold, which argues for its earliness.

c7.66

Sutta 12 has been recommended by Aśoka.²⁸ It uses the forms anuppavecche (12.20 & 3), muninam (12.2) and atitariya (12.13); and the old term Rajo.²⁹ For the goal it has the expressions Santipada (12.2) and Paramattha (12.13); the fact that whereas Muni occurs in evey vese, Bhikkhu is seen only in the last one, may also not be without significance. It delineates the ideal of the solitary hermit meditating in the forest, and the total impression that one gets after perusing it resembles that resulting from the Aṭṭhakavagga. It may be placed in the same stratum.

c7.67

²⁸ See D. C. Sircar—Select Ins. I. p. 78, fn. 2.

c7.68

²⁹ See the account of Jainism below, Chap. X.

c7.69

Sutta 14 regards the prohibition of meat-eating as only symbolical of good conduct. With this may be compared the early Buddhist reinterpretation of such terms as Brāhmaṇa, Vasala, Yañña, Aggi, Vedagū, Arahant and Āhāro. This reveals a process of transvaluation in progress. In the MN,³⁰ however, the authors are eager to refute the Jinist charge of meat-eating levelled against them. This would appear to represent a later stage, since Buddha is reported to have partaken of "sūkaramaddava" and a development in the reverse direction goes against general probability. The present sutta, therefore, may be considered earlier than MN (loc cit).

c7.70

³⁰ Sutta 55.

c7.71

The short uṭṭhānasutta (no. 22) is singular for its earnestness and power. It is reminiscent of Kaṭh.³¹ It describes Pamāda as Raja and in its spirit of Kiriyavāda certainly belongs to the earliest Buddhism. The composition too may be early.

c7.72

³¹ Kath. 3. 14.

c7.73

Sutta 24 addresses Buddha as Sakka (24.3) and calls him "Samantakkhu, Sakko va devānaṃ sahassanetto" It describes the Ariyadhamma as Parovara (24.11) and explains Taṇhā as Nāmarūpa—(24.13). It looks early.

c7.74

On the other hand sutta 13 (—Khuddakapāṭha VI) is the solitary example of a ‘Parittā’ in Sn. It pays homage to the Three “Jewels”, speaks technically of the 8 Puggalas, the Ariyasaccas, the 6 Abhabbaṭṭhānāni (13.10), the 4 upāyas of vippamutti (Ib) and refers to the theory of “Na te bhavam aṭṭhamam adiyanti” (13.9). It clearly belongs to a late stratum, although it appears less developed than the similarly purposed Āṭānāṭiya sutta of the DN.

c7.75

Of the remaining suttas it does not seem possible to assert anything except that sutta 20 gives the impression of earliness—for its “Inda” is the most venerable of gods; it develops a lively “upamā” (in the Pali sense) of “crossing the river” and concludes “aññāya atthaṃ paṭipajjamāno viññātadhammo so sukhaṃ labhethāti”, (30.8).

c7.76

Concluding table :

c7.77

| (a) Early | Late |
| The Aṭṭhakavg. (the 4 Aṭṭhakas appear earlier than the rest). | Suttas 27-28, 32-33, 37 (Vatthugāthās), and 38. |
| The Pārāyaṇavg. (minus the Vatthugāthās; Pārāyaṇa-sutta may also be late). | (Suttas 27-28 may be earlier than the rest). |
| (Suttas 1-3, 5, 12, 14, 22 and 24 appear early and may belong roughly to the same stratum as above). | Sutta 13 also appears late and may belong to the same stratum as above. |

c7.78

The rest uncertain.—(b) Suttas, 4, 7, 19, 30, 31 and 35 form a distinct class. Sutta 20 may be early.

c7.79

Itivuttaka.—According to Watanabe³² a collection of It, translated by Yüan Chwang about 650 A.D. exists in the Chinese Tripiṭaka. The scheme of its contents is indicated thus:—

c7.80

³² JPTS 1906; pp. 44-49.

c7.81

I. Ekadharmakhaṇḍa
1. sūtras 1-12
2. ,, 13-24
3. ,, 25-47
4. ,, 48-60

c7.82

II. Dvidharmakhaṇḍa
1. sūtras 1-12
2. ,, 13-24
3. ,, 25-36
4. ,, 37-50

c7.83

III. Tridharmakhaṇḍa
1. sūtras 1–13
2. ,, 14–25
3. ,, 26–28

c7.84

The Pali scheme, on the other hand, may be represented thus:—

c7.85

I. Ekakanipāto
1. Pātibhogavaggo suttas 1–10
2. Dutiya° 11–20
3. Tatiya° 21–27

c7.86

II. Duka°
1. Paṭhamo vaggo 28–37
2. Dutiyo° 38–49

c7.87

III. Tika°
1. Paṭhama° 50–59
2. Dutiya° 60–69
3. Tatiya° 70–79
4. Catuttha° 80–89
5. Pañcama° 90–99

c7.88

IV. Catukka° 100–112

c7.89

Fourth Nip. of doubtful authenticity.—The most striking difference is the absence of the 4th Nip. in the Chinese. It is noteworthy in this connection that at least four suttas of this Nipāta occur identically in the AN also—sutta 101=A.4.27; s. 105=Ib. 9; s. 108=Ib. 26; s. 111=Ib. 12. Furthermore, sutta 100 repeats after the third sentence the prose of sutta 98 which seems to have been suggested merely by word consonance. Sutta 102 in its prose portion describes the formula of the Four Truths, and its verses seem to distinguish three stages in the highest spiritual release—Āsavakhaya accompanied by Vimuttiñāṇa.³³

c7.90

³³ Cf. Mrs. Rhys Davids, Ind. Psy., p. 264.

c7.91

It is, however, not unambiguous in its expression (shall we read the fourth line after the first?).³⁴ Sutta 103 has the same theme. Its last verse seems to distinguish between Cetovimutti and Paññā.° The next sutta instead of the earlier triad of Sīla, Samādhi and Paññā, has the pentad of these plus Vimutti and° nāṇadassana, which is the scheme that occurs as a rule in the last books of the AN. The verses in sutta 105 seem to have been taken over from the sutta 15—they harmonize much more with the spirit and diction of the prose there. Sutta 109 presupposes considerable dogmatic development: it considers the 6 Ajjahttika-āyatanas and the 5 ‘Orambhāgiya-saṃyojanas’ and well-known technicalities. Sutta 111 appears to use “khandha” in the later technical sense of the five Khandhas. The next sutta adds to Diṭṭha, suta, muta, and Viññāta, three other terms viz. Patta, Pariyesita, and Anuvicarita.

c7.92

³⁴ Cf. the similar and more coherent version in the verses of sutta 62. Have we a garbled borrowing in sutta 102?

c7.93

It appears, thus, that the fourth Nipāta is, as a whole, not only doctrinally more advanced than the rest of the It, but also partly depends in its composition on the preceding suttas and the AN. This taken in conjunction with its total non-appearance in the Chinese It. certainly points to it as having the character of a later addition.

c7.94

The Third Nip.—Another point that emerges from the tables given above is the very regular and schematic arrangement of the suttas in the Pali Tika Nipāta. Now Watanabe informs us that whereas the Pali suttas of the first two Nipāta are almost all found in the Chinese, three fifths of those of the third are not.³⁵ Against this, of course, it has to be remembered that the third part of the Chinese is apparently defective since the regular “uddāna” fails to occur at its end, and also that the Chinese version of the suttas is in general more elaborate than the Pali one and appears later.

c7.95

³⁵ loc. cit.

c7.96

The internal evidence of the Pali version does not throw any certain light on the “stratigraphical” position of the 3rd Nipāta. Its first two vaggas are, however, on the average, considerably shorter than those of the next two vaggas. Also, the prose in suttas 50-67 follows the same mode of enunciation, which strengthens the general impression of their closer unity.³⁶ Further, some, at least, of the suttas of the last two vaggas of the 3rd Nip. may with some certainty be regarded “late”. Suttas 70 and 71 treat the theme of 64-65 in greater detail. Sutta 73 seems to take sutta 51 for its starting point. Sutta 74 speaks of the “Three Refuges”. Suttas 82-83 describe the occasions on which the gods shout (Devasaddo niccharati). Sutta 90 advocates faith in the “Trinity” and speaks technically of “Yadidaṃ cattāri Purisayugāni aṭṭha purisapuggalā.”

c7.97

³⁶ The prose and verse portions of the sutta 63 have nothing in common.

c7.98

It is thus likely that the third Nipāta has earlier and later elements, the latter being principally represented in the last two vaggas.

c7.99

The first two Nip.—The suttas of the first two Nipātas belong, in general, to the same stratum. The suttas 1-6 resemble each other obviously in prose and in verse so closely that they must be considered for all purposes a single group. They use “Anāgāmitā” in its general “pre-technical” meaning, and employ such forms as luddhāse, duṭṭhāse, mūlhāse, which are extremely rare in the It. This suggests that this is a very early group.

c7.100

The suttas 7-13 are indistinguishable from the previous group as far as the verse portion is concerned. The prose has a different form but one which is common within this group. The next two suttas belong to the earliest ideological stratum. The former equates —Nīvaraṇa=Avijjā=Moha=Saṃsārahetu. It uses Nīvaraṇa pre-technically, and represents the earliest form of the Paṭiccasamuppāda. The latter sutta pictures man (Puriso) as wandering through the worlds—through this station and that (Itthabhāvaññathābhāva) accompanied by his Thirst alone. The two suttas are similar in form and complementary in content.

c7.101

The suttas 16-17 are obviously a unity, the one dealing with the best internal and the other with the best external means of spiritual progress. The former describes the supreme goal as Anuttara yogakkhema and Uttamattha.

c7.102

The demarcation of the 3rd vagga of Nip. 1 from the 2nd is just as arbitrary as that of the 2nd from the 1st. In either case the last and the first suttas are closely interdependent. It is likely that these divisions which follow no other principle than that of grouping ten suttas together are later introductions.

c7.103

The prose of sutta 22 in which Buddha grandiloquently recounts his previous divine births is possibly later than the verse where we have Dāna, Samacariyā and Mettacitta for the Dāna, Dama, and Saṃyama of the prose. The difference is considerable.

c7.104

Sutta 23 advocates Appamāda as the sole means to Atthābhisa:naya, both diṭṭhadhammika and samparāyika. This has suggestions of earliness. The next sutta, on the other hand, belongs ideologically to exactly the same stage of development as the Ratana sutta of the Sn.

c7.105

Sutta 25 (Sampajānamusāvāda) may have formed the original for the Cūḷarāhulovāda of the MN.

c7.106

In sutta 27 (Mettabhāva°) it is noticeable that of the similes adduced only the first really fits in. The point to be illustrated is how “opadhikāni puññakiriyāvatthūni—mettāya cetovimuttiyā kalaṃ nāgghanti soḷasim.” The first simile is the transparent and very natural one of the stars and the moon. The second is: “just as the autumn sun shines forth, dispelling all darkness from the sky.” The relation between a lesser and higher virtue is, however, not that of darkness and light. The third simile speaks of the “osadhitārakā” shining in the morning, and has no reference whatever to any counterpart, similar or dissimilar. Again, if the following verses have any authentic reference to a simile, it is to the first one only. It is thus obvious that there has been a later addition to the prose text. Of such forced additions there are other examples in the Nikāyas.

c7.107

In the second Nipāta the verse portions of the suttas 30-31 are less tolerant and more dogmatic than the corresponding prose.

c7.108

The verse in sutta 38 has two distinct and unconnected divisions; the second occurs in MN and is more relevant there.

c7.109

In sutta 43 the verse has six lines arranged in two groups of three each, and since each consists of eight syllables the metre may be supposed to be a variety of Gāyatrī. Ideologically also it leaves no doubt of the positive connotation of “Nirodha” which is described as “Santaṃ atakkāvacaraṃ dhuvaṃ ajātaṃ asamuppannaṃ asokaṃ virajaṃ padaṃ (then, nirodho dukkhadhammānaṃ saṅkhārūpasamo sukhoti).” In sutta 44 there seems to be a difference between the idea of the prose and of the verse. Both distinguish between Sa-upādisesa and Anupādisesa Nibbānadhātu but according to the former the second also appears attainable during life: “Tassa idheva—sabbavedayitāni anabhinanditāni sītibhavissanti ayaṃ vuccati......anupādisesā nibbānadhātu,” while according to the latter “Anupādisesā pana samparāyikā yaṃhi nirujjhanti bhavāni sabbaso” (which is to be considered in contrast with the previous line: Ekā hi dhātu idha diṭṭhammikā saupādhisesā bhavanetisaṅkhaye).” According to the prose the distinction between the two Nibbutas lies in the fact that after Anupādisesa, experience is freed from all hedonic tone. This appears the earlier version since the view advanced in the verse is the one that became standard later. The earliness of the prose is also suggested by the fact that it speaks categorically of just five and not six senses.

c7.110

Conclusion.—The It. thus reveals at least two, if not three strata, of which the earliest is represented principally in the first two and a half Nipātas and the latest in the fourth.

c7.111

Udana.—Winternitz points out that the utterances are, “as a rule, older than the narratives into which they are inserted......”³⁷ The stories contained in many of the suttas are very simple and sometimes even inappropriate “to the pathos of the utterances themselves.”³⁷ This appears quite correct, and what is more, not only do the verses and the prose introductions form two distinct but also the only two distinguishable strata within the Ud. The verses are in style and ideas quite homogeneous and show many signs of earliness. Metrical irregularities are not uncommon but they only strengthen the impression of antiquity. At places archaic forms occur but they are very rare:—Esāno 13.1; lapetave 21.1; sutvāna 38.1; and laddhāna 50.1. The language is, on the whole simple, clear, fluent and not without a certain force. It is reminiscent of the Dhp. rather than of the Sn.

c7.112

³⁷ Op. cit. II. pp. 85-86.

c7.113

Of technicalities and formulae the verses are quite innocent, and yet they are, in most cases, more than merely moral or ascetic exhortations of a general character. They are, on the whole, distinctly Buddhist, but seem to belong to a pre-scholastic stage.

c7.114

The prose introductions, on the other hand, serve generally the commentarial purpose of indicating the occasion of the particular udāna. In some cases the prose narrative has the flimsiest of connections with the following verse and in certain others none. Besides, in a few instances their lateness is manifest from their contents. The probably genuine and early prose is extremely rare.

c7.115

The wholly mechanical arrangement of the Ud. contra-indicates a historical growth as among the various vaggas.

c7.116

Vg. 1.—The first or ‘Bodhivaggo’ can be much more suitably named the ‘Brāhmaṇavaggo’ as nearly all the verses contain “Brāhmaṇa” as the vital word in them. In the prose introductions to the first three suttas, which also occur in the Mahāvagga (Vin.), the full formula, of the Paṭiccasamuppāda is concisely stated in both forms (the general and the detailed) and orders (‘anuloma’ and ‘viloma’). The final verses are, on the other hand, of a much more general character. They simply speak of “Dhammapātubhāva.”³⁸ Suttas 4-9 are concerned with the true conception of “Brāhmaṇa” and in every case the prose appears to be the result of ad hoc invention. In sutta 10 the prose has as much independence as the verse, but they are connected all the same. The former beside the interesting narrative has a short and rather obscure sermon. It becomes, however, intelligible if placed by the side of the more explicit first sermon of the MN. The first line of the verse occurs in the DN and the following two lines resemble very strongly Kaṭh.³⁹ Besides, there is a homily on these verses contained within the Ud. itself (sutta 71). These are undoubtedly very ancient.

c7.117

³⁸ See below.

c7.118

³⁹ Kaṭh 5-15.

c7.119

Vg. 2.—The next vagga (Mucalindavagga) is better renamed "Sukhavagga" since all the verses in it are concerned with contrasting worldly and non-worldly happiness. In the first two suttas the prose has little relevancy to the verse, and in the next two ones it is ad hoc. In the following two suttas two different incidents are narrated to lead to the same moral contained in the final verse (the first line of which differs in form in the two suttas). In the other suttas of the vagga also, except sṭta 18, and perhaps s. 20 (cf. v.7.1.) the prose has the general appearance of having been built upon the verse. In sutta 18, however, it dominates and the verse is of the nature of a short and pointed moral following from it.

c7.120

Vg. 3.—In vagga 3 the verse of the first sutta has a very archaic appearance—it uses the form "lapetave" and its "dhunamānassa purekatam rajam" reflects the same idea as the "Dhona" of the Duṭṭhaṭṭhaka in the Sn. The expression "Sabbakammajaha" too is uncommon, and accords more with the Jinist than with the Buddhist emphasis. Whereas the verses of the suttas 22-24 vary but little, the corresponding prose narratives have little in common. The case is similar with the suttas 27-28.⁴⁰ In sutta 29 the expression Asippajīvī in the verse occasions a long list of the industries in the prose. In sutta 30 the form of metrical construction is quite unusual; and, more important, it speaks of only two "taṇhās", bhava-and vibhava°, instead of the usual three.

c7.121

⁴⁰ On sutta 27 see Winternitz II. p. 86.

c7.122

Vg. 4.—In the fourth vagga, the first sutta has a relatively long prose sermon employing the Aṅguttara form and connected but little with the final verses. The prose of sutta 34 illustrates the verse at the end by means of a very fanciful narrative. In the next sutta the verse may well have formed an original part of the prose. The same relation is possible in sutta 38, which relates the well known story of the Paribbājikā Sundarī. Sutta 40 appears a variant of the preceding sutta; its verse portion is a part of that of the latter, and the prose is brief and non-descript.

c7.123

Vg. 5.—In vagga 5, the first sutta forms an integral whole and recurs in SNI. Its view of Attā as the dearest is reminiscent of Bṛ. up. The next two suttas belong to the other end of the scale, and presuppose a well developed Buddhology—the former describes how the mothers of the Boddhisattas must be shortlived, and the latter recounts an event contemporaneous with the Paccekabuddha Tagarasikhin. The very long sutta 45 has by way of verse two brief lines at the end, couched in metaphorical language and wholly irrelevant to what precedes. The prose details the eight miraculous qualities of the ocean and the Dhammavinaya.

c7.124

Sutta 46 has already been considered in its various versions.⁴¹ It is plainly later than the collection of the sixteen Aṭṭhakavaggikas, and yet belongs, as Lévi has pointed out, to the earlier rather than the later parts of the canon. In sutta 48 a story relating Devadatta’s schism has been prefixed to a verse which just states that it is easy for the good to do good and vice versa.

c7.125

⁴¹ In connection with Sn.

c7.126

Vg. 6.—Of vagga 6, the first sutta recurs in the Mahāparinibbānasutta. It depicts Buddha as a vain magician who resorts to oblique hints in order to be requested to live longer. The next sutta partly occurs in SN I. pp. 77-9. The verse it has at the end is quite irrelevant to the preceding prose. The sutta which follows is an organic whole and contains the famous parable of the blind men and the elephant. It expresses Buddha’s opposition to Ekamsavāda so well, correlating it with his Abyākatavāda, that it is difficult to resist, considering it as old as Buddha himself. It stands on the same level as the parable of the poisoned arrow,⁴² to which it is a valuable supplement.

c7.127

⁴² MN. Cūḷamāluṅkyaputta°.

c7.128

The list of the Diṭṭhis mentioned in sutta 54 is extended in the next two suttas, which have an identical prose prefaced to variant verses. Sutta 59 is a typical instance of the ad hoc prose (see Winternitz, loc. cit.) Sutta 63 of the next vagga offers another equally patent instance. So does sutta 64. Sutta 67 is even less convincing. The author of the prose of sutta 69 seems to have grossly misunderstood the final verse, which intends “water” in no more than a merely figurative sense. Sutta 70 is a flagrant instance of the prose-verse unconnectedness.

c7.129

In suttas 71-74 the introductions are negligibly brief and nondescript. In sutta 71 there is no verse; instead, there is a short prose homily on the final verse of sutta 10.⁴³ The suttas 73-74 too have no verse, the central text of the one occurring in it, and of the other in MN.⁴⁴

c7.130

⁴³ Winternitz (loc. cit.).

c7.131

⁴⁴ Sutta 144.

c7.132

The Cunda-sutta, which recurs in DN.s.16, contains the "unorthodox" report of Buddha's partaking of "sūkaramaddava" at the place of Cunda. The two verses beginning "Cundassa bhattam bhuñjitvā etc." belong to the earliest stage of that genre which culminated in Aśvaghoṣa's Buddhacarita. The next sutta also forms part of DN.s.16, but beside the prophecy about the future of Pāṭaliputta it contains the miracle of crossing the river Gaṅgā; and both these, especially the latter, must be considered of dubious authenticity.

c7.133

The identical prose preface to the suttas 79-80 is clearly based on the final verses, which however must be understood with reference to the similar utterances in the Pārāyaṇa vagga of the Sn.

c7.134

Conclusion.—The verses of the Ud. thus represent a homogeneous early stratum while the prose is for the greater part later, and more heterogeneous.

Linked texts (34)