A Philological Approach To Buddhism

The Bukkyō Dendō Kyōkai Lectures 1994

K. R. Norman

SCHOOL OF ORIENTAL AND AFRICAN STUDIES UNIVERSITY OF LONDON

1997

Published by the School of Oriental and African Studies (University of London), Thornhaugh Street, Russell Square, London WC1H OXG

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

ISBN 0728602768 ISSN 0959-0596


Foreword	vii
Abbreviations	ix
Bibliography	xi
I Buddhism and Philology	1
II Buddhism and its Origins	21
III Buddhism and Oral Tradition	41
IV Buddhism and Regional Dialects	59
V Buddhism and Writing	77
VI Buddhism and Sanskritisation	95
VII Buddhism and Aśoka	113
VIII Buddhism and Canonicity	131
IX Buddhism and the Commentarial Tradition	149
X Philology and Buddhism	167
Word Index	185

Foreword

The lectures contained in this book are slightly revised versions of those I gave as Bukkyō Dendō Kyōkai Visiting Professor at the School of Oriental and African Studies from January to March 1994. My thanks are due to the Bukkyō Dendō Kyōkai for instituting the Professorship and for the authorities of the School of Oriental and African Studies for appointing me to it.

The lectures were intended for an audience which knew something about Buddhism but little or nothing about philology. The notes which I have added are intended for the general reader, rather than the specialist, to supply the sources for the information I give.

Abbreviations


AMg	Ardha-Māgadhī
AO	Acta Orientalia
B.C.E.	Before Common Era
BHS	Buddhist Hybrid Sanskrit
BHSD	Buddhist Hybrid Sanskrit Dictionary
BHSG	Buddhist Hybrid Sanskrit Grammar
BSO(A)S	Bulletin of the School of Oriental (and African) Studies
BSR	Buddhist Studies Review
BUp	Bṛhad-Āranyaka Upaniṣad
C.E.	Common Era
CP I-V	K.R. Norman, Collected Papers, Volumes I-V (1990-94)
CPD	A Critical Pāli Dictionary, Copenhagen
CUP	Cambridge University Press
DEDR	T. Burrow & M.B. Emeneau, Dravidian Etymological Dictionary (revised edition, 1984)
DPPN	G.P. Malalasekera, Dictionary of Pāli Proper Names (193738)
EV I, II	K.R. Norman, Elders’ Verses I (1969), II (1971)
GD I, II	K.R. Norman, The Group of Discourses I (1984), II (1992)
GDhp	Gāndhārī Dharmapada
IIJ	Indo-Iranian Journal
IT	Indologica Taurinensia
JA	Journal Asiatique
JBuRS	Journal of the Burma Research Society
JOI(B)	Journal of the Oriental Institute (Baroda)
JPTS	Journal of the Pali Text Society
JRAS	Journal of the Royal Asiatic Society
m.c.	metri causa
MIA	Middle Indo-Aryan
MRI	Minor Readings and Illustrator
MW	M. Monier-Williams, Sanskrit-English Dictionary (1899)
OUP	Oxford University Press
PDhp	Patna Dharmapada
PE	Pillar Edict
PED	The PTS’s Pali-English Dictionary
Pkt	Prakrit
PTS	Pali Text Society
RE	Rock Edict
SBE	Sacred Books of the East
Skt	Sanskrit
SOAS	School of Oriental and African Studies (London)
SWTF	Georg von Simson (ed.), Sanskrit-Wörterbuch der buddhistischen Texte aus den Turfan-Funden (1973-)
ūberbl	O. von Hinūber, Das āltere Mittelindisch im ūberblick (1986)
WZKS	Wiener Zeitschrift fūr die Kunde Sūdasiens

Abbreviations of the titles of Pāli texts are those adopted by the CPD. Editions are those of the PTS, unless otherwise stated.

Bibliography

Adikaram, E.W., 1953, Early History of Buddhism in Ceylon, Colombo.
Allon, M., 1994, “Some oral features of the prose portions of Pali canonical sutta texts”, unpublished conference paper.
Almond, P.C., 1988, The British discovery of Buddhism, Cambridge.
Alsdorf, L., 1962, “Uttarajjhāyā Studies”, IU VI, 2, pp. 110-36.
—_, 1965, Les études jaina: état présent et tâches futures, Paris.
Balbir, N., 1993, “Titres et Canons dans l’Hétérodoxie Indienne”, in La formation des canons scripturaires, Paris CERF, pp. 47-59.
Beal, S., 1884, Buddhist records of the western world, London.
Bechert, H., 1978, Buddhism in Ceylon and Studies on Religious Syncretism, Göttingen.
—_, 1980, The language of the earliest Buddhist tradition, Göttingen.
—_, 1985, Zur Schulzugehörigkeit von Werken der Hīnayāna-Literatur, Erster Teil, Göttingen.
—_, 1987, Zur Schulzugehörigkeit von Werken der Hīnayāna-Literatur, Zweiter Teil, Göttingen.
—_, 1991A, “Methodological considerations concerning the language of the earliest Buddhist tradition”, BSR 8,1-2, pp. 3-19.
—_, 1991B, The dating of the historical Buddha, Göttingen.
—_, 1992, “The writing down of the Tripitika in Pāli”, WZKS XXXVI, pp. 4553.
Bendall, C., 1883, Catalogue of the Buddhist Sanskrit manuscripts in the University Library, Cambridge, Cambridge (reprinted Stuttgart 1992).
Bernhard, F., 1965, Udānavarga, Vol. I, Göttingen.
Bhattacharya, K., 1989, “Brahman in the Pali Canon and in the Pali Commentaries”, in Amalā Prajñā: Aspects of Buddhist Studies, Delhi, pp. 15-31.
Bloch, J., 1950, Les inscriptions d’Asoka, Paris.
Bloomfield, M., & Edgerton, F., 1932, Vedic Variants, Vol. II (Phonetics), Philadelphia.
Bode, M., 1909, The Pali Literature of Burma, London.
Bollée, W.B., 1974, “Buddhists and Buddhism in the earlier literature of the Śvetāmbara Jains”, in Cousins et al., 1974, pp. 27-39.
Brough, J., 1953, The early brahmanical system of Gotra and Pravara, CUP.
—, 1954, “The language of the Buddhist Sanskrit texts” (Review of Edgerton, 1953A & 1953B), BSOAS, 16, 2, pp. 351-75.
—, 1962, The Gāndhārī Dharmapada, London.
—, 1964, “The Chinese pseudo-translation of Ārya-śūra’s Jātaka-mālā”, Asia Major NS 11.1, pp. 27-53.
—, 1980, “Sakāya Niruttiyā: Cauld kale het”, in Bechert, 1980, pp. 35-42.
Buddharakkhita, Acharya, 1985, The Dhammapada: the Buddha’s path of wisdom, Kandy.
Burford, Grace G., 1991, Desire, death, and goodness: the conflict of ultimate values in Theravāda Buddhism, New York.
Burghart, R., & Cantlie, A., 1985, (eds.): Indian Religion, London.
Burrow, T., 1973, “The Proto-Indoaryans”, JRAS, pp. 123-40.
Burrow T., and Emeneau, M.B., 1984, (second edition of) A Dravidian Etymological Dictionary (1961), Oxford.
Caillat, C., 1989, Dialectes dans les littératures indo-aryennes, Paris.
—, 1991, Middle Indo-Aryan and Jaina Studies, Leiden.
Carter, J.R., & Palihawadana, M., 1987, The Dhammapada, New York.
Chakravarti, N.P., 1930, L’Udānavarga sanskrit, Paris.
Childers, R.C., 1875, Dictionary of the Pāli language, London.
Clough, B., 1824, A Compendious Pāli Grammar with a Copious Vocabulary in the Same Language, Colombo.
Collins, S., 1982, Selfless Persons, CUP.
—, 1990, “On the very idea of the Pali canon”, JPTS XV, pp. 89-126.
—, 1992, “Notes on some oral aspects of Pali literature”, IIJ 35, 2/3, pp. 12135.
Cone, M., 1989, “Patna Dharmapada I”, JPTS XIII, pp. 101-217.
Cousins, L.S., 1983, “Pali oral literature”, in Denwood and Piatigorsky, 1983, pp. 1-11.
Cousins, L.S. et al. (eds), 1974, Buddhist Studies in Honour of I.B. Horner, Dordrecht.
Cribb, J., 1985, “Dating India’s earliest coins”, in South Asian Archaeology 1983, Naples, pp. 535-54.
Cutler, S. Mellick, 1994, “The Pāli Apadāna Collection”, JPTS XX, pp. 1-42.
Denwood, P., & Piatigorsky, A. (eds), 1983, Buddhist Studies: ancient and modern, London.
Deraniyagala, S.U., 1990A, “Radiocarbon dating of early Brahmi script in Sri Lanka”, Ancient Ceylon 11, Colombo, pp. 149-68.
—, 1990B, “The proto- and early historic radiocarbon chronology of Sri Lanka”, Ancient Ceylon 12, Colombo, pp. 251-92.
Dhadphale, M.G., 1980, Synonymic collocations in the tipitaka: a study, Poona.
Duroiselle, C., 1911, Review of Bode 1909, in JBuRS I, 1, pp. 119-21.
Edgerton, F., 1953A, Buddhist Hybrid Sanskrit Grammar, New Haven.
—, 1953B, Buddhist Hybrid Sanskrit Dictionary, New Haven.
Ehelolf, H., 1916, Ein Wortfolgeprinzip im Assyrisch-Babylonischen, Leipzig.
Elizarenkova, T.Y., 1989, “About traces of a Prakrit dialectal basis in the language of the Rgveda”, in Caillat, 1989, pp. 1-17.
Erdosy, G., 1993, “The archaeology of early Buddhism”, in Studies on Buddhism in honour of Professor A.K. Warder, Toronto, pp. 40-56.
Geiger, W., 1994, A Pāli Grammar, Oxford.
Gethin, R., 1992A, “The Mātikās: Memorization, Mindfulness, and the List”, in The Mirror of Memory. Reflections on Mindfulness and Remembrance in Indian and Tibetan Buddhism, Albany, pp. 149-72.
—, 1992B, The Buddhist Path to Awakening, Leiden.
Gombrich, R.F., 1971, Precept and Practice, OUP.
—, 1980, “The significance of former Buddhas in the Theravādin tradition”, in Somaratna Balasooriya et al. (ed.), Buddhist Studies in honour of Walpola Rahula, London, pp. 62-72.
—, 1988, “How the Mahāyāna began”, Journal of Pali and Buddhist Studies 1, pp. 29-46.
—, 1990, “Recovering the Buddha’s message”, in Skorupski, 1990, London, pp. 5-20.
—, 1992, “The Buddha’s book of Genesis”, IIJ 35, pp. 159-78.
Goody, J. (ed.), 1968, Literacy in Traditional Societies, CUP.
Grierson, Sir George A., 1926, (second edition of) Bihar Peasant Life (1885), Patna.
Hallisey, C., 1990, “Tuṇḍilovāda”, JPTS XV, pp. 155-95.
——, 1993, “Nibbānasutta: an allegedly non-canonical sutta on Nibbāna as a great city”, JPTS XVIII, pp. 97-130.
Hara, M., 1969, “A note on the folk-etymology of rājan”, The Journal of the Ganganatha Jha Research Institute, XXV, pp. 489-99.
Hardy, E., 1902, The Netti-pakarana, London.
Harvey, P., 1990, An introduction to Buddhism, CUP.
Heesterman, J.C., 1969, Pratidānam, den Haag.
Hinūber, O. von, 1971, “Die ‘dreifache’ Wirkung des Karma”, IU 13, 4, [1972], pp. 241-49.
-_, 1977, “Zur Geschichte des Sprachnamens Pāli”, in Beitrāge zur Indienforschung Ernst Waldschmidt zum 80. Geburtstag gewidmet, Berlin, pp. 237-46.
——, 1981, ‘The ghost word dvihitikā and the description of famines in early Buddhist literature”, JPTS IX, pp. 74-86.
——, 1982, “Pāli as an artificial language”, IT 10, pp. 133-40.
——, 1986, Das āltere Mittelindisch im ūberblick, Vienna.
——, 1989A, Der Beginn der Schrift und frūhe Schriftlichkeit in Indien, Wiesbaden.
——, 1989B, “Origin and varieties of Buddhist Sanskrit”, in Caillat, 1989, pp. 341-67.
——, 1991A, “Linguistic considerations on the date of the Buddha”, in Bechert, 1991B, pp. 175-81.
——, 1991B, The oldest Pāli manuscript. Four folios of the Vinaya-pitaka from the National Archives, Kathmandu, Mainz.
——, 1993, “From colloquial to standard language: the oral phase in the development of Pāli”, in Premier Colloque Étienne Lamotte, Louvain, pp. 101-13.
——, 1994, “Die neun Añgas: Ein frūher Versuch zur Einleitung buddhistischer Texte”, WZKS XXXVIII, pp. 121-35.
Horner, I.B., 1954, Middle Length Sayings, Vol. I, London.
——, 1957, Middle Length Sayings, Vol. II, London.
Horner, I.B., & Jaini, P.S., (tr.), 1985, Apocryphal Birth-Stories (Paññāsajātaka), Vol. I, London.
Hultzsch, E., 1925, Inscriptions of Asoka, Oxford.
Jacobi, H., 1884, Jaina Sūtras, Part I, SBE XXII, Oxford.
Jaini, P.S., 1986, Apocryphal Birth-Stories (Paññāsa-jātaka), Vol. II, London.
Jayasekera, U.D., 1992, The Dhammapada, Colombo.
Jayawickrama, N.A., 1962, The Inception of Discipline and the Vinaya Nidāna, London.
Jones, J.J., 1952, The Mahāvastu, Vol. II, London.
Jong, J.W. de, 1974, “A brief history of Buddhist Studies in Europe and America (Part II)”, The Eastern Buddhist New Series VII, No. 2 (October 1974), pp. 49-82.
Kalupahana, D.J., 1986, A Path of Righteousness: Dhammapada, Lanham.
Karashima, S., 1991, in Papers in Honour of Prof. Dr. Ji Xianlin on the occasion of his 89th Birthday, Beijing, pp. 607-43.
-, 1992, The textual study of the Chinese versions of the Saddharmapundarīkasūtra in the light of the Sanskrit and Tibetan versions, Tokyo.
Kloppenborg, R., 1973, The Sūtra on the Foundation of the Buddhist Order, Leiden.
La Loubère, S. de, 1693, The kingdom of Siam, London (reprinted 1969).
Lamotte, É., 1958, Histoire du Bouddhisme Indien, Louvain.
——, 1970, “Le Buddha insulta-t-il Devadatta ?”, BSOAS XXXIII, pp. 107-15.
——, 1988, History of Indian Buddhism (English translation of 1958), Louvain.
Law, B.C., 1957-58, “The Dīpavamsa”, Ceylon Historical Journal, pp. 1-266.
Lévi, S., 1912, “Observations sur une language précanonique du bouddhisme”, J A, pp. 495-514.
Luce, G.H., 1974, “The advent of Buddhism to Burma”, in Cousins et al., 1974, pp. 119-38.
Lūders, H., 1940, “Sanskrit muktā, muktāphala, phala”, in Philologica Indica, Göttingen, pp. 179-90.
——, 1954, Beobachtungen ūber die Sprache des buddhistischen Urkanons, Berlin.
MacQueen, G., 1984, “The doctrines of the six heretics according to the Śrāmanyaphala Sūtra”, IU 27, pp. 291-307.
Malalasekera, G.P., 1935, Vaṃsatthappakāsinī, London.
——, 1937-38, Dictionary of Pāli Proper Names, 2 volumes, London.
Matsumura, H., 1989, “Āyuhparyantasūtra”, in Sanskrit-Texte aus dem buddhistischen Kanon: Neuentdeckungen und Neueditionen, Göttingen, pp. 69-100.
Mayrhofer, M., 1976, Kurzgefasstes etymologisches Wörterbuch des Altindischen, Heidelberg, Vol. III.
Monier-Williams, M., 1899, Sanskrit-English Dictionary, Oxford.
Mori, S., 1988, “Chronology of the ‘Sihala sources’ for the Pāli commentaries (II)”, Buddhist Studies (Bukkyō Kenkyū), Vol. XVII, pp. 119-67.
Morris, R., 1884, “Notes and queries”, JPTS, pp. 69-108.
Mūller, F. Max, 1881, The Dhammapada, SBE X, Oxford.
Nakatani, H., 1987, Udānavarga de Subaši, Parts I & II, Paris.
Nārada, U, 1969, Conditional Relations, London.
Norman, K.R., 1967A, “Notes on the Aśokan Rock Edicts”, IIJ X, pp. 160-70 ( = CP I, pp. 47-58 [50-51)].
—, 1967B, “Notes on Aśoka’s Fifth Pillar Edict”, JRAS, pp. 26-32 (= CP I, pp. 68-76).
—, 1970, “Some aspects of the phonology of the Prakrit underlying the Aśokan inscriptions”, BSOAS 33, pp. 132-43 (= CP I, pp. 93-107).
—, 1973, “Middle Indo-Aryan Studies X”, JOI(B) XXIII, pp. 64-771 (= CP I, pp. 160-69).
—, 1976A, “Notes on the so-called ‘Queen’s Edict’ of Aśoka”, Studies in Indian Epigraphy III, pp. 35-42 (= CP II, pp. 52-58).
—, 1976B, “Middle Indo-Aryan Studies XIII: the palatalisation of vowels in Middle Indo-Aryan”, JOI (B), Vol. XXV, pp. 328-42 (= CP I, pp. 220-37).
—, 1976C, “Pali and the language of the heretics”, A O, 37, pp. 117-26 (= C P I, pp. 238-46).
—, 1977, “The Buddha’s view of devas”, in Beitrāge zur Indienforschung Ernst Waldschmidt zum 80. Geburtstag gewidmet, Berlin, pp. 329-36 ( = CP II, pp. 1-8).
—, 1978, “The role of Pāli in early Sinhalese Buddhism”, in Bechert, 1978, pp. 28-47 (= CP II, pp. 30-51).
—, 1979A, “Māgadhisms in the Kathāvatthu”, in Studies in Pali and Buddhism (in Honor of Bhikkhu Jagdish Kashyap), Delhi, pp. 279-87 ( = CP II, pp. 59-70).
—, 1979B, “Dhammapada 97: a misunderstood paradox”, IT 7, pp. 325-31 (= CP II, pp. 187-93).
—, 1980A, “The dialects in which the Buddha preached”, in Bechert, 1980, pp. 61-77 (= CP II, pp. 128-47).
—, 1980B, “Four etymologies from the Sabhiya-sutta”, in Buddhist Studies in honour of Walpola Rahula, London, pp. 173-84 (= CP II, pp. 14861).
—, 1981, “A note on attā in the Alagaddūpama-sutta”, in Studies in Indian Philosophy, Ahmedabad, pp. 19-29 (= CP II, pp. 200-209).
—, 1982, “The four noble truths”, in Indological and Buddhist Studies (Volume in honour of Professor J.W. de Jong), Canberra, pp. 377-91 ( = CP II, pp. 210-23).
—, 1983A, “The Pratyeka-buddha in Buddhism and Jainism”, in Denwood and Piatigorsky, 1983, pp. 92-106 (= CP II, pp. 233-49).
—, 1983B, Notes on the Ahraurā version of Aśoka’s First Minor Rock Edict, IIJ 26, pp. 277-92 (= CP II, pp. 250-68).
—, 1983C, Pāli Literature, Wiesbaden.
—, 1984-85, “A dialect variation in Aśoka’s Minor Rock Edict I”, Journal of Ancient Indian History XV, 1-2, pp. 1-14 (= CP III, pp. 126-36).
—, 1985, “Pāli Lexicographical Studies III”, JPTS X, pp. 23-36 (= CP III, pp. 83-94).
—, 1988, “The origin of Pāli and its position among the Indo-European languages”, Journal of Pali and Buddhist Studies 1, 1988, pp. 1-27 (= CP III, pp. 225-43).
—, 1989A, “Common terminology in early Buddhist and Jain texts”, in Jainology: manifold facets (Pt. Jaganmohanlal Shastri Sadhuvad Grantha), Rewa, pp. 393-97 (= CP IV, pp. 264-70).
—, 1989B, “The Pāli language and scriptures”, in Skorupski, 1989, pp. 29-53 ( = CP IV, pp. 92-123).
—, 1989C, “Dialect forms in Pāli”, in Caillat, 1989, pp. 369-92 (= CP IV, pp. 46-71).
—, 1990A, “Aspects of early Buddhism”, in Ruegg and Schmithausen (eds): Earliest Buddhism and Madhyamaka, Leiden, pp. 24-35 (= CP IV, pp. 124-38).
—, 1990B, “Why are the four noble truths called ‘noble’?”, in Ānanda: essays in honour of Ananda W.P. Guruge, Colombo, pp. 1-13 (= CP IV, pp. 170-74).
—, 1990C, “Pāli philology and the study of Buddhism”, in Skorupski, 1990, pp. 31-39.
-_, 1991A, “Brahmanical terms in a Buddhist guise”, in Skorupski, 1991, pp. 193-200 (= CP IV, pp. 271-80).
-, 1991B, “Observations on the dates of the Jina and the Buddha”, in Bechert, 1991B, pp. 300-12 (= CP IV, pp. 185-201).
—, 1991C, “Syntactical compounds in Middle Indo-Aryan”, in Caillat, 1991, pp. 3-9 (= CP IV, pp. 218-25).
—, 1991D, “Death and the Tathāgata”, in Studies in Buddhism and Culture (In honour of Professor Dr. Egaku Mayeda), pp. 1-11 (= CP IV, pp. 251-63).
—, 1993, “The metre of the Buddhist Hybrid Sanskrit Upāli-sūtra”, in Studies on Buddhism in honour of Professor A.K. Warder, Toronto, pp. 11323.
—, 1994, “Mistaken ideas about nibbāna”, in Skorupski, 1994, pp. 209-24.
Oberlies, Thomas, 1993, Āvaśyaka-Studien 2, Glossar ausgewāhlter Wörter zu E. Leumanns “Die āvaśyaka-Erzāhlungen”, Stuttgart.
O’Flaherty, Wendy Doniger, 1975, Review of B.L. Ogibenin: Structure d’un mythe védique; le mythe cosmogonique dans le Rgveda, in BSOAS 38, pp. 173-75.
Oldenberg, H., 1879, The Dīpavamsa, London.
Oldenberg, H., & Pischel, R., 1883, The Thera- and Therī-gāthā, London.
Paranavitana, S., 1970, Inscriptions of Ceylon, Vol. I, Ceylon.
Pinault, Georges-Jean, 1989, “Reflets dialectaux en védique ancien”, in Caillat, 1989, pp. 35-96.
Potter, Karl H., 1981, Encyclopaedia of Indian Philosophies, Vol. III (Advaita Vedānta up to Śamkara and his pupils), Princeton.
Pruitt, W., 1987, “References to Pāli in 17th-century French books”, JPTS XI, pp. 121-31.
Rādhakrishnan, S., 1950, The Dhammapada, London.
Rahula, Walpola, 1956, What the Buddha taught, London.
Rhys Davids, C.A.F., 1974, (third edition of) Buddhist Psychological Ethics (1901).
—, 1909, Psalms of the Early Buddhists. I — Psalms of the Sisters, London.
Rhys Davids, T.W., 1881, Buddhist Suttas from the Pāli, London.
Rhys Davids, T.W., & C.A.F., 1921, Dialogues of the Buddha, Part III, OUP.
Roth, G., 1980, “Particular features of the language of the Ārya-MahāsāmghikaLokottaravādins and their importance for early Buddhist tradition”, in Bechert, 1980, pp. 78-135.
Saksena, B., 1936, “Pāli bhūnahu”, BSOS 8, pp. 713-14.
Sander, Lore, “Parişad und parşad in Vinaya- und Hīnayāna-Sūtra-Texten”, in Bechert, 1985, pp. 144-60.
Sasaki, Genjun H., 1986, Linguistic approach to Buddhist thought, Delhi.
Schmithausen, L., 1981, “‘Liberating Insight’ and ‘Enlightenment’ in Early Buddhism”, in Studien zur Jainismus und Buddhismus (Gedenkschrift fūr Ludwig Alsdorf), Wiesbaden, pp. 199-250.
—, 1987, “Beitrāge zur Schulgehörigheit und Textgeschichte kanonischer und postkanonischer buddhistischer Materialien”, in Bechert, 1987, pp. 304-81.
Schubring, W., 1962, The doctrine of the Jainas, Delhi.
Schwarzschild, L.A., 1961, “The indeclinable je in MIA”, Bhāratīya Vidyā, XXXXII, pp. 211-17.
—, 1991, Collected Articles of L.A. Schwarzschild on Indo-Aryan 1953-1979, Australian National University, Canberra.
Shukla, N.S., 1979, The Buddhist Hybrid Sanskrit Dharmapada, Patna.
Simson, G. von, 1973-, Sanskrit-Wörterbuch der buddhistischen Texte aus den Turfan-Funden, Göttingen.
Skilling, P., 1993, “A citation from the *Buddhavamsa of the Abhayagiri school”, JPTS XVIII, pp. 165-75.
Skorupski, T., 1989, The Buddhist Heritage, Tring.
—, 1990, The Buddhist Forum, Volume I, London.
—, 1991, The Buddhist Forum, Volume II, London.
—, 1994, The Buddhist Forum, Volume III, London.
Southwold, M., 1985, “Nirvana in village Buddhism”, in Burghart & Cantlie, 1985, pp. 15-50.
Stargardt, J., 1990, The Ancient Pyu of Burma. Volume I, Early Pyu cities in a man-made landscape, Cambridge.
Takakusu, J. (tr.), 1896, A record of the Buddhist religion, London.
Tambiah, S.J., 1968, “Literacy in a Buddhist village in North-East Thailand”, in Goody, 1968, pp. 85-131.
Thomas, E.J., 1949, (third edition of) The life of the Buddha (1927), London.
—, 1951, (second edition of) History of Buddhist Thought (1933), London.
Trautmann, Thomas R., 1981, Dravidian Kinship, Cambridge.
Tsuchida Ryūtarō, 1991, “Two categories of Brahmins in the early Buddhist period”, The Memoirs of the Toyo Bunko, 49, pp. 51-95.
Turner, R.L., 1973, “Pāli phāsu and dātta”, BSOAS 36, pp. 424-28 (= Collected Papers, 1975, pp. 430-33).
Vetter, T., 1985, “The most ancient form of Buddhism”, in Buddhism and its relation to other religions (Festschrift for Dr. Shozen Kumoi), Kyoto.
Vogel, C., 1970, The teachings of the six heretics, Abhandlungen fūr die Kunde des Morgenlandes XXXIX, 4, Wiesbaden.
Walleser, M., 1914, Prajñāpāramitā. Die Vollkommenheit der Erkenntnis nach den indischen, tibetischen und chinesischen Quellen, Göttingen.
Walshe, M., 1987, Thus have I heard, London.
Warder, A.K., 1970, Indian Buddhism, Delhi.
Wiltshire, M.G., 1990, Ascetic figures before and in early Buddhism: the emergence of Gautama as the Buddha, Mouton de Gruyter, Berlin and New York.
Winternitz, M., 1933, A history of Indian literature, Vol. II, Calcutta.
Young, F., 1992, (third impression of) Can these dry bones live? (1982), London.
Yuyama, A., 1969, “The bodhi tree in the Mahāvastu-avadāna”, in Heesterman, 1969, pp. 488-92.

I - Buddhism and Philology

In the autumn of 1993 I attended a conference in America on the State of the Art in Buddhist Studies. ¹ At that conference speaker after speaker said that they felt marginalised, in effect left out on the edges of their subject. In some cases this meant that their interest in Buddhism was what might be called “on the fringe”, often combined with another discipline, e.g. social anthropology; in other cases it meant that in the departments of humanities in the various universities to which they belonged they were regarded with some sort of suspicion as being teachers of subjects which were not as important as other subjects, so that when money was allocated for teaching or library purposes they got the smaller share, or when posts were sacrificed in the interests of economy it was posts in their subjects which went.

Even those who might have been thought to be in the mainstream of Buddhist Studies said that they felt marginalised when they went as part of their field work to the country of their own specific interest only to find that they were regarded in some way as outsiders by the Buddhists in those countries.

My contribution to the conference was a paper on “Pāli studies in the West: present state and future tasks”, ² delivered in a session on Textual and Philological Studies. It was greeted with no great enthusiasm. People I met at various social functions thereafter during the course of the conference said “Ah, you are Mr Norman, aren’t you? Your paper was about texts, wasn’t it ?” and hurriedly changed the subject, although one or two people did express interest, in a way which showed that they had never thought seriously about textual and philological studies before. Despite this reception, I felt in no way marginalised, as did the many participants who spoke about what I might describe as the “trendier aspects of Buddhism”. This was strange because there were very few people at that conference who would claim to be philologists first and foremost, and to the best of my knowledge I was the only philologist there who disclaimed all knowledge of Buddhism. Nevertheless, I felt at the very core and heart of the subject - although this may be the equivalent of saying that I was the only one in step - because the branch of learning which I represented, i.e. philology, was to my mind the very basis of all the work which my fellow participants in the conference were doing, although my specific area of philology, i.e. early IndoAryan dialects, bore no relationship to the geographical areas of most of the participants’ interests.

It is, in fact, not surprising if some scholars in the field of Buddhist Studies feel marginalised, because to some extent they have marginalised themselves. There is a tendency in modern scholarship to look always for the new scholars entering the field are not content to tread the paths well-worn by their predecessors, even when it is clear that the work of their predecessors needs reworking. The cry is to find something new, something which has not been done before. I am confronted with this tendency all the time. Prospective research students visit me or write to me and ask what they can do for their doctoral thesis in the field of Pāli studies. I say: “What has not been done needs to be done, and what has been done needs to be done again”. Of these the second is the more important, because, by and large, the most important Pāli texts were published first, when little was known about the Pāli language - there were only inadequate dictionaries and grammars, and only a few manuscripts had come to Europe. Now more and better manuscripts are available, and we have superior grammatical and lexicographical aids, and so vast improvements can be made in editions and in translations based on those editions, and in books on Buddhism based on those translations. Unfortunately, the enquirers all want to do something new, so the study of the fundamentals is abandoned while they go after trendy trivia which they hope will have an earthshaking effect upon the world of Pāli and Buddhist studies when the result of their research appears.

Why did I feel the way I did at that conference ? I will answer that question with another question.

If you hear about a religion for the first time, and want to know about it, how do you set about gaining the knowledge you desire? You can of course read a book about it, but if you are not entirely satisfied with the answers you get, and read another book on the subject, you may well find that although the authors seem to be talking about the same thing they are not necessarily saying identical things about it.

An alternative method is to question a follower of the religion, but you will almost certainly find that precisely the same inconsistency in the views expressed appears, if you go on to question another follower. Sooner or later any serious investigator into a religion must examine the texts which the followers of the religion say they are taking as the basis for their religion. He must see how far what the texts say and what the followers do coincide, i.e. he must measure the difference between precept and practice.

If, by chance, the religion is one which still maintains an oral tradition, that is to say its “sacred” texts have not yet been written down, then problems may arise from the fact that the modern adherents of the religion are precisely those who are the authorities for their religious beliefs, and what they say and do is their religion, i.e. the practice is the precept, or the precept is the practice.

Such a situation presents difficulties of its own, but such difficulties are mercifully absent from the study of Buddhism. Here we are faced with a situation where the founder of the religion lived approximately 2½ millennia ago. There is a large body of material in an assortment of languages which is ascribed to the founder and much of it has been edited and published in a form which makes it easily accessible and available to investigators. Those wishing to know about Buddhism can consult these ancient sources, and there is a battery of ancillary material - dictionaries, grammars, translations, etc., - which enables them to do so.

There is, of course, the problem that if we set out to understand what the earliest texts say, i.e. those ascribed to the Buddha himself, or his followers during his lifetime, we have to consider the fact that the language which we find in such texts is not necessarily, and almost certainly is not, the language of the Buddha himself, i.e. the language has been changed both synchronically - it has been translated or transformed into other languages as the need arose, perhaps as Buddhism spread into neighbouring areas - and also diachronically, i.e. as the language of the readers or recensionists developed in the course of time, this had an effect upon the language of the texts. It is also possible and indeed probable that changes took place in what the Buddha is reported to have said and done, i.e. the tradition changed, unconsciously, the Buddha’s views because, as certain words fell out of use and were no longer understood, they were “brought up to date” and made more intelligible by having an interpretation inserted into the texts in their place. The account of what the Buddha said or did might also be changed consciously by having interpolations inserted, for various reasons. Sometimes it is because a passage seemed appropriate to the context. For example, when in the Mahāparinibbānasutta the Buddha has given eight reasons for an earthquake occurring, a number of other sets of eight phenomena are added. Sometimes an interpolation occurs because a person or a city or a sect wished to have some dogma or action authenticated, and a reference to the Buddha doing something or saying something was inserted into the text to give the authentication they desired.

Much of our knowledge about Buddhism, at least in the early days, came in a rather haphazard way, as the result of reports by travellers, envoys, etc. One such envoy was Simon de La Loubère, who in 1687-88 went to Siam as an envoy of the king of France (King Louis), and on his return wrote a book about the Kingdom of Siam, which was published in France in 1691. Two years later an English translation was published in England. ³ La Loubère gave a fascinating account of many aspects of Siam and Siamese culture, and he also included an account of the Siamese religion - Buddhism - and the sacred language of the Siamese. He noted the differences between that language, which he called Balie, and the Siamese language, and noted, correctly, the relationship between the former and Sanskrit. He also gave French translations of a few Buddhist texts.

It was until quite recently believed that he probably made these translations from Siamese translations from Pāli, but it has now been shown that, before La Loubère went to Siam, French Christian missionaries were already active in that country, and had compiled dictionaries of Siamese and Pāli, now unfortunately lost, and even a Pāli preface and postface to a translation of St Luke’s Gospel into Thai, which still survives in France in the archives of the Missions étrangères. ⁴ It is therefore possible or even probable that La Loubère gained his knowledge from those missionaries, rather than from the man in the street. We may assume that those missionaries learned Pāli and studied texts in Pāli in order that they might better confute their Buddhist opponents and convert them to Christianity.

Such was, indeed, the purpose of British missionaries in Sri Lanka a little more than a century later. Men like Benjamin Clough studied Pāli and wrote grammars, together with lengthy word lists based upon indigenous dictionaries. ⁵ If the original intention was to learn the language so that the religion it supported might be demolished, we may suspect that familiarity begat not contempt but affection, and translations of the Pāli texts, e.g. the Dhammapada, which those missionaries produced did more to make English readers want to learn more about Buddhism than to convert the followers of Buddhism to Christianity. In fact the impact of Buddhism upon nineteenth century Britain reveals a very varied response, well depicted recently by Philip Almond. ⁶

It is possible, however, that such study of Pāli would have been as ineffectual outside missionary circles as the French endeavours - although they sent manuscripts back to France there is little evidence of their being studied - had it not been for the sudden rise of a new field of study: comparative philology. In the few years since Sir William Jones’ well-known statement about the relationship of Latin, Greek, Sanskrit, etc., the subject had made rapid progress.

Although Jones was not the first to notice such relationships, his pronouncement seems to have attracted more attention than the suggestions of others. The need to study Sanskrit and to obtain manuscripts of Sanskrit works led to the Danish philologist Rasmus Rask going to India and Ceylon (Sri Lanka) c. 1820 and bringing back a fine collection of manuscripts, which is now housed in the Royal Library in Copenhagen.

Other manuscripts of Buddhist texts came to the West in different ways. Brian Hodgson was the British resident in Kathmandu in the 1830’s and began to send Buddhist Sanskrit manuscripts to Europe. These formed the basis of Burnouf’s studies of Buddhism. After the conquest of Upper Burma in 1885 a large part of the royal library at Mandalay came to England. It is unfortunate that the complete holdings of the library were not sent, for those manuscripts which did not reach England at that time have been lost. Daniel Wright was the brother of William Wright, the Professor of Arabic at Cambridge. When he became surgeon to the British Residency in Kathmandu in the 1820’s, he was asked to obtain copies of Sanskrit manuscripts for Cowell, the Professor of Sanskrit at Cambridge at the time. Wright found it cheaper and easier to buy originals, which are now in the University Library at Cambridge. ⁷ Once professors began to be appointed in European universities to teach Sanskrit and Pāli, the academic study of Buddhist Sanskrit and Pāli manuscripts made great progress, and Spiegel and Fausbøll had published Pāli texts by 1855.

These editions and studies were, inevitably, made from the very limited number of manuscripts which were available to the editors, and were restricted to the tradition of the country in which they were procured, in the case of Rask and Clough to Ceylon (Sri Lanka). Only gradually did it become possible to make use of materials from other traditions, and this produced surprises. Although Fausbøll had been able to make use of Burmese manuscripts for his edition of the Jātakatthavaṇnanā, for the most part he took the readings of his Sinhalese manuscripts for the text, and gave the readings of the Burmese manuscripts in the footnotes. When, however, he came to Volume VI, the Mahānipāta, he reported ⁸ that he had not taken full notice of his Burmese manuscript B^d, as the text for the Mahānipāta in that manuscript had been very much enlarged throughout, so as to make it, in many places, very different from the Sinhalese tradition. He reported that the aim of the Burmese redactor seemed to have been to make the tale more lucid and intelligible, but the consequential difference was in many particulars so great that he expressed the hope that some scholar would give a separate edition of the Mahānipāta according to the Burmese redaction. Fausbøll had been forced to realise the need for comparative studies, on a philological basis, of parallel versions of Buddhist texts.

I have entitled this lecture “Buddhism and philology”. I might just have well have entitled it “Buddhism and the philologist”, since I want to set out the pattern which I, as a philologist, shall follow in these lectures. The course is in effect an apologia pro vita mea, explaining my encounter with Buddhism through the medium of philology, and at the same time setting out some of the things which philology has told us about the Buddha and various aspects of Buddhism: its origin, the way in which the texts were transmitted, first orally and then in a written form, the way in which a selection of texts was given canonical status, and how there was need for commentaries upon them. In my last lecture I shall suggest what further contributions philology can be expected to make to the promotion of Buddhism by giving a better understanding of the texts which lie at the heart of Buddhist Studies.

I personally came to this body of material, or at least to a portion of it, not from my desire to investigate Buddhism but from purely philological considerations. I was trained as a classicist and studied classical philology, in the form which was current in my student days, i.e. the investigation of the relationship between Latin, Greek and Sanskrit in particular, and between other Indo-European languages in general. I went on to study Sanskrit and the dialects associated with Sanskrit - the Prakrits - and was appointed to teach the Prakrits, or Middle Indo-Aryan, as they are sometimes called, lying as they do between Old Indo-Aryan, i.e. Sanskrit, and New Indo-Aryan, i.e. the modern Indo-Aryan languages spoken mainly in North India.

I was interested in the relationship between the various Middle Indo-Aryan dialects, and began to study them comparatively, trying to use forms in one dialect to shed light on forms in other dialects. Because Middle Indo-Aryan includes Pāli, the language of the Theravādin canon and the commentaries upon it, as a matter of course I included the study of Pāli in my investigations. My interest in Pāli led to my being invited to join the Council of the Pali Text Society and I was persuaded to help with the continuation of the Pāli Tipitakaṃ Concordance, which had come to a standstill after the death of its first editor, E.M. Hare. This was essentially a philological undertaking, in that it consisted of presenting for publication the contents of the Pāli canon, word by word, with the material analysed and set out by case forms, tenses, etc., as was appropriate. The words were given a meaning merely for reference purposes, but no interpretation of meaning, or discussion from a doctrinal point of view was involved.

That work in turn led to an invitation to become involved with A Critical Pāli Dictionary, based in Copenhagen. That involvement led eventually to becoming Editor-in-Chief of that project, until the portion for which I had assumed responsibility (to the end of Volume II, i.e. to the end of the vowels) was completed at the end of 1990, some 66 years after the first fascicle of the dictionary was published by Helmer Smith and Dines Andersen in 1924. That work for A Critical Pāli Dictionary was also primarily a philological undertaking, since it involved an analysis of the usage of each word listed and the assignment of meanings for each usage, but not, for the most part, a discussion of the doctrinal importance of each word and the part it played in Buddhism.

What do I mean by saying that in A Critical Pāli Dictionary there was no discussion of the doctrinal importance of each word and the part it played in Buddhism? I mean that it is sufficient in a dictionary to state, for example, that jhāna means “meditation”, without any discussion of how one meditates, or the importance of meditation to a Buddhist or to Buddhism. As a result of my activities in this field, extending over nearly twenty years, I am in the habit of saying that I know nothing about Buddhism, but I do know a little about some of the words used in Buddhism and some of the languages of Buddhism. I regard one of my purposes in life as being an adviser to those who know a lot about Buddhism, but not a great deal about the languages of Buddhism. From time to time I receive enquiries from those wishing to write about aspects of Buddhism. They tell me what they think a particular passage means, and how they propose to incorporate their interpretation into their discussion or description of Buddhism, and they ask me whether I think that their proposal is possible, philologically speaking.

The difference between a student of Buddhism and a student of Buddhist philology is very considerable. It may be possible for someone to be an expert in all types of Buddhism. Since the age of the polymaths is not quite dead, it may be possible for someone to be an expert in all types and aspects of Buddhist philology. I do not know such a person, although I have been fortunate enough in my life to meet one or two scholars who come close to it.

But I am not one of them. Such expertise as I have is confined to the dialects of Middle Indo-Aryan used primarily in North India between about 500 BC and 1000 AD , but taken from there south to Ceylon (Sri Lanka) and north to Chinese Turkestan. My knowledge of Buddhist philology is therefore restricted to Indian Buddhism, and especially to those schools of Buddhism which made use of a dialect of Middle Indo-Aryan for their texts, or used a variety of Sanskrit which was greatly influenced by Middle Indo-Aryan. Nevertheless, much of what I shall say applies to the way in which we should approach the texts which form the basis for the study of any aspect of Buddhism.

It is time to say what I mean by a philological approach to Buddhism. By that I mean approaching Buddhism not merely by language - we all do that in one way or another - but by means of what we can learn from language. Not “what does it mean ?” but “why or how does it mean it ?”. Let me give an illustration. We might, in a philosophical passage, have a sentence which says ” X arises because of $Y$ ”. We can give tentative translations of both $X$ and $Y$ , and the task of interpreters is to refine these translations until they have arrived at what they think the author of the statement meant when the statement was made. There is no need for philology in my sense of the word there. Philology in my sense is required at an earlier stage. What is the structure of the sentence ? Are we correct in thinking that it actually says ” X arises because of Y ” ? In many cases, perhaps most cases, it will be clear that the structure of the sentence is sound as it stands, and there is no need for the philologist to consider the matter further.

Let me give you another illustration from my own experience. When I was first appointed as a Lecturer in Middle Indo-Aryan Studies, I found in my very first term that there was an option available called “Sanskrit and Indo-Aryan philology” and I was responsible for teaching the Middle Indo-Aryan part of the latter, centring the course around the Aśokan inscriptions. More than that, I discovered that there were, in my very first term, candidates taking this option. And so at high speed I set about producing a course of lectures on Middle IndoAryan philology beginning with the Aśokan inscriptions. I consulted all the editions of the inscriptions available to me, read all the secondary literature on which I could lay my hands, and I produced a course of lectures during which I dealt, in my opinion, in a satisfactory way with all the many problems in those inscriptions. I showed those attending my lectures how the inscriptions should be translated and interpreted. I was so pleased with my achievement that I inserted a course of lectures on the Aśokan inscriptions into my own standard teaching. And so, year after year, I lectured on the Aśokan inscriptions, using those same notes I wrote so long ago. Well, not quite the same notes, because as my understanding of Middle Indo-Aryan philology increased, and as my appreciation of the philological approach to these matters grew, I realised that I had tried to say what the words meant, but not how or why they meant it. As I came to grapple with the question of how they meant it, I discovered that, in many cases, I did not know how the inscriptions could possibly mean what I had said they meant, and as a result of not knowing how they could mean what I had said, I had great doubts about what they did actually mean. And so my study of the Aśokan inscriptions led to a situation where every year I understood less and less. To paraphrase the words of another honest seeker after truth: the only thing I know about the Aśokan inscriptions is that I know nothing about the Aśokan inscriptions.

I can almost hear some of you thinking, “If that is what philology does for you then thank goodness I am not a philologist”. And of course there is a great deal to be said for that view. If you are setting out to explain and translate a text then it is very frustrating, even humiliating, to have to confess that you do not understand how the words can possibly mean what you would like them to mean. And consequently many translators do not confess this, possibly because they do not wish to show their ignorance, or possibly because they do not realise their ignorance. Instead they translate by intuition. If you examine many translations of Buddhist texts the impression you will gain is that the translator has looked at the words, has perhaps looked them up in the dictionary and has ascertained meanings for them, and then has used his intuition to put those words together to reveal the meaning of the Buddhavacana, the word of the Buddha. Sometimes intuition leads a translator to the right result, and sometimes not. Many translators have worked on the basis of intuition, and for the most part it has got them along very well. Often self-taught, or taught by native scholars whose standard of proficiency in teaching was not necessarily high, they have been forced, especially in the days before such aids as good dictionaries and grammars in a European language existed, to translate by the “intuitive” method, whereby they examined the context and deduced from that what the meaning must be. As has been said by Professor O’Flaherty in the context of Rgvedic studies, where the situation is very similar: “Many of our most valuable insights into otherwise obscure terms have come from scholars who have seen what the meaning must be from [the] context, from an understanding of Vedic thought processes”. ⁹ The wonder is not that these intuitive translators were sometimes incorrect, but that they were correct so often.

In what I propose to say in these lectures I will be more specific and restrict myself to a field where I think, I hope correctly, that I know a little about the subject I am discussing, i.e. the fields of Sanskrit and Pāli. In those fields, the habit of translating by intuition - I could be unkind and say “by guesswork” has been in the past, and to some extent still is, the standard way of proceeding when faced with a difficult passage. When in doubt, make a guess. This is all very well, if the guesswork leads to the right answer, but guessing does not seem to me to be the right basis for the study of anything, whether it is religion or social conditions, political affairs or history.

The more we study the work done by our predecessors the more, I am sorry to have to say, we find that our idols have feet of clay. Thomas William Rhys Davids founded the Pali Text Society in 1881, and by his editions and translations of Pāli texts and by his contribution to Pāli lexicography he probably did more than anyone else to promote Pāli studies in Europe. My attention was drawn recently to a passage in the Dīgha-nikāya, or more precisely, to Rhys Davids’ translation of the passage. It comes in the Ātānātiya-suttanta, and outlines the sort of treatment which might be meted out to any non-human creature who approached someone who knows the Ātānāta spell. One sentence in that passage ¹⁰ consists of four words or phrases; a phrase meaning “empty bowl”, a word meaning “head”, a particle which has several meanings of which the concessive “although” and the emphatic “indeed” are perhaps the most common, and a verb meaning “to bend, to turn, to turn upside down”. Rhys Davids took these words and translated them as “They would bend down his head like an empty bowl”, ¹¹ without perhaps reflecting on whether there was any significant difference between a head being turned down like an empty bowl and being turned down like a full one. In fact, the translation is wrong. Rhys Davids was translating by intuition, or as we might say, he was guessing the meaning. The words for “empty bowl” and “head” are not in the same case, which is what his translation would require, and the particle does not mean “like”. The sentence means “they would turn an empty bowl upside down on his head” - the word for head is in the locative case, and the particle is used here to give emphasis to the word which precedes it. The emphasis is upon the fact that the bowl was empty. The bowl had to be empty, so that it would slip down on to the recipient’s shoulders. The tormentors then hit the bowl with an iron bar, as the commentary explains, ¹² and the resultant reverberations would presumably do the victim’s eardrums and brain no good at all.

The method of translation which leads to this type of misunderstanding of the text is very similar to the way in which some young children read. They learn the value of some of the letters, and the context in which they find the letters on the page - they are perhaps accompanied by a picture illustrating the story enables them to work out what is going on, and so to guess appropriate words which include the letters they can recognise and will also fit the story. I am reminded of a child, who at the age of nine was diagnosed as being dyslexic, with a specific reading problem. He had been “reading” in this way, and his problem had only come to light because he had reached the stage in arithmetic where he no longer had to add 193 to 476 , but was beginning to get questions which had to be read, e.g. of the “if two men can dig a hole five feet deep in two days how long would it take ten men to dig to Australia”-variety - the sort of question which you cannot guess from the context. That is exactly parallel to the intuitive translator. He knows the meaning of some, or even all, of the words, and he deduces from the context the way in which the words should be construed. Once the context gives no help for his intuition, he is reduced to blind guesswork, and he may not always guess correctly.

You may say that the example I have quoted from Rhys Davids is a trivial matter, and think that a trifling mistranslation of this nature is of no consequence. I cannot agree with that view. The important point about this matter is that the Pāli words are incapable of bearing the translation which Rhys Davids assigned to them - a translation which, I may add, was copied, perhaps unthinkingly, by a more recent translator. ¹³ The translator gives the impression of having made no attempt to see how the words could have the meaning he gave them, or to put it more bluntly, he showed himself unable to construe the words in a simple Pāli sentence. If he adopted such a method of translating in a simple sentence where we can see his error, how can we trust him in a more important context, where the language may be more complicated, and every atom of philological expertise is needed to unravel the meaning intended by the author of the text?

The situation can be seen very clearly from the way in which some people make translations. Hardly a year goes by without a new translation of the Dhammapada appearing. There are now probably forty or more available in English alone. And yet, if one compares a new translation with its predecessors, one notes that for the most part the new translation differs only in very minor details: the order of words, perhaps, or the choice of translations adopted to serve for specific technical or semi-technical terms. Anyone reading a translation of a Pāli text is dissatisfied with the translations given for words like dhamma, kamma, nibbāna, āsava, etc., and, sooner or later, is fired with the desire to make a new translation giving what they think is the “proper” or “correct” translation of these terms. Having done this, translators believe that they have made a better translation, without giving any thought to the question of whether they have gained a better understanding of the meaning of the phrase or the sentence as a whole. It has been pointed out ¹⁴ that quite reputable translators have been known to take over translations from their predecessors, forgetful or perhaps ignorant of the fact that the text they print with their translation follows a different reading, and cannot possibly mean what they say it does.

At this point, I should make it clear that the fact that we do not see a philological problem in a text does not mean that there is no such problem, because it sometimes happens that those unversed in philology do not understand that the language they are dealing with, which perhaps they think they understand very well, does not and perhaps cannot have the meaning which they ascribe to it. It is, for example, not always understood by non-specialists that an early Pāli canonical sutta is itself a translation, and forms which were left untranslated when the Pāli recension was made from some earlier version can sometimes be identified. ¹⁵ Although it may be possible to translate such Pāli texts into English, we must, if our aim is to find out how the words can have the meaning we assign to them, first try to find out what the author actually said, i.e. we must “back-translate” the text into a form of language as close as possible to that which we believe was spoken at the time of the Buddha. ¹⁶

This involves making use of all the resources of philological and literary criticism to establish the original form of the text which we wish to translate, which requires a knowledge of the languages of North India and Ceylon at the time of the Buddha and the centuries immediately after his death. This in turn necessitates expertise not only in the Middle Indo-Aryan languages, of which Pāli is one, but also in Classical Sanskrit, Buddhist Sanskrit and Buddhist Hybrid Sanskrit, since much of the language of early Buddhist texts is related to or taken over from Sanskrit, while parallel versions of many Pāli canonical texts exist in Buddhist Sanskrit or Buddhist Hybrid Sanskrit.

A very good example of a portion of text which seems to have no philological problems is something which is fundamental to Buddhist belief - the so-called four noble truths. Everyone who knows anything about Buddhism knows these four truths well. The shortest form of them that I know is that which says simply: “Four noble truths: pain, arising, cessation, path”. ¹⁷ This is, of course, abbreviation to the point of error. The first noble truth is not pain, but the realisation that “this (i.e. the whole of existence) is pain”. I have seen “this suffering is a noble truth”, “this origin of suffering is a noble truth”, and “this cessation of suffering is a noble truth” as translations of the first three noble truths, ¹⁸ in complete disregard of the grammar and syntax. It is only when the philologist examines the problem, analyses the relationship of the words, compares other versions of the noble truths found in Sanskrit, and establishes the syntax of each phrase that the correct translations “The noble truth that ‘this is suffering’”, “The noble truth that ‘this is the cause of suffering”’, etc., become possible. ¹⁹

Once the structure of the syntax and grammar of the four noble truths has been understood, then it becomes possible to give consideration to a number of other sentences and phrases found in Buddhist texts (and other Indian texts as well), which present comparable grammatical and syntactical problems for the philologist, but not always for other people, and to suggest explanations for them which enable satisfactory translations to be given. ²⁰

What else can philology tell us about concepts which are not entirely unimportant in the field of Buddhist Studies? The word nibbāna (or nirvāna in its Sanskrit form) is so common in discussions of Buddhism that it is very often not translated, but left in its Pāli or Sanskrit form. It is, however, possible for the philologist to take the word and suggest an explanation for the fact that while it appears to come from a root meaning “to blow”, the past participle which is used in conjunction with it comes from a root of entirely different meaning although of rather similar appearance. ²¹ The connection between them, when explained in this way, enables us to understand the word play which is found with both the noun and the past participle.

We can also explain the structure of the adjectives which are commonly used to describe nirvāna and thus enable better (and more accurate) translations to be given, so that we can better understand the reluctance of the Buddha to attempt to describe the nature of nirvāna, or of someone who has gained nirvāna.

For example, one of the epithets used of nirvāna is amyta (in Sanskrit), or amata (in Pāli) “not dead”, or “deathless”, or “undying”, or “immortal”, and we therefore sometimes find the compound amatapada “the amata place” or “the place of amata” being translated as “the immortal place” or “the place of immortality”. The former translation raises the difficulty of conceiving of a place that does not die, as opposed to a place which is mortal and does die. The latter translation implies that those who have gained nibbāna live for ever, which, as has been pointed out, ²² seems to be incompatible with the rest of the teachings of the Buddha, and must therefore be an untenable view in Buddhism. It is, however, possible to gain a better understanding of the meaning of the word, if we try to find out how it gets its various meanings. We can, in the light of the distinctions we can draw between the meanings of amyta, better understand the situation with regard to nirvāna: it is ajāta “without birth” and consequently it is amyta “without death”. ²³ That is to say: no one is born there, and therefore no one dies there.

To give another example. The concept of the pratyeka-buddha is well known as the middle element in the triad: buddhas, pratyeka-buddhas and śrāvakas, and the usual translation “a buddha (awakened) for himself” satisfies most people. This usage of the word pratyeka (pacceka in its Pāli form) is, however, unusual, and it is not at all clear how it could have the meaning given to it. When consideration is given to the fact that the concept of a similar type of Buddha is also found in Jainism, but there, in Jain Prakrit, the name is patteya-buddha, then it becomes possible for the philologist to suggest an etymology for the word which can explain both the form in Prakrit and the form in Pāli - explaining the form in Sanskrit presents no difficulty since it is merely a Sanskritisation of the Pāli form or something similar to it - and on the basis of that explanation to provide a more satisfactory translation. ²⁴

The basic work of philology is, of course, going on all the time when we consider Buddhist texts. We have to ask ourselves “do we know what each word is” before we can ask “what does each word mean”, and start to think about translating the words or interpreting them. In many cases we can see that the understanding of the structure of a particular word had already been lost before the commentaries were written, because it is clear that the commentators do not understand its structure. In this connection, the discovery of Sanskrit and Prakrit Buddhist texts in Chinese Turkestan, Afghanistan and India in the course of the last hundred years has provided an invaluable tool for the philologist. Quite often these new discoveries date back to a time before the confusion or corruption or whatever it was came into being; sometimes they present a corruption of a different nature which enables us to suggest a solution to the problem. A close comparative study of (say) the Pāli Dhammapada, the Gāndhārī Dharmapada, the so-called Patna Dharmapada and the Sanskrit Udānavarga is very rewarding, since it enables us to use each and every version as a control upon the others, and we can quite often identify the corruption even when, in the absence of parallel passages, no corruption was hitherto suspected - and, more important, suggest how it came into being.

Such philological studies have enabled us to identify nominal and verbal forms which were no longer understood, to identify verbal roots which are not otherwise attested in the language, ²⁵ and with our newly-found knowledge to correct corrupt readings and restore something which we hope might approximate to the original reading, and make better and more accurate translations as a result. It is very regrettable that many of these philological discoveries are still unknown to, or are perhaps ignored by, some translators, who consequently may well make mistakes in their translations, and in their understanding of Buddhism based upon those translations.

For example, philological investigation has shown that in Pāli the ending which is identical in form with the accusative singular can sometimes, although very rarely, stand also for the accusative plural or the ablative singular. ²⁶

Nevertheless, some recent translations reveal no knowledge of an accusative plural with this ending (in -am) in Pāli. There is in the section about the wise man (Pandita-vagga) in the Dhammapada a verse ²⁷ about abandoning the kanham dhammaṃ, literally “the black dhamma”, and cultivating the sukkaṃ [dhammam], “the white (or bright) dhamma”. Some translators take these forms as singular, and translate “Having forsaken a shadowy dhamma, the wise one would cultivate the bright”, ²⁸ or “A wise man should develop a bright character abandoning the shady one”, ²⁹ or “Let the wise man leave the way of darkness and follow the way of light”, ³⁰ whereas the commentary on the verse where it occurs elsewhere in Pāli, ³¹ makes it clear that it is referring to the very common categorisation of dhammas, in the plural, as akusala and kusala. I mentioned just now parallel versions of the Pāli Dhp, and if we examine the forms of the verse which occur in them, ³² we find that they have plural forms.

Such translations seem even more remarkable when translators reveal knowledge of parallel versions, but do not take advantage of their knowledge. There is in the Flower Section of the Dhammapada ³³ a verse which describes how a bee takes nectar from a flower and then flies away without harming it. The word for “flower” (puppham) appears to be in the accusative case, and this has caused problems for some translators, who find it difficult to fit an accusative form into the sentence. Recent translators ³⁴ state that “from a flower” would be a better translation for puppham, and they point to the existence of the ablative forms puspā and puspād in the parallel texts, but do not follow their own suggestion in their translation. They do not say that puppham might actually be an ablative, and they show no hint of any knowledge of the existence in Pāli of an ablative singular in -am.

Another way in which philology, the study of why words mean what they do, can be helpful is that we sometimes find that those who made the first English translations of Buddhist texts gave a particular meaning to a word which we have for the most part followed without change ever since. When we come to look at the words themselves we find that the meanings which we have accepted for so long are very often not the only possible meanings but in some cases not even the most likely meanings. Take for example, the phrase “noble truth”, which I mentioned a few minutes ago. It has become a commonplace to talk about the four noble truths, and this is a perfectly acceptable translation of the compound ariya-sacca: ariya means noble and sacca means truth, so ariyasacca means noble truth. This translation is so common and so fixed in our minds, that it seems almost like blasphemy to have to point out that not only is this not the only possible translation, but it is in fact the least likely of all the possibilities. If we look at the commentators we find that they knew this very well. They point out that the compound can have a number of meanings. It can mean “truth of the noble one”, “truth of the noble ones”, “truth for a noble one”, i.e. truth that will make one noble, as well as the translation “noble truth” so familiar to us. ³⁵ This last possibility, however, they put at the bottom of the list of possibilities, if they mention it at all. My own feeling is that it is very likely that “the truth of the noble one (the Buddha)” is the correct translation, although we must never lose sight of the fact that in Indian literature multiple meanings are very often intended, so that it is not always possible to say that there is a single correct meaning.

To give another example: we all know at least the title of the text called Saddharmapundarika, and following Burnouf we have always translated this as “the lotus of the good law”. Once again this is not the only possibility. It could, for example, be “the good lotus of the law”, or “the lotus of the law of the good one”, or ” … of the good ones”. Since in fact we find that there is in the Dhammapada a verse talking about the good teaching the dhamma to the good, ³⁶ we could very reasonably translate the title as “The lotus of the doctrine of the good one or of the good ones”, i.e. the Buddha or the Buddhas.

Another place where the early translators were probably wrong is in the translation which is given of the title of the text called the Mahāparinibbānasutta. Rhys Davids translated this as “The book of the great decease”. This is possible, although we might wonder what would constitute a small decease. We can, in fact, be fairly certain that the translation is not correct. We find quite commonly in the Pāli canon pairs of suttas, one called mahā “great” and the other cūla “small”, so that we might have a large sutta on a particular subject and a small sutta on the same subject. And we may assume that originally the small version was in fact shorter than the large one, and represented a contraction of the large, or alternatively the large represented an expansion of the small version. It does not, however, in the forms in which we have them, always work like that. Sometimes the small sutta is larger than the large sutta, and we must assume that further contraction or expansion has taken place since the suttas received their names. It is probable, therefore, that a more correct translation of Mahāparinibbānasutta is “the large text about the decease”, and it is just chance that we do not find a Cūlaparinibbānasutta in the Pāli canon. ³⁷ I may note in passing that Rhys Davids’ translation of the title is all the more strange because the preceding sutta in the Dīgha-nikāya is the Mahānidānasuttanta, which he correctly translates as “The great discourse on causation”.

Those who mistranslate in such a way are in good company, because there is another blatant, if I may be permitted to use the word, mistranslation of a text title of exactly the same type, namely that adopted for the Mahāprajñāpāramitāsūtra, which that great Belgian scholar Étienne Lamotte translated as “Le traité de la grande vertu de sagesse” instead of “Le grand traité de la vertu de sagesse”. To do Lamotte justice we must remember that he himself realised his mistake some time before his death, but at a time when it was not deemed feasible to change the title.

Another important point which philology has been able to clarify is the translation of the noun parinibbāna which Rhys Davids translated as “decease” in the title of the sutta I have just mentioned. Because that sutta deals with the nibbāna which the Buddha obtained at death, parinibbāna is often translated as “final nibbāna”, reserving the simple “nibbāna” for the experience which the Buddha had at the time of his awakening. Because of the use of the word parinibbāna in connection with the Buddha’s death, it is assumed by some that it can only be used of nibbāna at death. This interpretation is, however, based upon a misunderstanding of the significance of the prefix pari-. It can be shown very easily that it does not imply “final”, and parinibbāna, at least in its original usage, cannot mean “final nibbāna” because there are many references in the texts to living beings who are described as parinibbuta “having gained parinibbāna”. ³⁸

Another example of an on-going mistranslation is the word kusala, which I mentioned a few minutes ago. This is usually translated “skilled” or “skilful”. The word certainly has this meaning, but in Sanskrit it is the secondary meaning, the primary meaning being “good”. In the Pali Text Society’s Pali-English Dictionary, however, “clever”, “skilful”, and “expert” are given as the first meanings, although its opposite akusala is given only the sense of “bad, evil”. The two terms kusala and akusala are frequently combined as epithets of the dhammas “the (mental) phenomena”, sometimes translated as “states”, and we often read in translations of the abandoning of “unskilled states” and the arising of “skilled states”, although translators rarely try to explain what a skilled or unskilled state is. The commentaries make it clear that we are, in fact, dealing with good and bad (mental) phenomena.

Publication of the Pali-English Dictionary was completed in 1925, so the last fascicles are nearly 70 years old, while the earlier fascicles are more than 70 years old. At the time of its publication it represented the summit of philological achievement in the field of Pāli studies, but even so it contained numerous errors, as well as misleading statements of the sort I have just mentioned. Since 1925 editions of many hitherto unpublished Pāli texts have appeared, so that it is woefully incomplete, while better editions of many of the texts which are referred to in it have been published, so that many corrections of forms and meanings are needed. The Pali Text Society is fully aware of this, and a revised version of the Pali-English Dictionary, to be called the New Pāli-English Dictionary, is in preparation. ³⁹

Until that is published, the Pali-English Dictionary is the best complete dictionary of the Pāli language we have, and despite all its failings it is still used by most Pāli scholars. I have already mentioned the way in which would-be translators have perhaps looked words up in the dictionary and have ascertained meanings for them, and then have used these meanings as the basis for their intuition. It is, however, difficult to persuade many scholars that they must be very wary of taking everything they find in the dictionary as infallible. I remember asking one scholar how he proposed to solve some of the problems which I knew existed in a text he was proposing to translate. He told me that he would give every word in his text the meanings given in the PED, and he asked me very indignantly how I expected him to translate the text if the PED meanings were not correct.

These problems arise because there is a strange sanctity attached to the printed word. In the same way that people used to say, “It must be true, because I read it in The Times”, so when I offer critical comments on suggestions which people send me, and hint that their translation of some Pāli word or phrase should be changed, they reply that I must be wrong, because they took whatever it is they have written from the PED. When I come back with the retort, “Yes, I know it is in the PED, but it is wrong, and we are going to change it in the revised edition”, they are still very reluctant to believe that anything which has appeared in print can be incorrect. I do not know what sort of persons they think the editors of the first edition were, but they have certainly imparted some sort of infallibility to them. It is quite clear that they have never read the comments about Rhys Davids in the Afterword to the Dictionary, which William Stede wrote when the final fascicle appeared, some three years after Rhys Davids’ death: “His mind was bent on other aims than dictionary work, which was not his strongest point”, or Stede’s statement about the joint editors’ motto: “Better now and imperfect, than perfect and perhaps never !”⁴⁰

There is also, I believe, a reluctance to believe that present-day scholars, mere mortals, can improve on anything in the Dictionary. I can only hope that when the revised edition appears it too will gain the veneer of infallibility which the printed word imparts. I am encouraged to think that it will do so, because I find that the same sanctity is becoming attached to A Critical Pāli Dictionary, not simply the early fascicles produced by Helmer Smith and Dines Andersen, who were perhaps the greatest Pāli philologists that Europe has produced, but also more recent fascicles. In something that I was writing recently I had occasion to quote from an article I had earlier written for A Critical Pāli Dictionary, as part of my involvement in the dictionary. However, in the time since writing the article I had thought of a way of expressing my solution to a particular problem in a slightly different way. I showed what I had written to a colleague, and he returned my material to me with the comment that he thought I should quote what A Critical Pāli Dictionary said about this problem. I could go on at some length about the light which a philological approach has shed upon many of the key words in Indian Buddhism, with a resultant improvement in our understanding of such terms, and a greater accuracy in the translations we can give for them, but I have probably said enough for one lecture.

II - Buddhism and Its Origins

In the first lecture I gave a general account of the way in which I personally became involved in the philological study of Buddhism, and I gave some illustrations of the way in which philologists have been able to shed light upon a number of important concepts in Buddhism, and as a result of the shedding of that light to understand the concepts better, to give better translations for them, and to show more clearly how they fit into the general pattern of Buddhism.

I want in this and in the remaining lectures to deal with more specific aspects of Buddhism, and to continue to show how, if we approach them through the medium of philology, always asking “why or how do these words get their meaning”, we may be able to add something to the picture which we get through other means of approach.

Since I shall in these lectures be giving translations of many of the terms I shall discuss, it would perhaps be sensible to start now by giving some idea of the approach to translation which I have evolved over the years.

It is very difficult to give a one for one translation of Sanskrit and Pāli words into English. It is very rare that one Sanskrit or Pāli word has exactly the same connotations, no less and no more, as one English word. This means that if I wish to give a more adequate translation I am forced to give a phrase in English, or perhaps even a whole sentence, or in the case of a very difficult word with a wide range of connotations, even a whole paragraph. Consequently, if I am translating a Buddhist text into English, it is very difficult to produce something which approximates closely to the meaning of the original, and yet appears in good, clear, concise and readable English.

It is for this reason that many translators do not translate the difficult words, but leave them in their original Sanskrit or Pāli form. And this is fine, for them. As they read through their translations, every time they come across a Sanskrit or a Pāli word, they know, within limits, what it means and they can mentally substitute that meaning. It is not, however, so good for the rest of us, who have little idea of what these scholars have in mind, since their translations may consist of little more than strings of Sanskrit, Pāli or Tibetan words linked together with “ands” and “buts”. Reviewers sometimes complain of translations which are so literal and so full of foreign words that they hardly read as English. ⁴¹ Such translations are of little value, and one might just as well leave the whole thing in the original. Only an expert can understand all the words left in the original language, and the expert needs no English words at all.

And so what I have favoured over the years, when translating, is to leave a minimum of these difficult terms in their original form, but the first time they occur, to include a note as lengthy and as detailed as I think is required, giving some idea of what I think the word means, and why I think it means it. In effect I am saying that, every time the reader comes across this word thereafter, he must remember to consult the relevant note to find out what it means in the context. This system does have defects. As anyone who has looked at any of my translations knows, the actual translation is only a small proportion of the book. The notes are far longer, and then there are all the indexes which will enable the reader to find where I dealt first with the problematic word.

However, in the context of lectures such as these, that system is clearly not possible, and so I shall have to give a single word translation, for the technical terms, etc., that I am talking about, but this will be simply for the purpose of identification. In no way do I mean to imply that the one word translation I give is an adequate translation of the technical term we are discussing. And, with that proviso, I will go on now to talk about the information that philology can give us about Buddhism and its origins.

We are all familiar with the account of the origin of Buddhism which we find in the Indian tradition. The Buddha-to-be was the son of an Indian king. Despite his father’s attempts to ensure that his son should see no sign of old age, sickness or death, he became acquainted with the suffering existing in the world and the advantages of the ascetic life by seeing four signs at the age of twenty-nine. He left his wife and new-born son and became a wanderer. He tried severe ascetic practices, and followed various teachers, but found that he could not obtain the goal he was seeking. By meditation he obtained nirvāna, and then began to teach to others the way which he had found to be successful, beginning with those with whom he had earlier practised asceticism.

What light can a philological approach throw on this narrative?

It goes without saying that the origins of Buddhism lie in the political, economic, social and religious environment of the time.

The political and economic picture which we gain from early Pāli texts is one where the urbanisation of the Gangetic plain was well under way. There were large, well fortified cities, with powerful rulers. Movement between those cities was easy, and trade between them was flourishing. We read of merchants setting out with large caravans, and there are frequent references to the coinage which must have facilitated the growth of trade. ⁴² The ksatriyas, the ruling class, had gained a more imposing identity - they were no longer minor chieftains - and the vaiśyas, the mercantile class, had begun to gain wealth, and power arising from that wealth, as opposed to an earlier situation where they were merely itinerant traders. This must have lead to a situation where the ksatriyas and the vaiśyas would be very’open to a religion which gave them a social position equal to, or even superior to, that of members of the brahmanical caste. It is very clear that the brahmanical caste was regarded as superior, at least by the members of that caste. As a ksatriya the Buddha might be expected to oppose the brahmanical caste and much of his teaching was devoted to defining the word brāhmana as a moral term, and denying that one became a brahman simply by birth. He insisted that it was actions which made a brahman. A consequence of the Buddha’s teaching about this was that (although there are many references to brāhmanas becoming his followers ⁴³ ) the main support for his religion came from ksatriyas and vaiśyas, particularly from the latter. This was presumably because they were wealthy and were well placed to gain merit by dāna “giving, generosity”. They also travelled widely, and were able to act as missionaries, taking the message to other vaiśya communities. It is a striking fact that, as Buddhism spread, it followed the trade routes, being propagated either by “missionary” traders or by bhikkhus who travelled with the caravans under the protection of the traders.

The Buddha was born in Nepal, and his name was Siddhattha. The traditional story states that his father Suddhodana was a king, that is to say a rāja. This word, however, may mean nothing more than a man of the royal tribe or the military caste, i.e. a ksatriya, ⁴⁴ and in this context, in a place some way away from the Gangetic plain, it is probable that it still meant a minor tribal chieftain, at the head of the Śākya clan. Siddhattha’s gotra name was Gotama, but Gotama is not a ksatriya name, so it probably represents a borrowing of the family purohita’s gotra name ⁴⁵. This suggests that the Śākyas were a fairly recent entrant into the caste system, which in turn suggests that perhaps the Buddha’s family was not in origin Indo-Aryan. There are other examples of clans or tribes being assimilated into the caste system in a comparable way. ⁴⁶ It has been pointed out that important parts of the commentarial tradition concerning the Buddha’s family relations followed Dravidian marriage patterns, which is taken, by some, as proving that the commentarial tradition must have been composed in areas where Dravidian marriage patterns prevailed, i.e. in the southern half of India or in Sri Lanka. ⁴⁷ The tradition, it is suggested, must therefore be a late story from South India. It is, however, quite arguable and, I think, more likely that it represents the actual clan relationships at that time among non-Indo-Aryan tribes in the North, or among tribes which had until recently been non-IndoAryan.

What do we know about the religious beliefs of the Buddha’s contemporaries, those beliefs which he found around him? When we come to examine the teaching of the Buddha we find that we can make certain deductions about those beliefs from his reactions to them. Some beliefs he accepted as they were, others he accepted nominally, but gave a changed meaning to them, others he opposed outright.

There are, for example, traces in the Pāli texts of a belief which we may assume goes back to an early date. We find references to what one does in this world, and what one consequently experiences as a result of that action, when one has “passed away” (pretya). For example, when one has passed away one who has done good rejoices, and one who has done evil laments. ⁴⁸ This seems to go back to an earlier reward-and-punishment idea of the after-life. There is no implication that this rejoicing or lamentation will be endured again and again for the whole of eternity. A simple reference in a Pāli sutta ⁴⁹ to going “beyond this shore and the far shore”, caused great problems for the commentators. My personal belief is that this statement was first formulated in a situation where the author was considering two stages only, i.e. this world and the afterlife, rather than the endless stream of samsāra. The commentators, ⁵⁰ however, found the statement difficult to explain, because when they wrote many centuries later, this shore and the far shore meant samsāra and nirvāna, and to pass beyond nirvāna was a Mahāyāna idea which had no place in a Theravādin text.

At the same time we find abundant evidence that this earlier stage of religion had been replaced, or at least complemented, by a belief in samsāra, the journeying on in a series of existences with no beginning and no end, with the precise nature of each existence dependent upon the actions which one had done in a previous existence, i.e. one’s karma. The origin of this doctrine of reincarnation is not clear. There are those who think that the Indus Valley civilisation was the source. It can nevertheless perhaps be explained as a development of the old idea that there was reward or punishment at the end of life. It is possible that the idea grew that the reward or punishment could be another (good or bad) existence, and if, at the end of that existence, there was an imbalance in the reward or punishment, yet another existence, and so on and so on. ⁵¹ Others think that the idea was imported into India from some other culture, perhaps from the Greek cities of Asia Minor, where we know that the idea of a cycle of births was current. Whether there was any connection between these two areas and, if so, which way the borrowing went, is dependent to some extent upon the dates which we assign to the first appearance of the idea in both areas.

The Buddha accepted the Vedic gods, but although he accepted their existence, he denied them any causal role in the universe. When they had been deprived of their causal role, there was no longer any point in sacrificing to them, and they became simply super-men, enjoying the benefits of the good karma they had amassed in a previous existence. ⁵² The Buddha made it clear that the devas, like all others in samsāra, were subject to death and rebirth. As super-men, they had a longer life-span than men, and Sakka, the king of the devas, had a longer life-span ⁵³ than the other devas. Nevertheless, there was an important difference between devas and men. When a man died, that was the end of him in that existence; at the death of a king, his successor, a different person, took his place. Not so with the devas. When Sakka dies, as he must do, ⁵⁴ being finite, his place is immediately taken by another Sakka. ⁵⁵

There are references to an individual being reborn as Sakka a number of times. The Buddha, for example, stated that he had been reborn 36 times as Sakka. ⁵⁶ The commentator Buddhaghosa tells a story that Sakka once died while listening to a discourse from the Buddha, and was immediately reborn again as Sakka, ⁵⁷ so that he could continue to hear the sermon. It is clear that the singularity of this occurrence was not lost upon Buddhaghosa, for he proceeds to explain how it was that this was not noticed by the other devas. ⁵⁸

We find echoes of Upanisadic statements in the Buddha’s sermons, and it would therefore seem likely that any technical terminology he employed which has parallels in the Upaniṣads would be heard by those who were already conversant, if only to a limited extent, with the Upaniṣadic usage. It is, for example, clear from the way in which the Buddha was able to assume that his hearers understood such concepts as nicca “permanent”, anicca “impermanent”, sukha “happiness”, and dukkha “misery”, ⁵⁹ that they had already heard teachers speaking about such things. In the absence of any evidence to the contrary, it would therefore seem likely that any mention the Buddha made of attā “self” and anattā “not self” would be interpreted, rightly or wrongly, by his hearers in the light of contemporary usage, and that, as far as we can tell, is Upaniṣadic usage.

This being so, it is hard to see why almost all writers about Buddhism accept the statement often made that the Buddha makes no mention of the Upaniṣadic concept of a Universal Self, an ātman or brahman. ⁶⁰ When the Buddha stated that everything was anattā “not self”, we should expect that the view of attā “self” which he was denying was that held by other teachers at that time. We can, in fact, deduce, from what the Buddha rejected, the doctrine which the other teachers upheld. Once we know that the Buddha was using words in this way, then we are aided in our attempt to understand them and translate them.

The Buddha vigorously denied the brahmanical idea of the existence of the ātman, the idea that there is no difference between us and a world spirit, the standard advaitavāda “non-dual” doctrine, ⁶¹ which is expressed in the Upaniṣads with the words tat tvam asi “You are that”. You are that. You are identical with that world spirit - the view that we all have a portion of that spirit in us and when everything which hides that identity is removed then we can be absorbed into ātman/brahman.

The Buddha, on the other hand, specifically condemned the view that the world and the self were the same thing, and that after death one might become permanent, lasting, eternal and not liable to change, and he rejected the idea that one could look at the various aspects of the world and say “that is mine, I am that, that is my self’, which is a clear echo of tat tvam asi, expressed from a different point of view. When the Buddha said: “rūpa ‘form’, etc., are not mine”, ⁶² he was denying the view that there is no distinction between knower and known.

The Buddha’s rejection of the existence of the attā, i.e. his view that everything was “not self” (anattā), was based upon the brahmanical belief that the ātman was nitya “permanent” and sukha “happiness”. Hence the Buddha could refute this by pointing out that the world, which was supposed to be part of ātman, was in fact anicca “impermanent” and dukkha “misery”⁶³ - his belief that the world was dukkha was, of course, the first noble truth.

The Buddha’s teaching about this is, however, not always understood. The word anattā is sometimes translated “having no soul”, and various things which are specified to be anattā are thought of as having no soul. We can see the need for better understanding of this vital concept of Buddhism when we survey the range of translations given for the phrase sabbe dhammā anattā which occurs in the Dhammapada ⁶⁴ and elsewhere: “All forms are unreal” (Max Mūller); ⁶⁵ “All the elements of being are non-self” (Radhakrishnan); ⁶⁶ “All things are not self” (Acharya Buddharakkhita); ⁶⁷ “All phenomena are nonsubstantial” (Kalupahana); ⁶⁸ “All things are ego-less” (Jayasekera); ⁶⁹ “All dhammas are without self”, says a very recent translation (Carter and Palihawadana). ⁷⁰

Translations such as “without self” and “having no soul” cannot be correct, because the grammar and syntax show that anattā is not a possessive adjective, which it would need to be to have such a meaning. It is a descriptive compound, and if the correct translation for attā is “soul”, then the word would mean that these various things are “not soul”. This, however, cannot be correct, because the Buddha sometimes exhorted his followers to regard these things as parato, i.e. “as other”. We cannot, however, co-relate “as other” and “not soul”. It is clear that the only translation which it is possible to co-relate to “as other” is “not self”. We are not to regard these things as part of the self, and to clarify the point the Buddha asked his followers whether, when they saw wood being burned, they felt any pain. The answer was “No”, and the explanation given was that they did not feel any pain because the wood was not part of the self. ⁷¹ It was other than the self. We might regard the Buddha’s refutation of the ātman/brahman idea as being somewhat empirical - as was Dr Johnson’s refutation of Bishop Berkeley’s theory of the non-existence of matter - but it was no less effective for all that.

The Buddha also rejected the alternative brahmanical view of ātman as brahman. There seems to be no occurrence in Pāli of the uncompounded neuter word brahma ⁷² in the sense of the Upaniṣadic brahman, but brahma is used in compounds, apparently in the sense of “excellent, perfect”. In its basic brahmanical sense brahma-carya means “the practice of a brāhmana”, i.e. the living of a celibate life, learning the Vedas. The Buddha used the phrase in the more general sense of “to live the best life, i.e. a holy, celibate (or in the case of married couples, a chaste and moral) life”. In the Upaniṣads brahma-patha means “the way to brahman or Brahmā”. The Buddha used it in the sense of the way to the best, i.e. nibbāna, and it is explained as being the same as brahmavihāra. ⁷³

It is possible that brahma-vihāra was in origin a brahmanical term. ⁷⁴ It would literally mean “dwelling in brahman or with Brahmā”, although it is not attested in that usage in Sanskrit. It perhaps shows a trace of its original meaning in a sutta ⁷⁵ in which the Buddha speaks to young brahmans who were disputing the correct way to obtain brahma-sahavyatā. In the context this would seem to mean “union with brahman”, but the Buddha, perhaps jokingly, interprets it as meaning a state of union with the god Brahmā. He explains that someone who practises the four types of concentration ⁷⁶ called brahma-vihāra is reborn as a Brahmā in the Brahma-world. ⁷⁷ It is to be noted that this means only being born in the same heaven as Mahā Brahmā, not union with the Upaniṣadic brahman.

Contemporary with the Buddha there was a growth of non-brahmanical śramana “ascetic” movements, and we find in Buddhist texts a list of names of six teachers and something about their teachings. ⁷⁸ This seems to be a very old list because the texts are not consistent about which beliefs they ascribe to which teacher, ⁷⁹ and as we have them they are an unreliable guide to what was really going on at the time of the Buddha. We can, however, see that some of the teachings were a reaction to the idea of samsāra, the endless series of rebirths. We can then see a pattern of development of thought in Indian religion: first, the view that there is a single existence at the end of which one was judged and punished or rewarded; then a view that there is an endless series of punishments or rewards in an endless series of existences; this led on to attempts to gain release from this endless series. We can see that some teachers taught a way out of it, by specifying that samsāra is finite, so that, when we have finished a certain number of rebirths, that will be the end. Others taught various methods of gaining mokṣa “release” from the endless series of rebirths.

The Buddha’s way to release, as we shall see, was by means of meditative practices, and this method followed closely and was developed from, it seems, the teachings of other śramanas. The tradition tells us that he went to two teachers, Ālāra Kālāma and Uddaka Rāmaputta, but he left them both after a short while because their teachings did not lead to the goal he desired, and they had only reached certain stages in their meditative practices. He went away, and one version of the tradition ⁸⁰ tells us that he remembered that as a boy he had entered the first jhāna, the first state of meditation. This would seem to presuppose that there was a series of meditative practices which had already been categorised and numbered or - perhaps more likely - that the Buddha was employing a later categorisation and imposing it on an earlier occurrence, i.e. he had had some sort of meditative experience, as a boy, which he equated with the first stage of the code of meditative practices we read about later on. Repeating his boyhood experience, the Buddha then went on to a second and a third and a fourth jhāna. I would personally doubt that at this early time the four jhānas were so rigidly delineated. I would assume that his meditative experience simply flowed on, and it was only later, when the Buddha came to teach these meditations to his followers, that they were codified and categorised as the four rūpa-jhānas “the meditations about form”.

From the fourth jhāna he gained bodhi. It is not at all clear what gaining bodhi means. We are accustomed to the translation “enlightenment” for bodhi, but this is misleading for two reasons. First, it can be confused with the use of the word to describe the development in European thought and culture in the eighteenth century, and second, it suggests that light is being shed on something, whereas there is no hint of the meaning “light” in the root budh- which underlies the word bodhi. The root means “to wake up, to be awake, to be awakened”, and a buddha is someone who has been awakened. Besides the ordinary sense of being awakened by something, e.g. a noise, it can also mean “awakened to something”. The desire to get the idea of “awakened” in English translations of buddha explains the rather peculiar Victorian quasi-poetical translation “the wake” which we sometimes find. ⁸¹

It is not clear what the Buddha was awakened to, or at what particular point the awakening came. In some texts he stated that he was awakened to the destruction of the āsavas “the influxes”. He was therefore khīnāsava “one who has destroyed his āsavas”, an epithet of an arahat. Elsewhere the Buddha said that he was awakened to the knowledge and insight that this was his last existence.

The shortest account of the Buddha’s bodhi in Pāli is that found in the Ariyapariyesanasutta, ⁸² and for that reason some scholars believe that this is the earliest account available to us. ⁸³ We may assume that in the shortest account of his bodhi the Buddha would deal with the most important part of the experience, and it appears fom this version that this was the gaining of nibbāna. This view is supported by the fact that he left the teachers he had before his bodhi, because their teachings were inadequate, as they did not lead to nibbāna. Of each of them he said, “This doctrine is not conducive to disgust (with the world), nor dispassion, nor cessation, nor quiescence, nor super-knowledge, nor awakening, nor nibbāna”. ⁸⁴

We may deduce from this that the concept of the attainment of nibbāna existed, even though the Buddha-to-be and his teachers were unable to achieve it, that is to say that people knew that there was a state to which they gave the name nibbāna, even though they could not attain it. Whether this meant that some had indeed already attained it, but had not passed on their method to others, or whether it was a concept of some sort of utopia which had been proposed and named, and towards which men were struggling, is not clear.

We may also deduce that the words in the Buddha’s statement are in the order in which the various states mentioned in it are to be realised, starting with disgust with the world, and going on to awakening and nibbāna. This would support the belief that the Buddha’s aim was to free himself from samsāra, and all aspects of his teaching were concerned with the acquisition of means to do this, either in this life or a later one, and with finding out how best to dwell in samsāra until release was obtained. ⁸⁵

There is an interesting point which arises in connection with the four jhānas which the Buddha practised at the time of his bodhi. I have already mentioned the account of the Buddha’s pre-bodhi visits to Ālāra Kālāma and Uddaka Rāmaputta. With them he practised meditation and reached “the state of nothingness” (ākiñcaññāyatana) ⁸⁶ and “the state of neither perception nor nonperception” (nevasañña-nāsaññāyatana), ⁸⁷ respectively. It is strange and noteworthy that, although he rejected both of these as not leading to nibbāna, nevertheless in his own teaching after his bodhi he included them as stages on the way to nibbāna. As taught by the Buddha, they are the third and fourth of the arūpa-jhānas “the meditations about non-form”, i.e. they are devoted to the contemplation of things outside the physical world of form in which we live, and they are therefore the seventh and eighth of the samāpattis “attainments”, since they come after the four rūpa-jhānas “the meditations about form”. It would appear, then, that the Buddha had already attained the first four attainments with those teachers before he gained the seventh and eighth attainment, and we have the statement of the commentator Buddhaghosa to this effect. ⁸⁸ This would make the story of his boyhood memory seem very strange, and we should perhaps follow the view that the four rūpa-jhānas and the four arūpa-jhānas were originally two quite separate sets of states of meditation.

It would seem very probable that the four arūpa-jhānas were not discovered by the Buddha, and were not in origin Buddhist, and that is why they were included in the accounts of the non-Buddhist teachers’ views. If the suggestion of some scholars that the story of the Buddha being taught by these teachers has no historical basis, we must conclude that the inclusion of a mention of the arūpa-jhānas in the Buddha’s life history was intended to show that they were inadequate, when compared with the Buddha’s method. They did, however, lead to a state which seems to be equal to nibbāna, which presumably means that some, at least, of these non-Buddhist teachers had also succeeded in finding a way out of samsāra. It was possibly because the arūpa-jhānas were successful in gaining the desired end that they were incorporated into the Buddhist scheme of jhānas, not as simultaneous means (which would have been better, because they are really an alternative) but as consecutive.

In the Buddha’s accounts of the eight attainments, however, we read of a ninth state, that of the “cessation of feelings and perceptions” or “cessation of the feeling ⁸⁹ of perceptions” (saññāvedayitanirodha). ⁹⁰ In this state, for one seeing with perceptive knowledge (paññā), the āsavas are destroyed (paññāya c’ assa dīsvā āsavā parikkhīnā honit). ⁹¹ This would seem to imply that, if we equate the destruction of the āsavas (āsavakkhaya) with nibbāna, this was another way of attaining nibbāna, and some scholars believe ⁹² that this state of the “cessation of feelings and perceptions” (saññāvedayitanirodha) and nibbāna were originally identical.

We must, however, note that there is no reference to the four arūpa-jhānas in the accounts of the Buddha’s own attainment of nibbāna at the time of his bodhi. In the story of his death, in the Mahāparinibbānasuttanta of the Dīgha-nikāya, ⁹³ we read that the Buddha went through all the stages of the rūpa-jhānas and the arūpa-jhānas, and then entered the “cessation of feelings and perceptions” (saññāvedayitanirodha). He was then thought by Ānanda to have attained nibbāna. ⁹⁴ Anuruddha, however, pointed out that he had only attained the “cessation of feelings and perceptions”, ⁹⁵ which clearly, as far as Anuruddha was concerned, was not identical with nibbāna, but was probably some sort of death-like trance. From there the Buddha went back, in due order, to the first jhāna, and then up to the fourth jhāna, from which he died, and presumably attained nibbāna.

The root budh- in Sanskrit means not only “to be awake, to be awakened”, but also from the earliest texts onwards “to perceive, to notice, to learn, to understand,” and buddha in non-Buddhist texts means “intelligent, clever, wise”. Bodha as an adjective means “knowing, understanding”, and bodhi probably has the idea of “knowledge, understanding”, possibly the knowledge that release from samsāra is possible. It is also very likely that the knowledge itself is efficacious, i.e. “I know that it is possible to be released and merely by knowing I am released”. This theory fits in very well with the fact that immediately after his awakening the Buddha rehearsed the 12 -fold pratitya-samutpāda, the chain of interdependent elements, the so-called chain of dependent causation - the arising of things dependent upon other things.

The 12 -fold chain of causation is probably not the earliest form of the chain, and certainly portions of it with less links are found elsewhere in the canon. In discussing the 12 -fold chain I am not implying that that was necessarily the form in which the Buddha rehearsed it immediately after the awakening, but it is clear that that was the standard form at some time or other. It is interesting because it seems to be a calculated attempt by the Buddha to work out what he had just achieved. It goes, as is well known: avijjā “ignorance” produces sañkhārā “compounded formations”, these produce viññāna “consciousness”, this produces nāmarūpa “name and form”, this produces salāyatana “the six senses”, they produce phassa “contact”, this produces vedanā “feeling”, this produces tanhā “craving”, this produces upādāna “clinging”, this produces bhava “existence”, this produces jāti “birth”, and this produces jarāmarana “old age and death”.

In other places the chain is set out in reverse order, and I think that that is how the Buddha must have enunciated it when he was trying to explain what had happened to him. It is a statement of where the Buddha was, and how he had got into that situation. It ends, in the traditional form, with old age and death. That I think is where the Buddha started. He started from the position in which he found himself: he knew that he would grow old and die. Why was he going to grow old and die? Because he had been born. Why had he been born. Because of existence. Why was there existence? Because of clinging. Why was there clinging? Because of craving. Why was there craving? Because of feeling. Why was there feeling? Because of contact. Why was there contact? Because of the six senses. Why were there six senses? Because of name and form. Why was there name and form? Because of consciousness. Why was there consciousness? Because of the compounded formations. Why were there compounded formations. Because of ignorance. That is to say: the beginning of all this existence, which we know is suffering, is ignorance. We can now see what has happened to the Buddha. If any link in the chain is removed, then whatever depends upon it cannot arise. It is very clear that if I am not born, then I cannot die. If we look at the pratitya-samutpāda as a whole we can see that if there is no ignorance, then there are no compounded formations, and therefore everything else which depends upon the compounded formations will not arise. If instead of “ignorance” we translate avijjā as “lack of knowledge”, then we can see that avijjā is destroyed by vijjā “knowledge”, and knowledge is what the Buddha had just acquired. His knowledge, his bodhi, has therefore destroyed all the subsequent links of the chain. He has therefore destroyed future birth, and can exclaim triumphantly: “This is my last birth. I shall not be born again”.

One of the most interesting aspects of Buddhism and one of the strong points of Buddhism - one which undoubtedly appealed to converts who were ksatriyas and vaiśyas - was that it presented two ways to salvation. This perhaps reflects the fact, which I have already noted, that the accounts are not consistent in saying what the Buddha was awakened to. One way, the immediate way, was the one which the Buddha himself had employed, the so-called jhānic way, by meditation. This was in effect, although not in theory, restricted to those who had abandoned the world to become wanderers, and who had the time, and the inclination, to meditate. When the Buddha began to preach, however, he preached about the four noble truths, not about the destruction of the āsavas. The fourth noble truth is about the path which leads to the destruction of suffering, and this was a more gradual way to release, making use of the precepts of the eight-fold path to gain a better rebirth. This is the so-called kammic way. One might hope in time, after entering the stream, by amassing good kamma, to get to the point where the number of future birth in this world would be limited. Rebirths after that would be heavenly rebirths, leading at last to release from samsāra.

The whole point about Buddhism is that by its very nature it requires these two methods of attaining release. The system whereby some abandon the world and become wanderers in the hope of gaining nibbāna in that same life can only work as long as there are those who have decided not to abandon the world, not to become wanderers, but to continue as householders and make donations to the community of begging wanderers, the bhikṣus or bhikkhus. A system which depends upon dāna, “giving, generosity”, depends upon there being those who are able to make dāna, to give generously, to be donors.

Another of the other non-brahmanical śramana movements was Jainism, or to be more precise the teaching of Vardhamāna, known as Mahāvīra, or the Jina “the conqueror”. Although we get a picture of Jainism from the Pāli canonical texts which indicates that Jainism and Buddhism were strongly opposed to each other, of all the śramana religions of which we have knowledge, Jainism comes closest to Buddhism, in a number of ways. When Western scholars began to investigate Buddhism and Jainism in the nineteenth century they found that the two religions had so much technical and other terminology in common, that Jainism was in fact thought to be an offshoot of Buddhism, although the precise meanings of such terms did not always coincide. ⁹⁶

One such word is āsava, ⁹⁷ for the Buddhist usage does not fit the etymology of the word, while the Jain usage does. The etymology of this word (the preposition ā “towards” + the root sru- “to flow”) implies something flowing in, and this suits the Jain usage well, since there the āsavas are influences which flow into a person, and discolour his soul. ⁹⁸ We find illustrations of this in Jain manuscripts, with people ranging from white, through yellow, red, blue and green to black, depending on the amount of āsavas which has flowed into them. This does not suit the Buddhist idea, where the āsavas are not attributes which are capable of flowing into a person. They are, in fact, identical with the four oghas “floods”, ⁹⁹ and it seems clear that in Buddhism the word has lost its original meaning. So, although the translation “influence” or “influx” suits the Jain usage well, on etymological and exegetical grounds, it is not entirely satisfactory for Buddhism. This accounts for the number of translations which have been suggested for the word, including “passions”, “intoxicants”, “cravings” and “cankers”. ¹⁰⁰ The latter to me is a disease of dogs’ ears and of roses, and I am always surprised when I find the arahat, the one whose āsavas have been destroyed (khināsava), being described as “canker-waned”.

A comparative study of the terminology of the two religions gives some idea of the religious and cultural background in which Buddhism and Jainism came into being. The explanation for such parallels in terminology as āsava can sometimes be seen as a borrowing from one religion to the other or, perhaps more often, a common borrowing by both from a third religion or from the general mass of religious beliefs which we may assume were current at the time the two religious leaders lived, i.e. the beliefs of the śramanas.

It is to this general background of religious thought that we can probably assign most of the vocabulary of the ascetic type of religion, e.g. such words as śramana “ascetic”, pravrajyā “going forth”, pravrajita “one who has gone forth”, tapas “mortification”, and ṛṣi “sage”, found in both Buddhism and Jainism, and we may assume that these were terms which were common to many of the religious movements, in which the adherent went forth from the life of a householder and became a wanderer. Much of the terminology used for their religious experiences was also common to the two religions. They had terms such as nibbāna (nivvāna) in common and there is the strange fact that they both use with it the past participle of another root (nibbuta, nivvua), with a different meaning, which suggests that, just as I have suggested that nibbāna as a concept was pre-Buddhist, the word-play on the two words was also earlier than both religions.

It was long ago noted ¹⁰¹ that the Buddhists and Jains “give the same titles or epithets to their prophets”, e.g. (in their Sanskrit forms) arhat, Sugata, Tathāgata, Jina “conqueror”, Mahāvīra “great hero”, sarvajña “omniscient”, Siddha “perfected”, Buddha “awakened”, Sambuddha “id.”, parinirvrta “gained parinirvāna”, mukta “released”.

Two words in this list in particular merit close attention, namely Buddha and Jina. We are accustomed to think of these words as distinctive of their religions. Buddhism is called after the Buddha, and Jainism after the Jina. If this specific distinction was attached to the words at the time of the founding of the two religions, one might have expected that the other religion, in each case, would have avoided the words, or at least have said “our buddha is better than your buddha” or “our jina is better than your jina”. From the fact that they did not do so but continued to use the terms Buddha and Jina in both Jainism and Buddhism, we may deduce that these words were in common use prior to the origin of both religions and were taken into both of them with a non-distinctive sense. That is to say that there were those who were spoken of as buddha “awakened” to some sort of truth - doubtless about the possibility of release from samsāra, and those who were called jina “conqueror”, doubtless conquerors of samsāra, before the words were taken over into both religions, and it was then a matter of the historical development of the terminology of both religions that the specific distinction which those words denote now in those two religions arose. That is to say, there were buddhas and there were jinas before the beginning of both Buddhism and Jainism. The fact that Gotama was not the first buddha, and Mahāvīra was not the first jina helps us to understand how both religions evolved a theory of previous Buddhas and Jinas.

It is not known where the idea of a specific number of previous prophets came from, but it may be no coincidence that the Jains have 24 Jinas, while the Buddhists have 24 previous Buddhas, ¹⁰² plus Gotama Buddha. The addition of three extra Buddhas, which we find in the Buddhavamsa, is clearly a late extension of the general idea in Buddhism.

I have already mentioned in my first lecture the fact that the philological explanation for the difference between Pāli pacceka-buddha and Prakrit patteyabuddha lends support to the idea that the concept of this particular class of buddhas was also earlier than Buddhism and Jainism. The information which the two religions give about these buddhas, ¹⁰³ including the details down to the names and the causes of their awakening, suggests that the concept was also something which was part and parcel of the śramana movement and not exclusive to any one religious group, and was taken into those two religions from a third, earlier, source.

From the Pāli texts then we see that there was simultaneously a brahmanical orthodoxy and a non-brahmanical śramana movement. At a slightly later time Aśoka could refer to all religious persons as being either brāhmana or śramana, as the compound brāhmana-samana which he uses in his inscriptions shows. That śramana movement produced not only religious ideas which went against the brahmanical way of thinking, but also literature making the same point. We find that both the Buddhist Jātakas and the Jain Uttarajjhayana-sutta contain stories referring to the way in which brāhmanas mis-treated śramanas who came to their sacrificial enclosure to beg for food. The stories clearly had no specific class of śramana in mind, which made it easy for both religions to take over such stories and incorporate them into their collections of texts. Both religions shared texts (once again, probably of a common origin) defining a brāhmana by his conduct, not by his birth. Such literature emphasised the fact that it was support of the śramana movements which would bring merit, not support of the brahmans.

The Jains too objected to the brahmanical idea of the ātman, and refuted it in a way as empirical as the Buddhists. There is a Jain verse which says : “As a whole mass of earth, with all its manifold nature, is seen as one, so the whole world, with all its manifold nature is seen as the intelligent principle. Some fools, intent on their (bad) activities, say that it is so with the individual. (But) the individual who does an evil deed goes by himself to a harsh misery”, i.e. if we were all part of the same intelligent principle, namely the ātman, we as a whole would be responsible for evil deeds as a whole, and would suffer the punishment for them as a whole. It does not happen like that. Only the evil-doer suffers the punishment.

Nevertheless, the Jains differed from the Buddhists in believing that, although there was no world ātman, there was nevertheless a permanent, everlasting, individual ātman, which was the element which transmigrated. The Jains were able, therefore, to say that when the personal ātman gained release it was called siddhattha “that which has gained the goal”, and it went to the place of siddhi “perfection” at the top of the world, where all the siddhas “perfected ones” went. In the diagrams of the world which we find sometimes in Jain manuscripts, the universe is very often depicted in the shape of a man, with the earth at his waist and the hells below the waist and the heavens above the waist, the place of siddhi is right at the top, on his forehead.

The Buddha, on the other hand, had problems. Not only did he deny the existence of a world attā, but he also rejected the idea of a permanent individual attā. It was therefore very unclear what it was that transmigrated, and he found it very difficult to describe the state of one who had gained nibbāna. It would probably be more accurate to talk of the non-state of one who had gained nibbāna, because it is not possible to say what nibbāna is like, or what someone who has gained nibbāna is like. How can you describe the condition of a nonbeing in a non-state? It is that conundrum which led to the Buddha refusing to discuss the existence or non-existence of the Tathāgata after death. All we can do is say what nibbāna is not like. It is not like samsāra. Consequently it is often defined in terms of negatives or opposites. It is “blissful” (siva) or “happy” (sukha) as opposed to the dukkha of existence. It is “unmoving” (acala) as opposed to the endless movement of samsāra. It is “without death”¹⁰⁴ (amata) as opposed to the repeated deaths of samsāra. It is “without birth” (ajāta), “without beings” (abhūta), “without made things” (akata), and “without formed things” (asankhata) as opposed to the world, which has birth, beings, made things and formed things.

I said at the beginning of this lecture that the origins of Buddhism lie in the political, economic, social and religious environment of the time.

What time? I do not wish to say much about the date of the Buddha, but I must say something about it, because it has some bearing upon what I have been discussing. There are various ways of calculating the date of the Buddha’s death, and the one which is perhaps most commonly accepted in the West is c. 486 B.C.E. This depends upon a statement found in the Pāli chronicles that the Emperor Aśoka was consecrated 218 years after the death of the Buddha. Aśoka in his inscriptions mentions the names of a number of contemporary Greek kings, which enables us to fix Aśoka’s dates within fairly narrow limits, and by adding 218 to the date we can calculate for his consecration (c. 268 B.C.E.) we get a date c. 486 B.C.E. The Pāli chronicles also give the regnal years of the kings of Magadha between the death of the Buddha and Aśoka, and these dates too support the theory of an interval of 218 years, although they allow only 22 years for the ten sons of Kālāsoka and a similar period of 22 years for the nine Nanda kings who followed them, which seems to represent a manipulation of the chronology to make things fit.

This date of 486 B.C.E. causes difficulties, because the archaeological evidence, such as it is, suggests that, at that time, some of the places which the Buddha is said in the tradition to have visited had barely, if at all, been founded, while the evidence for a monetary economy ¹⁰⁵ suggests that it is to be dated somewhere around 400 B.C.E. If this archaeological dating is to be believed, then the picture of social and political life which we get from the early Pāli texts is misleading, since it reflects the conditions of a later time, not those contemporary with the Buddha.

Fortunately, however, philology has come to our aid. Recently attention has been drawn to the fact that although the Pāli word sata and the Sanskrit word sata do undeniably mean “one hundred”, they are also used to mean a large number, any large number. It has also been pointed out that the number “eighteen” has some sort of auspicious significance. Kings are very often said to reign for eighteen years, or important events to happen in the eighteenth year of their reign. To say, then, that something happened 218 years after the death of the Buddha, probably means no more than saying that it happened after a large number plus a large number plus an auspicious number of years, and this can in no way be taken as firm evidence that the whole period of time came to 218 years.

Other ways of calculating the date of the Buddha’s death have therefore had to be employed, and recent attempts making use of the information we find in both Buddhist and Jain texts about the life spans of theras have to some extent settled on about 400 B.C.E. [^2-66] If we accept the tradition that the Buddha was 80 years old when he died, then we are talking about a life-span covering the period 480 400 B.C.E. and a teaching period of c. 445-400 B.C.E., which fits the archaeological evidence better.

For the forty-five years prior to 400 B.C.E., then, the Buddha moved around Magadha and the surrounding areas, preaching his doctrine which grew out of the general śramana tradition of opposition to the brahmanical view of the social superiority of the brahmanical caste, and to the brahmanical belief in the ātman/brahman world spirit. The Buddha’s meditative techniques and terminology owed much to other śramana movements, and his establishment of a mendicant community followed the general śramana pattern of abandoning the world, and relying for food and other necessities of life on the generosity of householders who were lay-followers. With those other śramana movements he shared technical terms and epithets to describe those who had escaped from samsāra, as well as anti-brahmanical literature.

Looked at in this way we can see that, although the Buddha’s experience and his way of making known his experience to others were unique, there is far more that Buddhism held in common with the movements which were contemporary with it than we would normally assume, and we can see that there is an element of truth in the well known statement that the only original feature of the Buddha’s teaching was his combination of a belief in the transmigration of the self with the denial that there was a self to transmigrate.

The Buddha’s message about release from samsāra and the attainment of nibbāna, as explained in his sermons, was remembered by his hearers, and repeated by them, as they had heard them. After his death, collections were made by the expedient of asking his leading followers to recite what they remembered the Buddha saying. What they said was in turn recited by the rest of the monks, and arrangements were made for the safe keeping of these texts.

III - Buddhism and Oral Tradition

In the second lecture I spoke about the way in which the Buddha’s teaching evolved from the general śramaṇa religious movement, which had grown up in opposition to the brāhmaṇa caste, as both a social and a religious entity. After seeking - unsuccessfully - release from samsāra by means of severe ascetic practices and the meditation techniques taught by contemporary teachers, he succeeded by means of his own meditative practices - unique as far as we can tell, since no one else is known to have gained release in precisely that way before him. His message that it was possible to gain release (moksa) from samsāra, and to attain nirvāna, in this unique way, was taught, by word of mouth, first to his former associates in ascetic practices, and thereafter it was spread by the Buddha and his followers throughout Magadha and the surrounding areas.

There is no agreement among scholars about the date when writing first came into use in India but everyone, I think, agrees ¹⁰⁶ that during the early period of Buddhism, even if writing was available, all teaching was by oral methods, and the Buddhist scriptures were transmitted orally, as was also the case with the brahmanical texts.

If writing was in use during the early period of Buddhism, we should have expected to find rules laid down in the Vinaya governing the proper use and storage of writing implements and materials, in the way in which we find instructions about everything else which concerns a monk’s daily life. There are no such instructions in the Vinaya, and there are in fact only two mentions of writing in the whole text, both of them in the Parivāra, the last section of the Vinaya.

The Theravādin tradition confirms the absence of writing by stating that the canonical texts were first written down during the reign of Vatṭagāminī Abhaya, in the first century B.C.E. ¹⁰⁷ in Sri Lanka, implying that before that date they had not been written down - at least in their entirety. As we shall see in the fifth lecture, there is no good reason for doubting that the two statements about writing in the Vinaya, which I have just mentioned, are interpolations, added when the Parivāra, together with the rest of the canon, was written. The absence of writing is also confirmed by the terminology employed in the texts themselves, which enables us to deduce a great deal about the manner of teaching.

The vocabulary of the early texts is centred around the words for “hearing” from the root śru “to hear”, and for “speaking” from the root vac “to speak”. So a learned man is spoken of as bahuśruta (Pāli bahussuta - “having much hearing [śruti]”), and the word for “to teach” is vāceti “to make someone say something, to recite something (after his teacher)”. Such examples can be multiplied endlessly, and collections have been made of the words and phrases which imply reciting (an orally transmitted) text, rather than reading (a written) one. ¹⁰⁸ With the aid of such collections we can make certain deductions about the transmission of texts before the writing down of the canon. It must, however, be noted that this terminology does not, in itself, prove that that the texts were not in written form, because we know that the Aśokan inscriptions, which by definition are inscribed, i.e. written, on rocks and pillars, employ the same type of terminology. Aśoka says at the end of some of his inscriptions, “This edict is to be listened to” (tyam lipī sotaviyā), suggesting that perhaps only the administrators were able to read, and their duty was to recite the contents of the edict to an audience, on the dates and in the manner specified by Aśoka.

We find, indeed, that the early terminology - the use of the verbs ṣru and vac, for example - was widely used at a much later time when writing was well known, because it was the standard terminology. That is to say: the technology changed, but the terminology did not, in the same way as the French for “pen” is “plume”, or I still find at the bottom of copies of letters which are sent to me “carbon copy to Norman”, when what I receive is not an almost illegible letter written on some sort of flimsy tissue paper, but an impressive document printed on a LaserWriter, and indistinguishable in every respect from the top copy sent to the original addressee.

We must assume that in the early days the Buddha’s followers spread the message as they had heard it from the Buddha himself, and from his chief disciples, and as they had remembered it. We have no idea of the amount of divergence which began to creep into the message as a result of poor memory, or other defects in such a method of transmission, but it would be surprising if transmission in such a haphazard way had no effect upon the words and form of the message, if not its basic content. It is possible that in the very earliest stages of the oral tradition the situation was much the same as we are told exists in the recitation of modern oral epics, where no two performances are exactly the same, and the order of episodes may be changed from recitation to recitation, and episodes may even be omitted. It does not, however, matter very much if the reciter leaves out a portion, or adds something, as long as the main narrative is maintained and the desired end is achieved. The same may well have been true of early recitations of Buddhist suttas. The name of the place where the sermon was preached, and the identity of the audience, and other such relatively unimportant details, may well have varied from recitation to recitation.

The way in which groups of synonyms were used to explain or elaborate concepts, which we find in particular in the commentarial portion of the Vinaya, in the Niddesas and in some of the Abhidhamma texts, also suggests that texts of this type were composed and then transmitted orally. In their earliest form, they must have represented the attempts of individual bhikkhus to explain something, with a resultant variation in the details of the sermon, although the main theme would doubtless have been retained unchanged. We can imagine that if someone was reciting, for example, the Dhammasangani, he might insert synonyms and quotations from other texts, as they occurred to him, and on some occasions he might remember more or less of them, but it made little difference if synonyms were omitted or their order was changed. It is possible that, as the result of such inconsistencies, the need for some sort of codification was realised, and it might even have begun to take place during the lifetime of the Buddha.

We know that the Pātimokkha, the body of rules which governed the monks’ behaviour, already existed as a collected corpus of material because the monks came together twice a month to recite it, and we may assume that each recitation was identical, unless new rules had been promulgated, or old ones modified. The fact that the Pātimokkha was structured so early doubtless accounts for the fact that the formulation of the monastic rules, although not their number, is very similar in the Pātimokkhas of the various schools. We also know that four times a month the monks preached sermons to the lay-followers, so we could speculate that such public recitations, presumably in the presence of other monks, might have led to some inconsistencies in recitation being noted and as far as possible eliminated. This would have been the beginning of codification. We can only guess at the amount which had taken place during the Buddha’s lifetime.

We do, however, know that the tradition records that after the Buddha’s death a meeting was held at which the Buddha’s teachings were recited at a joint recitation (sangiti). In the form in which the tradition tells the story, one thera recited the Vinaya rules, and another the Suttas, and they were accepted by the rest. This was the beginning of canonicity, about which I shall speak in the eighth lecture, but I must anticipate my future remarks by saying that it seems to me to be most unlikely that all these sermons were already in the shape, and in the order, in which they are said to have been recited on that occasion, and it is probable that even the use of the names of the nikāyas at the first joint recitation is an anachronism.

The tradition tells us that, after the first recitation, measures were taken to ensure that this body of material was preserved from then on. The Buddhavacana was divided up into parts and given to groups for safe keeping. Buddhaghosa tells us that the Vinaya was entrusted to Upāli and his pupils. Upāli was the expert who had replied to Mahākassapa’s questioning about the rules of the Pātimokkha, giving information about where the offence was laid down, with respect to whom, on what subject, etc. In the same way the Dīgha was entrusted to Ānanda (who had recited the Sutta-pitaka), the Majjhima to Sāriputta, the Samyutta to Mahākassapa, and the Añguttara to Anuruddha, and their respective pupils. ¹⁰⁹ The way in which the texts were said to have been shared out to these various groups implies that they were already organised into the nikāyas as we know them: long (Digha), middle length (Majjhima), those linked by associated subject matter (Samyutta) and those arranged in numerical order, smallest first, with each section increasing by one (Añguttara).

It is generally accepted that this distribution was probably the beginning of the bhānakas system. The word bhānakas means “speaker”, from the root bhan “to speak”, and is another of the items of vocabulary which suggest that the early Buddhists used an oral tradition. There are references in the Pāli commentaries to bhānakas of the first four nikāyas, and also of the Jātaka and the Dhammapada, but it is probable that there were also bhānakas of other individual Khuddaka-nikāya texts. ¹¹⁰ There seems to be only one reference in the early literature to the Khuddaka-bhānakas. Since the Jātakas are part of the Khuddaka-nikāya, the relationship between the Khuddaka-bhānakas and the Jātaka-bhānakas, who are mentioned in the same sentence by the author of the Milinda-pañha, ¹¹¹ is not clear.

If the material entrusted to the groups of bhānakas was at first of a somewhat haphazard nature, as I have suggested, then the first task would have been to start some sort of editorial process, to make the material more consistent and to devise ways which would ensure that the consistent whole which the bhānakas produced could easily be handed on to their successors. We can surmise that the language was homogenised to a large extent. Once we have determined the nature of the language which we call Pāli we can see that in the canon as a whole there are very few non-Pali characteristics and most of those are due to a consistent introduction, at a later date, of Sanskritisms, which are restricted in number, the most obvious being the absolutive in -tvā. By eliminating anomalous dialect forms, such an editorial process may be presumed to have made learning the texts easier, by simplifying forms which at the beginning were probably quite divergent.

Once the language was homogenised, then the lists of synonyms, etc., which I have just mentioned, would probably have been fixed both in number and in order, and we can see signs of this in some of the categories of the Dhammasangani, where we commonly find a trio of words together, in a list of synonyms - a short a - or long ā-stem noun, then an action noun made from the same root with the -ana suffix, and then an abstract noun made by adding -tta or -tā to the past participle of the verb - always in the same order.

It is clear from a comparison of the way in which the nikāyas are formulated that there are certain differences between them, presumably arising from the fact that they were remembered in a slightly different way by the bhānakas responsible. We would expect the editorial process to vary from bhānakas group to bhānakas group, so that it is not surprising to find that formulation also varied from one group to another.

Nevertheless, it is interesting to note that, by and large, the differences in the language and other features of the various Pāli nikāyas, to which I have just alluded, are not great. Considering the disparate nature of the material with which the bhānakas must have started, this is rather surprising, because, although we know little about the relationship between the various groups of bhānakas in the Pāli tradition, we can deduce that they did not consult with each other to the extent of making their recitations of individual suttas or groups of verses identical, as we can see, for example, in the case of the thera Vangīsa’s verses, for a comparison of the versions of the verses ascribed to him in the Theragāthā, the Samyutta-nikāya and the Suttanipāta, shows that the versions preserved by the bhānakas of the Samyutta- and Khuddaka-nikāyas do not agree in every way, although it is possible that Vangīsa repeated his verses in different ways on different occasions.

The bhānakas also had different ideas about matters of Buddhist history, e.g. whether the four nimittas were seen on the same day or not, and why Ānanda arrived late at the recitation, ¹¹² and also about the distribution of texts between the various pitakas. Nevertheless, despite these differences, the fact that in general duplicated texts do not differ so very much, when they occur in different nikāyas, would indicate that there was, or had been, some sort of co-operation between the bhānakas or their predecessors. In the case of the bhānakas belonging to different traditions, or their equivalents in other system of transmissions, we can assume that the co-operation between them was less, although recent investigation has shown that communication between some schools, at least in later times, was closer than has sometimes been thought. ¹¹³

If, however, the bhānakas had started their conservation work immediately after the first recitation, without consulting one another, it is not easy to see how the homogenisation of the language of the texts, and in particular the translation from one dialect to another, which must have happened on more than one occasion in the history of the Theravādin canon, as the Buddha’s message spread - as we shall see in the fourth lecture - could have happened in such a consistent way between the various nikāyas, with no one nikāya showing a greater proportion of anomalous sound changes than another. This would suggest that a certain amount of homogenisation had already taken place before the bhānakas took over, and the delay entailed in this standardisation would also, of course, have allowed time for the texts to be collected, classified and codified into nikāyas.

The fact that differences of dialect are detectable in the Pāli canon shows that the form of the texts was certainly not fixed unalterably immediately after the first recitation. If changes could be made, this too would suggest that the bhānakas system was not yet in operation, or at least not in the form of caretakers of an immutable body of material. The possibility that change could still take place would suggest that similar changes could occur when comparable texts were being remembered by the monks belonging to other traditions. This would perhaps account for the differences which we find in related suttas belonging to other schools.

We must accept, then, that the story that the bhānakas system was instituted at the first joint recitation creates great difficulties. Although it seems clear that the origin of the bhānakas system must have been on the lines that have been suggested, it clearly cannot have happened in exactly that way. It is most unlikely, to say the least, that within a very short time of the Buddha’s death suttas had already been collected and categorised by length and subject matter into the form in which we have them in the Theravādin canon, and it is most unlikely that the Sutta-pitaka was in its present form at that time. It is obvious that if the Dīgha-nikāya, Majjhima-nikāya, etc., had not yet been formulated and named, there could hardly have been Dīgha-bhānakas and Majjhima-bhānakas, etc.

The Khuddaka-nikāya presents even greater problems, e.g. there are references in the Apadāna to the Kathāvatthu. The Kathāvatthu is, however, acknowledged by the Theravādin tradition to be a very late text, composed on the occasion of the third recitation in Asoka’s time. It is obvious, then, that portions, at least, of the Apadāna must be very late additions to the Khuddaka-nikāya, and yet the Apadāna is included in the lists of texts recited at the first recitation. As I have suggested, it is probable that even the use of the names of the nikāyas at the first recitation is an anachronism.

It is therefore possible that the information about the texts which were recited at the first recitation was given with the benefit of hindsight, by those who knew what form the canon had in their day, and who thought or at least maintained, perhaps for the purpose of authenticating their own school, that that form was precisely what had been recited at the first sangiti. Similarly, the bhānaka tradition was perhaps a later invention in the form in which we hear about it, being restricted to the Theravādin tradition, and the story of its early foundation was invented to give authenticity to the Theravādin canon. As to when this was done, we cannot say much, except to draw attention to the occurrence of the names of the nikāyas and the word banaka (of the majhima, eka-uttiraka and śayutaka nikāyas ¹¹⁴ ) in early Sinhalese inscriptions probably datable to the second century B.C.E. The institution of the bhānaka system might then be the result of a decision taken after the second recitation, or even as late as the third recitation, i.e. the bhānaka system arose before the formation of the Abhidhamma-pitaka.

On the other hand, we should perhaps note that there is no reference to the existence of Abhidhamma-bhānakas in the sentence in the Milinda-pañha which I mentioned earlier. We find ābhidhammikas included there, presumably experts in the Abhidhamma, but there is no mention of bhānakas. This may mean that the bhānakas system was closed, with no new groups being set up by the time the Abhidhamma was formulated, so that the Abhidhamma as a whole was composed too late to be incorporated in the bhānakas system. The Milinda-pañha sentence might indicate that there was a difference between an ābhidhammika and an Abhidharma-bhānaka. It is probable that the bhānakas were something more than mere caretakers of the texts entrusted to them. It has been suggested that they were also professional reciters, and this seems to be the sense of the word in Buddhist Sanskrit, where it occurs most commonly in the compound dharma-bhānaka “a preacher of the dharma”. ¹¹⁵ They could perhaps be called upon to deliver a sermon when required, and someone asking for a sermon to be recited could specify the type of sermon by length. As I mentioned in the first lecture, we sometimes find two version of a sutta, one long and one short, and so it would be possible for someone to say “I would rather like a middle length sermon, about something or other, and by the way I would prefer the shorter rather than the longer version of that sermon”.

The differences of view between the bhānakas about historical matters, which I have mentioned, perhaps indicates that a bhānakas did not merely recite, but could also put his recitation into some sort of context. The same distinction between a caretaker-cum-historian and a reciter is perhaps to be made about the word tepitaka “knowing the three pitakas”, which is used to describe, for example, the monks at the third sangiti. There was presumably a distinction between knowing the tipitaka and being a member of a group whose purpose in life was to teach a text to others and to be able to recite it to order. If the purpose of the bhānakas was to recite sermons to lay-followers at their request, then we may assume that few lay-followers wanted a recitation of a portion of an abhidhamma text. The teaching of abhidhamma was perhaps still possible by means of the elaboration of the mātikās - which I shall mention in a moment on an ad hoc basis.

We can only speculate about the way in which the bhānakas system operated. We might assume that the junior bhānakas sat around their seniors and learned the texts from them. Since the oral tradition is still strong in the Buddhist countries of South and South-east Asia it might be thought to be a simple matter to visit a monastery, and see just how the oral tradition is preserved, and we might have hoped that descriptions of the way in which texts are remembered and recited in modern times would have thrown light upon the situation at earlier times. This unfortunately does not appear to be the case. There is, for example, an account, by Tambiah, of the way in which texts were learned by junior monks when they were first admitted to the circle of reciters. He reported ¹¹⁶ that, in the monastery which he visited, texts were chanted by monks morning and evening. Newcomers repeated what they heard and memorised the chants fairly quickly, e.g. chants used in the daily worship of the Buddha - parittas such as Mangala, and other texts such as the Pātimokkha (which as we have seen was one of the first Buddhist texts to be recited). ¹¹⁷ This would seem to be a good guide as to the way in which memorisation was done in early times. Unfortunately, the picture is somewhat ruined by Tambiah’s discovery ¹¹⁸ that when the monks practised chants individually, they did so with the aid of printed texts, to ensure that they got them right.

We do not know how long the bhānakas system remained in being. Buddhaghosa refers to it as though it was still in operation in his time, although by then the canon had been written down for some hundreds of years. This is, however, not conclusive, because when Buddhaghosa says that the Dighabhānakas and the Majjhima-bhānakas do such and such he may simply be repeating what he found in the commentarial literature available to him in the Mahāvihāra in Sri Lanka, and it seems very clear that this commentarial literature was composed some centuries before the time when Buddhaghosa wrote and presumably referred to conditions at that time, which perhaps no longer pertained by Buddhaghosa’s time.

I said at the beginning of this lecture that one of the first acts of the bhānakas, or their predecessors if they were a late institution, must have been to attempt to homogenise the language. Side by side with that imposed consistency of language we can see that there is a certain consistency of structure.

There has been in the past a certain reluctance to believe that all early Buddhist teaching could be based entirely upon an oral/aural tradition, because of the sheer quantity of the material which would have to be remembered. In this connection we must have regard for the tradition, which I have mentioned, that the texts were divided up into nikāyas and shared out to different theras and their pupils for safe keeping and onward transmission, so that no thera was responsible, as a bhānakas, for more than one nikāya.

When oral literature became the object of academic study, and research was carried out into the recitation of oral epics in various parts of the world, it was discovered that so-called “primitive” peoples were able to remember and recite very long texts. Doubts about the ability of the early Buddhists to memorise their material were thought, by some, to have been dispelled by reference to such feats of memory, since it was thought that similar recitation might explain how the oral tradition of Buddhism and other religions could be maintained.

It became clear after a while that the situation was not as simple as might appear. The oral literature which was being studied was essentially of a verse nature and therefore the comparison would seem to apply only to Buddhist verse texts, but even there the situation was not entirely comparable, because the very nature of Buddhist verse texts, and the metres in which they were written, demanded complete accuracy of memorising, whereas the oral literature which has been studied is essentially of an epic nature where, as I said earlier, it is alleged that no two performances are ever identical because the reciter is free to insert, at any point, material of a formal nature, the so-called formulae which can be used to keep the recitation going while he remembers what happened next in the story. The great majority of Pāli canonical texts, however, are in prose, and complete accuracy of reproduction is required at each recitation. In these circumstances the findings of modern investigators of oral epic literature seem to have little relevance.

It seems that we must seek elsewhere for the explanation of the way in which the Pāli material could be remembered and handed on. We know that elaborate systems of recitation were and to a limited extent still are employed in Vedic recitation, but there is no evidence known to me for the employment of such methods by the Buddhists. Nor would they be entirely appropriate for prose texts.

I have already mentioned a consistency of structure as being one of the features which we can detect in the Pāli canon, and if we examine the canonical texts we can, in fact, find many features which we may assume were employed for mnemonic purposes to make the memorisation, and therefore the recitation, of these texts easier.

Rhys Davids a century ago ¹¹⁹ drew attention to some of the features which, he suggested, aided the power of memory for Buddhist Sutta and Vinaya texts. He pointed out “firstly, the use of stock phrases, of which the commencement once given, the remainder followed as a matter of course and secondly, the habit of repeating whole sentences or even paragraphs, which in our modern books would be understood or inferred, instead of being expressed”.

It is not precisely clear what he meant by stock phrases, and I suspect that he was referring to the way in which suttas tend to start in the same way - “Thus have I heard” - this is said by the cties to be a reference to the way in which Ānanda at the first recitation repeated what he could remember of the Buddha’s sermons ¹²⁰ - “At that time the Buddha was staying at such and such a place with a group of bhikkhus, and one day the bhikkhus decided to do something, or ask a question, etc.” Many of the introductory paragraphs to these sermons carry on with stereotyped phrases - someone approached the Buddha, and having approached him sat down at one side; to him seated at one side the Buddha said something or other.

In a paper read at the conference of the International Association of Sanskritic Studies in Australia in 1994, ¹²¹ it was shown that the consistency in the way in which these introductory paragraphs are structured is, in fact, more meticulous than might at first appear. It can be seen that the wording changes subtly in conformity with a fixed pattern to specify who is approaching whom. In each case the wording is slightly different and a further result of this is that, once a story teller has remembered that the particular sermon he is about to recite is, say, about one or more bhikkhus approaching the Buddha, to ask a particular type of question, the form of wording to be used is prescribed. And therefore he does not, so to speak, have to remember the form of words to use because the circumstances fix the form for him.

Interestingly enough, it has been pointed out that if we examine such stereotyped phrases in one nikāya and compare them with the phrases in another, we find that the forms which are employed do not necessarily agree something which leads us to the conclusion, I think, that, as we would expect, once the texts had been distributed among groups to preserve and hand them on to their successors, the precise methods of stereotyping which were employed, in an attempt to make remembering easier, were not necessarily the same for each set of bhānakas.

Another way in which we can see that these these texts were recited rather than read is the way in which they include in them lists of contents or indexes the so-called mātikās, ¹²² sometimes translated as matrixes. It has been pointed out that these lists play a mnemonic role, ¹²³ which, as we have seen, is important in a tradition which is founded essentially upon oral transmission. The word mātikā is sometimes used to designate the precepts of the Vinaya, ¹²⁴ and mātikās form, as it were, compendia of the doctrine. They are, however, also lists, especially in the Abhidhamma, which are presented in a numerically progressive form and they have a creative role, in as much as they allow the Abhidhamma to be composed and thus, it has been suggested, the idea of “mother” is contained in both mātikā and matrix. It has been said with regard to the Paṭthāna, for example, that if one knows the Mātikā to that text with its 24 conditional relations, and relates them to the 22 triplets and 100 couplets of the mātikā to the Dhammasangani, in all the permutations and combinations in which these can be taken, one can reconstruct the whole of the Paṭthāna. ¹²⁵

Something similar can be seen in the uddānas (the lists of contents) which are found in, or at the end of, a number of texts, e.g. in the Dhammapada where the uddāna gives the names of the vaggas, and in the Thera- and Therī-gāthās, i.e. the verses ascribed to male and female elders, where we are told how many elders there are and how many verses as a whole they have recited. Despite their position in our texts, these uddānas were probably intended originally for use at the beginning of the recitation, and even as the reciter progressed. Someone reciting the Dhammapada, for example, had a guide to tell him which vagga came after which, so that he could keep them in order. It must be pointed out that the system is not foolproof, because the numbers given in the uddānas to the Thera- and Therī-gāthās do not agree entirely with the numbers as we have them now, ¹²⁶ and they presumably refer to an earlier recension of the text, where such numbers were relevant and of value to the reciters. They have been retained in a written recension, even though they are no longer of any value.

Listing things by numbers is a very common mnemonic device in India, and both the Buddhists and Jains make use of this idea. The Jains have the Sthānānga and the Samavāya, the contents of which are listed numerically, and the Buddhists have something very similar in the Sangiti-suttanta of the Dighanikāya. The name Sangiti is reminiscent of the sangitis at which the texts were recited in a joint recitation, and the name of the sutta suggests that it represents a recitation of doctrinal matters, arranged in a numerical way, and intended for chanting together, perhaps in an attempt to provide a summary of the doctrine as a precaution against the confusion among the Jains after the death of their leader Nāthaputta, which was the occasion for the preaching of the Sangitisutta. Whether the recitation was at one of the great sangitis or at some other chanting is a matter for investigation.

We also find this “more by one” principle used in the Anguttara-nikāya, which is listed numerically for the same reason. It is quite clear that this method of enumeration is purely for artificial purposes, because we find that although the lists of ones and twos and threes and fours, etc., are genuine groups, when we get to the higher numbers the authors of the text were obliged to make combinations, e.g. one group of ten is made of the five fears (bhaya), the four elements of stream-entry (sotāpattiyanga), and the ariyan method (ariya ñāya). [^3-22]

Another guide for oral recitation which we can find in the Pali texts is the principle which appears in the list of abbreviations in A Critical Pāli Dictionary as wax. comp., i.e. the rule of waxing components. It is a translation of a German term, which was first used, as far as I know, in the field of Assyrio-Babylonian studies. [^3-23] It perhaps sounds odd in contemporary English, where the word “waxing” in the sense of “growing larger” is almost entirely restricted to the moon growing larger, as opposed to its waning phase. If we are to retain the word “waxing” then it is perhaps clearer if we call it the Waxing Syllables Principle, because the phrase refers to the fact that strings of words which make up a group, e.g. a number of Pāli epithets describing a city as “large, rich, prosperous, flourishing, crowded”, are very often put into order depending upon the number of syllables in each word, with the words with the fewest syllables coming first, and then the word with the next fewest syllables next, and so on until the word with the most syllables comes last.

Such a principle clearly guards against the way in which the order of such a string of adjectives might be shuffled around. They were put into a fixed order depending on this Waxing Syllables Principle and therefore they were always remembered in that fixed order. If anyone recited the string of words and put one word in the wrong place, then the change in the number of syllables would immediately reveal it. Once this principle became second nature, so to speak, then a reciter could not go wrong, because he would automatically recite words in the right order.

Something similar in English, but based on an entirely different principle, is the way in which we might talk about, say, a big red armchair. For some reason, of great interest to linguists, we never say a red big armchair, and so if we are telling a story where there is frequent reference to a big red armchair, to some extent our task of remembering that part of the story is eased, because we do not, each time, have to remember the order in which to say the words. The order is fixed for us in a pattern by something we do not know about and do not have to worry about. It is, so to speak, automatic.

The interesting thing about this is the way in which we occasionally find that words are not in the Waxing Syllables Principle order. If we find a set of words which we might suspect was at one time arranged on that principle - and if the suspicion is supported by the fact that in another tradition or in a comparable phrase in a Jain text it does occur in that order - but it no longer is, we may well be able to find a reason for the change. Perhaps one term has been replaced by a synonym, with a different syllable length, or a word in its Pāli form with a svarabhakti vowel is out of place, but if we calculate what the Sanskrit form or the form in another dialect was, we can see that the principle is indeed retained intact. This gives us information about the dialect in which the phrase was first composed. Sometimes we may suspect that the replacement took place at a time when the Waxing Syllables Principle was no longer operative, i.e. when the oral tradition which required memorisation had given way to a written tradition, which did not require it, and the Waxing Syllables Principle was of less importance.

Another principle to which Rhys Davids drew attention a hundred years ago is the way in which we find in Buddhist texts frequent examples of repetition.

The way in which repetition operates varies considerably. Sometimes it is exact repetition. If something happens twice or more, or something is said twice or more, then the exact passage is repeated verbatim on each occasion, e.g. stock sets of words, such as the description of a city, which I have just mentioned. Sometimes the repetition is partial, for example, we may find a statement saying that the good man does a series of actions, followed by a statement that the bad man does the opposite of these, each word being the same words used for the good man, with a negative prefix a - added to each. The reciter had therefore in effect only to remember one set of adjectives, and to put the negative in front of them, e.g. the good man does kusala deeds, while the bad man does akusala deeds. Once again the effort involved in memorising the sermon which contains these phrases is reduced. Similarly, there might be a passage followed by its opposite with the negative particle na in front of each verb. The good man does this, does that, does something else. The bad man does not do this, does not do that, does not do something else. Or there might be a set of phrases in each of which one word is changed: “We go to the Buddha for refuge, we go to the Dhamma for refuge, we go to the Sangha for refuge”.

Although it is unlikely that the circumstances were completely identical, the stories in the Majjhima-nikāya about the Buddha-to-be visiting the two teachers, which I mentioned last week, starts off with identical phrases, with changes made only to cover the different level of meditation reached, and the different response of the teacher when his pupil has successfully imbibed the teaching. Sometimes translators get carried away when they find such repetition. It is not, in this particular example, as exact as one would believe from reading Miss Horner’s translation, in which she makes the stories even more parallel, even more repetitive, than the Pāli justifies. [^3-24]

Repetition reaches its highest (or lowest, depending on how you look at it) level in Pāli, in my experience, in the Alagaddūpamasutta of the Majjhimanikāya. [^3-25] In that sutta a bhikkhu called Arittha developed the erroneous view: “In so far as I understand the dhamma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is no stumbling block at all”. [^3-26] The text tells us that other monks heard that Ariṭṭha had developed the erroneous view that “In so far as I understand the dhamma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is no stumbling block at all”, and so they went to him and said; Is it true Ariṭṭha that you have developed the erroneous view that “In so far as I understand the dhamma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is no stumbling block at all”, and Ariṭṭha replied; Yes, I have developed the view that “In so far as I understand the dhamma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is no stumbling block at all”. The text goes on to say that despite all their efforts, Ariṭṭha maintained his view that “In so far as I understand the dhamma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is no stumbling block at all”. So the bhikkhus went to the Buddha, and told him that Ariṭṭha had developed the view that “In so far as I understand the dhainma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is no stumbling block at all”, and that they had heard that he had developed the view that “In so far as I understand the dhamma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is no stumbling block at and so they had gone to him and asked him if it was true that he had developed the view that “In so far as I understand the dhamma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is no stumbling block at all”; and he had replied that he had developed the view that “In so far as I understand the dhamma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is no stumbling block at all”, and yet despite their efforts, he had persisted in his view that “In so far as I understand the dhamma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is no stumbling block at all”. And so the Buddha sent for Ariṭṭha, and asked him if he had developed the view that “In so far as I understand the dhamma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is “no stumbling block at all”, and Aritf11a replied that he did have the view that “In so far as I understand the dhamma taught by the Bhagavat, it is that, in following those things called stumbling blocks by the Bhagavat, there is no stumbling block at all”. Whereupon the Buddha condemned it as a wrong view.

That is to say that, if my arithmetic is correct, the identical passage occurs 12 times, and it was partly for that reason that I used to include that sutta very early in my Pāli course - the doctrinal importance of its contents was another reason. For beginners in Pāli the sheer size of the vocabulary to be mastered is a deterrent, since every line of every sentence presents new words which have to be understood and committed to memory. The amount of repetition in the Alagaddūpamasutta means that students suddenly find that they already know the words, and they begin to think that they are making progress as the same sentence describing the erroneous view occurs again and again, until they are able to chant out the heretical statement as well as any bhāṇaka, proving the value of repetition when memorising.

To us, repetition carried to such an extreme length is ludicrous, and we should certainly try to avoid it, introducing such phrases as “This view”, “such a view”, “the view which you mention”, etc., yet quite clearly such repetition was not regarded as otiose at the time of the oral tradition, although as we shall see in the fifth lecture when we consider Buddhism and writing, the situation changed somewhat at a later time.

Stock phrases, lists, the waxing syllable principle, repetitions, are all things which make the memorising and recitation of prose texts easier. In addition to these there are the usual literary features of alliteration, etc., which help to determine the choice of words to use, and to assist in remembering them.

But despite the help which such aids gave, and despite the care which the bhānakas may be assumed to have taken in reciting the texts allotted to them, it is inevitable that mistakes would creep into the tradition by reason of the type of error which is inherent in oral recitation, particularly when monks from different parts of India, with different pronunciations of Pāli, were assembled together. Buddhaghosa gives a list of errors which would invalidate a kammavācā - an official action of the sangha: uttering an aspirated consonant when it should be unaspirated, and vice versa, de-voicing a voiced consonant and vice versa, etc., all of which would be of general application in oral recitation.

In addition to such errors, the dangers inherent in an oral tradition are obvious. Handing texts on orally depends on two essentials: a donor to hand them on, and a recipient to accept them. If interest in a religion wanes, and no one wants to hear a text, it dies out, or if the number of donors of a text are reduced in number to a few old men, with failing memories, then the text is likely to be handed on partially or incorrectly, or not at all.

At some time after the introduction of Theravādin Buddhism into Ceylon, war and famine and the destruction of vihāras led to a breakdown in the bhānakas system, and to a situation where some texts were known to a very few bhikkhus. Buddhaghosa records the fact ¹²⁷ that there came a time when only one bhikkhu knew the Niddesa, and urgent measures had to be taken to have him repeat his text to receivers before he died. From fear of the Niddesa disappearing completely, the thera Mahārakkhita was persuaded to learn it from this one bhikkhu, and other theras learnt it from Mahārakkhita, so that the future transmission of the text was assured. ¹²⁸

Such incidents no doubt had an effect and gave a warning to the sangha, and made the theras in Ceylon realise that the whole canon could disappear if the oral tradition died out. This was probably one of the factors which persuaded the bhānakas that it was time they made use of the new-fangled writing. The interaction between the oral tradition and the written tradition, and the effect which that had upon Buddhism are the subjects of my fifth lecture.

Before then I want to talk about the way in which the Buddha’s message spread from Magadha, where he had first delivered it, throughout the North of India, down south to Sri Lanka, and North through Chinese Turkestan to China and elsewhere. And it is the first part of that movement, away from Magadha, which will be the subject of the fourth lecture. In that lecture I will deal with the philological information which we have about the form of the language which was used in the very early stages of Buddhism, and the way in which we can interpret philological material to give us some idea of the way in which Buddhism began to spread from the boundaries of its origin.

IV - Buddhism and Regional Dialects

In the third lecture I dealt with the way in which the early Buddhist texts were transmitted by an oral tradition, and I mentioned the difficulty I had in believing that the bhānaká tradition could have come into effect within a very short time of the Buddha’s death, because this did not seem to leave time for the homogenisation of language, which would seem to have been necessary if the Buddha preached, as is generally accepted, in a number of dialects. In this lecture I want to talk about the philological information which we have about the form of the language which was used in the very early stages of Buddhism, and the way in which we can interpret philological material to give us some idea of the way in which Buddhism began to spread from the boundaries of its origin.

We very commonly find in books and articles about early Buddhism such statements as: “The Buddha preached in the Prakrits, the language of the common people, and resisted the suggestions of some of his ex-brahman followers to translate his sermons into Sanskrit”. There is frequently no hint that these statements are anything other than accepted fact, but readers need to be very wary, because such statements are frequently not fact, and are anything but accepted by all scholars working in the field.

To tell the truth, there is a great deal of the Bellman principle in the academic world. You all know about the Bellman in Lewis Carroll’s The Hunting of the Snark, who maintained that “what I tell you three times is true”. I am as guilty in this respect as anyone else, I fear. I may have an idea about something, and so I incorporate my idea, as a suggestion, in an article I am writing, and wait for someone to reject or disprove it. No one does, and I repeat the idea, still as a suggestion, in another article. Again, no one rejects it. I do this a third time, and if there is still no reaction, it becomes fact in my mind - I have said it three times, so it must be true, and I consequently refer to it in future publications as an established fact. The thought that no one ever reads my articles and so no one has ever seen my suggestion, and so no one has had any desire to reject it, or the alternative explanation, that those who read my first article thought that the idea was so preposterous that it was not worth wasting paper and ink refuting it, so that the second and third repetitions were dismissed as “I see Norman is still pushing that stupid idea of his”, does not enter my head.

This reference to the Buddha refusing to allow his sermons to be translated is based upon a passage in the Vinaya, ¹²⁹ but to translate the passage in that way is to ignore the very real difficulties which are presented by two words in the original Pāli, chandaso and nirutti. The first has been translated “into Vedic”, or “into Sanskrit”, or “metrically”, or “as desired”. I cannot believe that “into Vedic” is a possible translation, but if “into Sanskrit” is a possibility, then this means, of course, that the Buddha’s words were not already in that language. In support of this translation, it has been pointed out that the bhikkhus who made the suggestion were brahmans by birth, who might be thought to favour Sanskrit in preference to the local vernaculars.

On the other hand, we might have thought that such learned people would have the sense to realise that the Buddha preached in Prakrit because that was what the local people spoke and understood, and it would be an act of folly to turn his sermons into something which would be unintelligible to them. There is the additional point that, if chandaso does mean “into Sanskrit” this would appear to have been unknown to later translators who did turn the Buddhavacana into that language. Buddhaghosa seems to have understood the bhikkhus’ suggestions as meaning: “Let us translate, as we translate the Veda into Sanskrit”, presumably referring to a situation where brahmans would be able to explain difficult words in the Vedas, when they recited them, in the same way as we know that Sanskrit commentaries were written upon the Vedas. This, then, might perhaps mean that the ex-brahmans were asking permission to gloss, i.e. explain, the Buddha’s words as they recited them.

If, however, we accept either of the translations “metrically”, or “as desired” for chandaso, then this tells us nothing about the language which the Buddha used. In view of this uncertainty, we should perhaps not put too much weight upon this problematic passage, and we should instead try to work the matter out for ourselves in another way.

Is there any evidence that there were, in fact, dialects in use at the time of the Buddha? We have Buddhist texts in both Sanskrit and dialects of MIA. How do we know that the Buddha did not preach in Sanskrit, and the translations from Sanskrit into MIA were not made later?

We must assume that the language(s) of early Buddhism reflected the linguistic background of the times. The beginnings of Prakritic tendencies in Indo-Aryan can be placed very early. In his examination of the history of the Proto-Indoaryans, Burrow pointed out ¹³⁰ that in the traces of their language found in the Hittite archives, the word for “seven” is satta (cf. Skt sapta), and he referred to this “as having evolved beyond the Proto-Indoaryan stage”. ¹³¹ We find in the Rgveda a number of characteristics which we would regard as Prakritic, i.e. they do not present the form which our knowledge of Indo-European philology tells us we should expect in early Sanskrit. Many lists of such forms have been published. ¹³² They include the semi-vowel r appearing as u in muhur and in pitus; r disappearing and making a following dental -t - retroflex in vikata, beside the expected form vikrta. We can see that in some places the metre demands the insertion of a svarabhakti vowel, e.g. Ind ¹³⁰ ra. Even though we must make it clear that we are talking about the recension of the Rgveda which we possess now, which is perhaps no earlier than the seventh century B.C.E., ¹³³ nevertheless we can see that such features were already in evidence well before the time of the Buddha. If there were Prakritic features in the literary language of the Rgveda, the assumption is that the spoken languages of the ordinary people had many more of these features.

Our view about the prevalence of MIA features at this time is reinforced by the fact that about 400 B.C.E., i.e. about the time of the Buddha’s death, the Sanskrit grammarian Pānini produced his grammar, the prime aim of which was, we may assume, to establish the form of Sanskrit inflexibly, so that MIA could not affect it - although, as will be clear from what I have just said, to some extent he was closing the stable door a little late. Patañjali’s Mahābhāsya, dating from the second century B.C.E., actually quotes some of the forms which must be avoided, and they include forms which are well attested in MIA. ¹³⁴

We need then have no doubt that the MIA dialects had been formed, and were in use during the time when the Buddha was teaching, but we have little knowledge about the dialect geography of the fifth century B.C.E. If we ask: “What was the language of the Buddha?”, then we have to admit that we do not know for certain what language or languages the Buddha spoke.

Any movement, whether religious or social, which opposed the power and status of the brahmanical caste might be expected to make use of a nonbrahmanical language. Not only would this be an anti-brahmanical gesture, but it would also aid the non-Sankrit-speaking element of the population, i.e. those who spoke a vernacular language. It therefore seems very likely that the Buddha’s sermons were preached in a non-Sanskritic language, i.e. a Prakrit, and from the fact that he moved about preaching in various places we can assume that he preached in a number of dialects, varying his language to suit his audience. Much of his teaching life was spent in Magadha (although none of the great events of his life occurred in that area). We can therefore assume that on some occasions, at least, he used the Magadhan dialect of the time, which we may call Old Māgadhī.

This belief might appear to be supported by the fact that the Pāli commentarial tradition tells us that the Buddha’s language was Māgadhī. Unfortunately the commentators use this name about the form of the canon which they had before them, i.e. the language which we call Pāli, and as we can be quite certain that Pāli is not Māgadhī, we have to examine carefully just what they meant when they said this. It is probable that they believed that the Buddha spoke the words as they were in the canon, i.e. in Pāli, and as they knew he lived and uttered the words in Magadha, they believed that Pāli was Māgadhī.

We do not know precisely the form which Māgadhī had at the time of the Buddha. We can deduce to some extent what form that dialect had at the time of Aśoka, and we know what the grammarians writing some centuries later said were the characteristics of the dialect. Extrapolating from this, we can gain what we may regard as a fairly accurate idea of the main features of that dialect at the time when the Buddha was teaching. Its main characteristics would have been: l for r, palatal s for all sibilants, -e as the nominative singular ending. If this is correct, then, as I have said, the language of the Theravādin canon which we have is certainly not Māgadhī. That would indicate that the language has been translated on at least one occasion, presumably to meet the needs of the situation. As Buddhism moved from the land of its origin into areas where different dialects or languages were spoken, and as those dialects or languages developed and changed over the course of time, it would seem to be inevitable that some sort of translation process was needed if the Buddha’s teaching was not to become unintelligible to those to whom it was preached. Are there any traces of this? Yes, there are. If we analyse the language of the Theravādin canon, i.e. the language which we call Pāli, we can see that for the most part it has features which we would class as western, using this in a linguistic rather than a geographical sense, since we find eastern forms in the version of the Aśokan inscriptions in the West at Sopārā. Nevertheless, we can detect anomalous forms in it, i.e. forms which do not seem to follow the western patterns of phonology and morphology of the language. Some of these have features which are more appropriate to Sanskrit, e.g. consonant groups containing -r-, and the absolutives in -tvā. Disregarding these for a moment, we can see that there still remain a number of other anomalous forms. There are forms with eastern kkh where we should expect western cch, e.g. bhikkhu and bhikkhunī; forms with l where we should expect r, e.g. verbs with the prefix pali- instead of pari-; forms with -e where we should expect -o, e.g. bhikkhave instead of bhikkhavo, and also some nominative singular endings in -e, instead of -o; examples of the voicing of consonants, e.g. yādeti; forms with intervocalic y instead of k or t; forms with j where we should expect y, e.g. jantāghara; dental n instead of retroflex n, e.g. in nibbāna; v where we should expect y, e.g. āvuso from the noun āyu(s); and a small group of words where a consonant group including a nasal has developed in an unexpected way, e.g. n t > n d, n d > n n and m b > m m.¹³⁵

It is generally agreed that these are dialect forms, from one or more dialects, and they may be regarded as being remnants of dialects through which the Buddhavacana was transmitted before it was translated into the language which we find in the Theravādin canon, i.e. Pāli.

In what I say in this lecture I am, to some extent, a prisoner of my own nomenclature. I shall be using “Pāli” as the name of a language, and it has become conventional to do so. But Pāli is an abbreviation of pāli-bhāsā, which means the language of the pāli, i.e. the texts, or the canon. It follows, then, that every feature of the language of those texts, however strange, and however inconsistent with the rest of the language of the texts, is Pāli. We can define it, if we wish, as a western dialect with some eastern features, although this in itself is a matter for debate, as we shall see, but it would, technically, be incorrect to talk about these possible eastern features as anomalous forms. It is even more incorrect to call them non-Pāli forms. Nevertheless, I must call them something in the course of my discussion about what to call them, so I shall in general call them “anomalous” or sometimes “eastern”, and when I want to stress that the Pāli being used is free from such eastern forms, I shall call it western. This may seem confusing, but I hope that all will become clear as I go along.

The presence of these anomalous forms in the Theravādin canon was, of course, noticed a long time ago, and their value as indicators of an earlier form of the Buddhavacana has been much studied and debated. We have to ask ourselves what their significance is. Why they were retained from an earlier version, and what conclusions can we draw from their retention? Some scholars have regarded them as evidence of an Ur-kanon, (“a primitive canon”), ¹³⁶ while others have seen them as pré-canonique (“pre-canonical”). ¹³⁷ The first of these views pre-supposes that there was a canon in existence at some pre-Pāli time, and while this is possible and indeed quite likely - although this depends to some extent upon the definition which is given to the word canon (something to which I shall return in the eighth lecture) - these forms cannot be taken as proving this. All they prove is that some texts, i.e. the ones in which we find the anomalous forms, existed at an earlier date in a dialect or dialects other than Pāli. Even this is overstating the case, because it is quite clear that the use of some of these forms was extended by analogy. I shall return to this in a moment. At this point I will only say that there is probably no one reason for their retention.

If these anomalous forms are remnants of dialects through which the Buddhavacana was transmitted, then our task is to identify the dialects and the areas in which they were used, in the belief that this will give information about the regions of North India through which Buddhism spread. The problem is to see whether our anomalous forms exhibit any of the characteristic features of any of the dialects about which we have knowledge, in the hope that this will tell us about the regions through which the texts, or parts of them, were transmitted. To guide us in our search we have two main aids: the inscriptions found in India, from the time of Aśoka onwards, and the statements of the grammarians. Both sources are to some extent unreliable, and in any case refer to a time later than the Buddha. I mentioned in the second lecture the belief of some that the Buddha’s death is to be dated c. 400 B.C.E. This means that the Aśokan inscriptions are about 150 years after that date, and so we have to allow that much time for linguistic change, and we have to accept that the Aśokan inscriptions cannot be entirely satisfactory as a guide to earlier dialects and dialect forms. Furthermore, although we can surmise that Aśoka intended his inscriptions to be carved in the dialects appropriate to each site, there is clear evidence that this was not always done.

The grammarians were writing many centuries later, from the fifth century C.E. onwards, when the dialects had, in any case, been standardised by being used for literary purposes, especially in the dramas, although a literary dialect or language very often represents a fossilised version of a language used as a vernacular long before, and so may retain very old features. As a third aid, there is also the information which can be gained from Jain texts, which underwent the same type of language changes as the Buddhist texts, i.e. different dialects were used as Jainism spread. As in the Buddhist texts, there is the same difficulty of identifying the languages and the areas where they were used. Nevertheless, we can, to some extent, use Jain texts to help with the identification of the dialects of Buddhist texts, and vice versa.

We know nothing about the translation techniques which were employed by those taking the Buddhavacana to an area where a different dialect was used, and who were therefore faced with the task of translating into a new dialect. If the donor dialect (i.e. the dialect from which the translation was being made) and the receiving dialect (i.e. the dialect into which the translation was being made) had differences of phonology and morphology, then much of the translation process could have been done almost mechanically. For example, if the translation was being made from a dialect which did not voice intervocalic consonants into one which did voice them, than the translator simply had to voice all unvoiced intervocalic consonants. If the translation was being made from a dialect where the nominative singular of short a-stem nouns was in -e into a dialect where the nominative singular was in -o, then the translator had simply to change all nominative -e endings into -o.

There would, however, have been certain items which remained unchanged: 1) words which had a specific sanctity attached to them because they were regarded as technical or semi-technical terms, and were therefore, despite the air of strangeness which they must have presented, too important to change; 2) words which had no equivalent in the receiving dialect, and which therefore had to be retained; 3) words retained by an oversight.

The accuracy of such a translation process depended on the knowledge and ability of the translator. Clearly, the translator had to know something about the characteristics of the donor and receiving dialects. If, by any chance, the donor dialect was one which voiced intervocalic consonants, and the receiving dialect did not voice them, then an mechanical translation technique was likely to produce errors, because it would lead to a situation where consonants which should be voiced in the receiving dialect might become unvoiced. It was all very well changing all past participles in -ida in the donor dialect into -ita in the receiving dialect, but what about a form like uppāda? Since there is a root padas well as a root pat-, should this be translated into uppāta, or left as uppāda? Only the sense of the passage could determine this, and if the passage was ambiguous, or for some other reason it was difficult to understand the meaning of the passage, then a translator was left to guess, and his guesses might not always be correct. If the receiving dialect had the nominative singular of short a stem nouns in -o, then to turn all -e endings into -o may well have meant that forms which should have had the -e ending in that dialect, e.g. locative forms, were also changed, incorrectly, into -o.

Let us look at some of these anomalous forms and see what we can deduce from them.

It has been pointed out that the Buddha’s mother Māyā was probably not called “Delusion”, but “Mother” (Mātā). ¹³⁸ This development of intervocalic -t ->-y - is a characteristic of Māhārāstrī, but it is scarcely conceivable that this detail of the Buddha’s life story was transmitted from Magadha to Mahārāstra (where our evidence is in any case from a later time - probably after the writing down of the Pāli canon) before it entered the Theravādin tradition.

The change of intervocalic consonants to -y - is not a feature of the eastern dialect of the Aśokan inscriptions - there are only one or two examples, which can probably be explained otherwise ¹³⁹ - but the -ikya forms found at Kālsī, where the other versions have -ika, may reflect some such change. We do not know precisely what the writing of the ky ligature means, but it is possible that the scribe was writing it because he received k in his exemplar, but wanted to show that his own pronunciation or the pronunciation in his area was nearer y. Lūders has given some examples of -ika/-iya alternations, ¹⁴⁰ and the fact that the change of -k- > -y- occurs after -i - seems to support the view that ikya at Kālsī does indicate that the change >i y a had taken place or was beginning to take place. ¹⁴¹ It is, therefore, possible that the sound change found in Pāli Māyā reflects the fact that in some part of Magadha, perhaps in the West, towards the Kālsī area, this change was operative.

Moreover, it has been shown that some of the etymologies in the Sabhiya-sutta of the Pāli Sutta-nipāta depend upon their being first pronounced, i.e. composed in, not just transmitted through, a dialect where some, at least, of the intervocalic consonants had developed into y.¹⁴² The age of the sutta is shown by the fact that a Buddhist Hybrid Sanskrit version of it exists in the Mahāvastu, where the bhikkhu is called Sabhika - the -iya/-ika variation of the name in the two versions being in itself an example of the very change I am talking about. If the age of the text supports the view that the version underlying the Pāli and Buddhist Hybrid Sanskrit versions was composed in Magadha, then this is additional evidence for the existence of this sound change in Magadha at an early date.

Similar considerations would apply to the examples of voicing found in Pāli. It seems clear that they are evidence for transmission through a dialect where voicing was usual, and such hyper-forms as uppāta < Sanskrit utpāda show that the translator had knowledge of a dialect where voicing occurred. Voicing is a typical feature of Śaurasenī, but there would be considerable problems if we had to assume transmission through that dialect. Although it is not a consistent characteristic of any of the Aśokan dialects, nevertheless there is some evidence for voicing in the Aśokan inscriptions, e.g. libi for lipi, thuba for thūpa, and the hyper-forms which occur, e.g. the root pat- written for pad-, show that the scribes had knowledge of a dialect where voicing occurred.

We find a very small number of forms which show an anomalous development of a consonant group containing a nasal: hanta > handa; the Buddha’s charioteer name Chanda > Channa; ālambana > ārammana. These changes are typical of the Gāndhārī dialect as seen in the Gāndhārī Dharmapada, but although this text is said to show the characteristics of the Gāndhārī dialect several centuries earlier than the manuscript itself (which dates perhaps from the third century C.E.), these features are not found in the version of the Gāndhārī dialect which we find in the Aśokan inscriptions in the North-west. They therefore seem to be a later development in that dialect, and it is most unlikely that these forms could be borrowings into Pāli from the Gāndhārī dialect. I should rather favour the view that since our knowledge of early dialect geography in India is so unreliable, there may well have been unattested dialects which left some mark upon early Buddhist texts.

The ending -e in place of -o is perhaps the most widely attested of the anomalous forms. It is a standard feature of the Māgadhī dialect and the eastern versions of the Aśokan inscriptions, and it also occurs in the Gāndhārī dialect and the Sinhalese Prakrit. We find in the Digha-nikāya descriptions of the teachings of the six teachers who were contemporary with the Buddha, and some of these descriptions include nominative singular forms in -e,¹⁴³ as well as a number of other anomalous forms.

Another place where such -e forms occur is in the framework of the Kathāvatthu, a work which is acknowledged by the Theravādin tradition to be one of the last additions to the canon, since it is said to have been recited at the time of the third sanggiti which was held at Pātaliputra during the reign of Aśoka. We know, then, both where and when it was composed, and it is therefore not surprising that we find in it a number of features of both phonology and morphology which coincide with the eastern versions of the Aśokan inscriptions, e.g. the way in which evam occurs with an emphatic h-, i.e. hevaṃ. ¹⁴⁴

One of the interesting features of this text is the way in which the translators seem to have understood very well the difference between the predominantly western Pāli forms and the eastern forms which are found on almost every one of its 600+ pages. The text consists of a discussion of certain statements, of which 500 were orthodox and 500 unorthodox according to the commentary, although these numbers must have been used in a typical “round number” way, since, in its present form the text, in fact, contains less than 300 unothodox statements. They are set in a framework of question and answer. All the statements, heretical or otherwise, are free from anomalous eastern forms in -e, except for one stock phrase which is repeated a number of times. The framework, however, shows two speakers using -e and -o dialects which were probably completely differentiated at one time, but which, by the time the text had become fixed in its present form, had become mixed in a consistent way. We cannot say whether from the time of the text’s composition one of these dialects used exclusively -o forms, as it does in the present version. If it did, then this would be evidence for the existence of a dialect with -o forms in Magadha c. 250 B.C.E. The original differentiation could, however, have been between two sub-dialects of Māgadhī, which later translators thought were too similar to be easily distinguishable by non-Māgadhī speakers.

Among the anomalous forms in Pāli, attention has been drawn recently ¹⁴⁵ to the word je, which is a particle used when addressing women of a lower class. It is explained as being a shortened form of the word ajje, which would be an eastern form of ayya < Sanskrit ārya “noble”. I have great doubts about this etymology. It may be possible to find evidence for a word which originally meant “noble” being used in a pejorative sense, but I would find it difficult to accept this derivation when the word ajje from which it is said to have been derived was still in full use as an honorific form of address in the eastern dialect in which je must have developed its pejorative sense. I personally believe that je is the emphatic particle, which appears as ye in Pāli and the Aśokan inscriptions, ¹⁴⁶ and I believe that it is an example of the very rare change of initial and intervocalic y>j in Pāli, which is seen also in the word jantāghara. My doubt about the derivation < ajje is supported by the occurrence in Jain Prakrit of hamje, which is used in a very similar sense to je.¹⁴⁷

There are also hyperforms based upon the development of y > j. John Brough ¹⁴⁸ remarked: “They were aware of the Prakritic tendency to voice intervocalic stops of the literary language, and in attempting to combat this tendency, they occasionally overreached, and produced monstrosities such as Yamataggi for Jamadagni”. From the existence of the name Yamataggi in the Theravādin canon we can deduce that the person responsible for the production of this form was aware of the fact that the dialect from which he was translating sometimes showed initial j - where his own dialect showed y-, and intervocalic -d - where his own dialect showed -t - When he came across the form Jamadagni (or more likely Jamadaggi), he did not know its correct form in his own dialect, doubtless because it was a name new to him, whose meaning and etymology were unknown. In the absence of any knowledge of the correct form of the name in his own dialect, he was obliged to back-form by rule, which led to the hyperform. This shows that there was a pre-Pāli dialect where the changes y→j-¹⁴⁹ and -t→-d - occurred. This conclusion supports the possibility that the words showing the change of initial y→j - which I have already mentioned are dialect words in the Theravādin canon.

There is another word in Pāli which seems to be an example of this anomalous sound change, i.e. niya < Sanskrit nija “own, belonging to oneself”. This may simply be an example of an intervocalic consonant being elided and replaced by a glide -y-, i.e. it may be a genuine Māhārāṣ̣̣rī-type form < nija, but it is also possible to explain it as a hyper-form, i.e. a translator who knew that intervocalic -y - became -j - in the donor dialect, wrongly back-formed the -j - which he found in his exemplar into -y-, presumably not recognising the fact that in this particular case nija was correct in the receiving dialect also.

Brough was doubtful about the date when this sound change was operative. He said: “The mere existence of the [Pāli] form Yamataggi then forces upon us the conclusion that parts at least of the Pāli canon were translated from a MIA dialect in which initial y - had already become j-. This seems to demand a seriously late date. But I can only pose the question …”. ¹⁵⁰

Brough’s problem was that he knew of no evidence for the existence of a dialect which turned y into j, either initially or intervocalically, at the time of the Buddha. We know that some at least of the Jain canon was transmitted in such a dialect, where the stem of the relative pronoun is, for example, j a, where Pāli has y a, and the optative ending is -ejja, where Pāli has -eyya, but the Jain tradition tells us that the canon was not written down until some nine centuries after Mahāvīra’s death, in the fifth century C.E., and although parts, at least, of the Jain canon certainly existed before that date, there is no reliable way of dating any of the phonological features of the languages of the Jain canon. The answer to Brough’s question is that we must surmise that there was an earlier dialect where this change took place, and we have, by chance, a minute portin of evidence to support this view. It has been noted that there is one occurrence of an intervocalic y becoming j in the Aśokan inscriptions, ¹⁵¹ and that is in the word for “peacock” in RE I(G). It occurs as majūla at Kālsī and Jaugada, with the eastern I replacing the r of Sanskrit mayūra, and as majura in Sh and M, with the expected western r. The appearance of j at four sites (the fifth, Girnār, has mora, just like Pāli) suggests very strongly that j was taken over without correction by the scribes at all those sites from the exemplars which they received. The dialect at Sh has a tendency to turn -j - into -y-, and would therefore be unlikely to do the opposite, unless there was a good reason for doing so.

The change of -y- > -j- is, however, anomalous in the Aśokan inscriptions. It seems not to be a feature of the dialect which Aśoka’s secretariat at Pātaliputra used, and it presumably represents a remnant of Aśoka’s own dialect, which was probably a sub-dialect of Māgadhī, i.e. a dialect spoken in just the area where we would expect the early language of Buddhism to be in use. It seems very probable, then, that these words with j in place of y came into Pāli from an eastern dialect. This might have been the actual dialect used by the Buddha, but it was at any rate one of the dialects into which his teachings were translated at an early stage in the history of Buddhism.

It is to be noted that this deduction, made on the basis of one single word, implies that Aśoka’s own dialect differed from the eastern dialect attested in his inscriptions, because if we assume that his dialect changed y > j, then the nominative singular of his form of the relative pronoun would be je, whereas we know that the eastern versions of his inscriptions have a form without initial consonant, i.e. e. This implication need present no difficulties, because there is also evidence that Aśoka’s own dialect had palatal ṣ for the sibilant, ¹⁵² whereas no version of the Aśokan inscriptions, other than those from the North-west, has this sound except by scribal idiosyncracy.

From this, then, we can deduce that at the time of Aśoka, and earlier in the case of some Pāli texts which we can, with great probability, date to a preAśokan time, almost all the sound changes which we have marked out as anomalous in Pāli can be shown to occur somewhere in the East, although sometimes only as a single example in the Aśokan inscriptions.

We must therefore conclude that the information we have hitherto had about dialects in Aśoka’s time, let alone at an earlier time, is deficient, since it by no means tells us about all the dialects which were in use. The fact that, if we look carefully, we can augment this information by detecting hints of other dialects is, in itself, of great importance, because it means that we can use the same technique of marking out anomalous forms as we have employed in the Pāli canon, to mark out the anomalous forms in the eastern versions of the Aśokan inscriptions, in an attempt to find the dialect which Aśoka himself used - a dialect which is not represented, in its entirety, in any version now extant.

If our conclusion is correct, then, we can assume that the dialects in use, even in a limited area such as Magadha, showed variations. This should not surprise us, because we should expect neighbouring areas, even neighbouring villages, to have very slightly divergent dialects. We can also assume that traces of this linguistic diversity were retained when the sermons, which had been preached in different areas, were first collected together and their language was homogenised. We can guess - but it is nothing more than a guess - that when the first collection was made and homogenisation began to take place, examples of divergence were even more numerous, since, as sub-dialects of the Magadha area, these variations were probably not sufficiently great to cause difficulties for speakers of other sub-dialects, and there was therefore no need to remove them.

Our question must be: on the assumption that the process of homogenisation was continued, while Buddhism was still, for the most part, confined to the Magadha area, so that fewer and fewer of the anomalous forms remained, why do we find that some of those forms survived the major translation into a western dialect which must have taken place at a later date?

Some of these anomalous forms that I have been discussing are called, by some, “Māgadhisms”, because they conform to the pattern of Māgadhī, as described by the grammarians. Others, however, say that the term Māgadhism is misleading, because it takes for granted that these forms are taken over from Māgadhī. ¹⁵³ It is true that it is probably not entirely accurate to regard all such anomalous forms as Māgadhisms, in the sense of their being words which were originally in a Māgadhī version of the Buddhavacana, but were retained deliberately or accidentally at the time when translation took place into a nonMāgadhī dialect.

Some of these forms are certainly found in other dialects, e.g. nominative singular forms in -e occur in the Gāndhārī Prakrit, probably as the remnant of a linguistic area which was split into two by a later tribal movement, after which the two halves migrated to very different areas, one to Gandhāra and the other to Magadha. Nevertheless, to suggest that the North-west was the source of such forms would demand a complete re-appraisal of our ideas about the way in which the Buddha’s message was spread.

It has also been pointed out that other explanations can be given for some of these forms, but some of these alternative explanations seem unnecessarily complicated. To say that pure < Sanskrit purah was taken over into Pāli as puro, but then became pure by analogy with agge, etc., or by the dissimilation of the vowels u and o, leads to a situation where we have to assume that a Māgadhī form was converted into the correct western form, but then underwent a change which converted it back to the same form which it had in Māgadhī, but we are not, nevertheless, to regard it as a Māgadhism. ¹⁵⁴ There is unlikely to be agreement about such views, since there is no way of proving or disproving either hypothesis. In such circumstances, I should like to propose a theory of “the economy of development”, which means that an explanation of any sound change by a single stage of development is more likely to be correct than an explanation by a multiple development. If we adopt this theory, then to take it as a Māgadhism provides a simpler explanation.

What is, however, clear is that every Māgadhism does not prove that the passage in which the Māgadhī form occurs is old and dates from the Māgadhī period of Buddhism, and certainly those who say ¹⁵⁵ that we can rely too much on such details are correct. Once a standard procedure had been adopted, e.g. of writing bhikkhave in a specific context, and it was applied to newly created texts, the occurrence of such Māgadhisms tells us nothing about the original language of the text in question, i.e. the occurrence of a Māgadhism in a text does not prove that the text was originally in the Māgadhī dialect. There was a great deal of extention of use by analogy. In just the same way, as we shall see in the eighth lecture, phrases in common use in canonical texts do not prove that a text is canonical, because such phrases were used by medieval writers to give their texts a veneer of canonicity.

A recent discussion of some of the features I have just been talking about was entitled “From colloquial to standard language”. ¹⁵⁶ I find the use of the word “colloquial” slightly strange, because to my ear it has something of a pejorative sense. If the Buddha preached in the dialect of the local people wherever he went, then we would say, I think, that he made use of the local vernaculars. Many of these features also occur, as we have seen, in the language of the eastern versions of the Aśokan inscriptions, which is normally identified as the administrative language of the secretariat, and which can scarcely be described as colloquial. The same discussion describes the difference between the particle je and the feminine vocative ajje as representing a distribution of colloquial and standard language. ¹⁵⁷ Again, this usage of “standard” is strange, in the first place because it is not justified, but it is simply taken as fact that there was a standard language in Buddhism, and in the second place because the term “standard” is not defined. We are left to deduce that it means a language without “colloquialisms”, on which, perhaps, all existing versions of Buddhist IndoAryan dialects and languages are based. It is hard to see how Pāli could fit into such a pattern, because, as we have seen, Pāli does include such forms. Perhaps “standard” means a literary type of language, but no evidence is given that such a dialect ever existed. I would doubt that changes were made by translators simply because some words were thought to be “colloquialisms” and therefore “non-literary”. It seems much more likely that eastern forms were unacceptable, for phonological and morphological reasons, to the western Prakrits into which the early Buddhist texts were translated, and they therefore had of necessity to be changed, except when there were reasons for their retention.

It is clear that many of these anomalous forms were retained because they were technical or semi-technical terms, e.g. bhikkhu, bhikkhunī, nibbāna, bhūnahū, ¹⁵⁸ with the last three showing a replacement of -n - by -n - which in the Aśokan inscriptions is typical of the eastern dialect(s). These words were, in fact, part of the basic vocabulary of early Buddhism. We can surmise that the spelling of khana < Sanskrit ksana “opportunity” was retained because of its repeated occurrence in the phrase “do not let the opportunity pass you by”. The expected western form chana exists, but in the sense of a particular sort of opportunity, namely “a festival”. The particle je was doubtless retained in its eastern form because the genuine western form ye was not used in this particular sense in the receiving dialect into which the sermons were translated, although ye is used as an emphatic particle after infinitives in Pāli in exactly the same way that je is used in Jain texts. ¹⁵⁹ A word like āvuso also had a semi-technical sense, in as much as the Buddha had laid down the specific circumstances in which it was to be used. ¹⁶⁰

It seems to me that the words attributed to the six teachers probably reflect (despite the view of some to the contrary ¹⁶¹ ) the actual dialects of their teachings, at least as they were remembered at the time of the composition of the texts. In support of this suggestion is the fact that comparable views ascribed to heretical teachers in Jain texts show close verbal similarities. Other forms were probably kept by mistake, or because the translators did not recognise the verbal root, e.g. āvudha in place of āyudha “weapon”. In some cases it seems clear that the sense was ambiguous, and the translator could not determine the correct form, e.g. there is a verse in the Theragātha, ¹⁶² where we have three forms ending in -e, sacce, atthe and dhamme, which can be eastern nominative or western locative case forms, and we cannot be certain how we should translate them. The Pāli commentary takes all three as locatives, but this seems rather forced, and if we look elsewhere for an interpretation we have a choice of translating “In truth the meaning and the doctrine are grounded” or “Truth is grounded in the meaning and the doctrine”. ¹⁶³ The BHS translator took it in the second way, and changed one -e form to -am, making a nominative satyam.

After examining the anomalous forms in Pāli, we can therefore say that we have evidence that the texts of the Theravādin canon was transmitted through a mixture of dialects or sub-dialects, almost all of which can be shown, or can be surmised, to have been employed in the East at the time of Aśoka, and probably earlier.

I said earlier that the great majority of forms in Pāli are what we would call western, using this in a linguistic rather than a geographical sense. This would indicate that the last recension before the writing down of the canon was in an area where a western style Prakrit was in use, at least for literary purposes. This was likely to be in the West of India, but not necessarily so, since, if an eastern dialect could be used in the West, as we have seen at Sopārā, there is no reason why a western dialect should not be used in the East of India. Theoretically, the recension could have been made at the time of writing the canon down in Sri Lanka, but this suggestion causes problems, because the Sinhalese Prakrit which we find in the inscriptions of the first century B.C.E. does not resemble Pāli very closely - it has, for example, eastern nominative forms in -e - and if we are looking for an area where a western dialect was used, then this does not seem to be a likely candidate.

The presence of Sanskrit forms in Pāli used to be taken as evidence for the belief that Pāli was one of the oldest of the MIA dialects. This idea was based upon a view that MIA showed a steady progress from Sanskrit to the last form of MIA, i.e. Apabhramśa - the very latest stage before the emergence of the New Indo-Aryan languages. A dialect which retained a proportion of Sanskrit forms was therefore thought to be closer to Sanskrit, not only in form but also in time. As more was learned about MIA, however, and as more and more texts were found in Sanskrit or various forms of Sanskrit, owing a smaller or larger debt to underlying MIA dialects, it became clear that this view was wrong. As some of these Sanskrit forms in Pāli were recognised to be incorrect, and not genuine Sanskrit forms at all, e.g. attaja is Sanskritised as atraja, instead of ātmaja, it became clear that the Sanskrit forms in Pāli were late additions to the texts, i.e. they represented the result of a limited re-introduction of Sanskrit or quasiSanskrit forms into the MIA original, not the preservation of old forms.

We have no direct evidence as to where and when the Sanskritisms were introduced into the Theravādin canon. Although a start had probably been made before Theravādin texts were taken to Sri Lanka, nevertheless, we believe that the greater part of the Sanskritisms were introduced in Sri Lanka, if only because we would start to date the start of Sanskritisation rather late, probably not before the second century B.C.E.

Whether other changes, beside Sanskritisation, were being made to the language of the canon at that time we do not know. Since it is clear from the commentators’ explanations that updating of the language did occur, e.g. present participles with the old historical nominative singular ending -am in the canonical texts are explained by younger forms with the analogous ending -anto in the commentaries, we may well be correct in believing that such things could also happen in the canon, although not in verse texts, where the metre would act as a constraint upon any changes which would alter the metrical length of a word.

We have no evidence that Pāli, either with or without Sanskritisms, coincided with any historical language or dialect. It is not clear what we should call such a language. Some ¹⁶⁴ call it “artificial”, and certainly there are artificial features, such as the incorrect back-formations just mentioned. The English language, however, has similar artificial features, e.g. the g in sovereign, by analogy with reign, the s in island by analogy with isle, or the p in receipt because it is derived from Latin receptum, but no one regards English as an artificial language, reserving the term for such invented languages as Esperanto. Others call the Pāli language “literary”, and certainly it is the language of a literature, although it is not literary in the sense that it represents the refined form of a popular dialect.

However we choose to describe Pāli, the chronicles tell us that in the first century B.C.E. the Pāli canon was written down in Sri Lanka, and it is with the use of writing and the effect which it had upon Buddhist texts and Buddhism itself that the fifth lecture is concerned.

V - Buddhism and Writing

I want in this lecture to say something about the impact which writing had upon Buddhism, and about the deductions which we can make about the changes which took place when Buddhist texts began to be written down - first, the changes in those texts and second, the changes in Buddhism itself.

The Dipavamsa, our earliest authority for the statement that the canon was written down, states: “Before this time (i.e. the reign of King Vattagāmini Abhaya, in the first century B.C.E. ¹⁶⁵ ), the wise bhikkhus had handed down orally the text of the three pitakas and also the atthakathā. At this time, the bhikkhus, perceiving the loss of living beings, assembled, and in order that the religion might endure for a long time, they recorded (the above-mentioned texts) in written books”. ¹⁶⁶ It is not clear whether “perceiving the loss of living beings” is to be taken generally - “everyone is going to die” - or whether it refers to the specific events of the time. If the latter, then it could well be a reference to the partial breakdown of the bhānaka system. I mentioned in the third lecture that Buddhaghosa records the fact ¹⁶⁷ that, at some time after the introduction of Theravādin Buddhism into Sri Lanka, there came a time when only one bhikkhu knew the Niddesa, and from fear of its disappearing completely a thera was persuaded to learn it from this one bhikkhu, and other theras learnt it from him. This probably made the sangha in Sri Lanka realise that the whole canon could disappear if the oral tradition died out.

Nevertheless, it is probable that there were other pressures too. It is very likely that the so-called Brāhmanatissa famine, foreign invasion from South India and also the political and economic circumstances of the times played a part in persuading the bhikkhus to make this decision. It has also been suggested that the growing power of the newly founded Abhayagiri vihāra, and the threat which this offered to the Mahāvihāra, could not be ignored.

Other, later, sources give more information about the writing down of the canon, presumably making use of information either unknown to, or at least unused by, the author of the Dipavamsa. Some authorities talk of a joint recitation (sañgīti) being held in the Āloka-vihāra before the writing down, and this is referred to as being “like a fourth sanggiti” by some sources. ¹⁶⁸ This may, of course, be an invention dating from a later time when there was some doubt about the authority of the written canon, because it appeared not to have been authenticated by a sanggiti. In this connection we should remember the pattern at the fifth sanggāyana in the nineteenth century and the sixth sanggāyana in the 50 -s of this century: The canon was recited and then what had been agreed was carved and printed, respectively. This clearly was the pattern which was expected if an authentic version of the canon was to be produced. It is, of course, possible that the committing of the canon to writing was a rather drawn-out affair, and was not the result of a single sanggiti, which is why there is no mention of one in the earliest texts. If this was so, then it is certain that the story of the sanggiti was a later invention. There is, however, little doubt that we can accept that the writing down of the tipitaka during the reign of Vattagamini Abhaya was an historic fact. ¹⁶⁹

The Dipavamsa gives no information about the way in which the tipitaka was written down. We must assume that what was currently being remembered and recited was repeated in the presence of scribes, who wrote it down from dictation. Unless the various sections of the bhānakaka tradition had been cooperating and had been making simultaneous changes, e.g. in the Sanskritisations which had been introduced, we must assume that immediately prior to writing down the canon there were divergent features in the nikāyas. Writing down would have been an excellent opportunity for the homogenisation of forms - all absolutives in -ttā being changed to -tvā, and the forms containing -r - being standardised, etc. The references to the bhānakas in the commentaries suggest that the oral tradition continued to function alongside the written tipitaka, and if the bhānakas continued to act in the same way as before, they had perhaps to relearn or, at least, revise their nikāyas, with these homogenised readings in them.

In the case of the Pāli tradition, then, we may assume that the writing down of the canon may have had some effect on the inter-relationship of the various bhānakas traditions, if they were compelled to co-operate in this way. It is noteworthy that, as I said in the third lecture, there are differences of readings in parallel passages in different nikāyas, as we have them now, and we find that the use of stock phrases, e.g. the formulae used to describe bhikkhus approaching the Buddha, etc., are not identical. We may conclude that any editorial process which occurred was not complete, but was restricted to certain particulars.

The date when writing might have first been used by the Buddhists depends, of course, on the date of the introduction of writing into India, and there is no agreement about this among scholars. As is well known, writing was widely used by the emperor Aśoka c. 260 B.C.E. in two scripts: Kharoṣthi and Brāhmī. The former, it is agreed by all scholars, I think, was derived from the Aramaic script which was used by the administrators of the Persian Empire, and its use was confined to the areas of North-west India which were at one time part of the Persian Empire, i.e. the region around Gandhāra, and also Chinese Turkestan into which the script was taken from Gandhāra. We have no inscriptions in Kharoṣthi earlier than Aśoka, but the references made by the Greek historians to the use of writing in the North-west of India at the time of Alexander presumably refer to it, and if Pānini refers to writing, then since he came from the North-west, it was probably Kharoṣthi that he meant.

There are, to simplify the discussion, three views about the origin of the Brāhmī script. ¹⁷⁰ It is undoubtedly the case that the script used by the Indus Valley civilisation c. 2000 B.C.E. contained a small number of characters which were identical with, or very similar in appearance to, characters in the Aśokan form of Brāhmī (although, of course, until the Indus script is deciphered we do not know if the phonetic values were the same). Some scholars, therefore, think that writing continued from the Indus Valley civilisation right through to the third century B.C.E. Others, basing their theory on a small number of characters which seem to have parallels, of both form and phonetic value, in an early Semitic script, believe that writing came to India from further west. The third view is that the script was invented by Aśoka’s scribes, or those of his immediate predecessors, specifically for the purpose of inscribing imperial edicts. There are objections to this view, in as much as the variety of forms which we find in the Aśokan version of Brāhmī suggest that there was already a history of development of the script before Asoka’s scribes used it. Support for this idea of pre-Aśokan development has been given very recently by the discovery of sherds at Anuradhapura in Sri Lanka, inscribed with small numbers of characters which seem to be Brāhmī. These sherds have been dated, by both Carbon 14 and Thermo-luminescence dating, to pre-Aśokan times, perhaps as much as two centuries before Aśoka. ¹⁷¹

It is hard to believe that traders, operating throughout the Near and Middle East and meeting fellow-traders from countries where writing had been used for centuries, would not see the advantages of adopting a writing system. The same applies to administrators. Once the North-west of India had become part of the Persian empire, if not before, writing would have been employed by the Aramaic scribes in that area, and it is likely that neighbouring rulers would realise the advantages of keeping records of regnal years and royal accounts, and treasury and armoury details, in some tangible form. Aśoka’s so-called “Queen’s edict” (if my interpretation of it is correct) ¹⁷² specifically states that certain charges are to be set against the queen’s name, in a way which implies that this system of accounting was well known.

My own belief is that although one correspondence in both shape and sound between Brāhmī and an early Semitic script might be simply chance, it is beyond the bounds of coincidence for what I regard as at least three certain parallelisms of both form and phonetic value to have occurred. ¹⁷³ I therefore take it as certain that at least those three characters of Brāhmī are connected with the Semitic script.

There are other similarities of form between Brāhmī and early Semitic, but the phonetic values are not the same. This partial correspondence does not, however, present an insurmountable difficulty. There is precisely the same problem with the Carian alphabet which was based upon the Greek alphabet, but did not always have the same phonetic value as the Greek. What was borrowed was the idea of writing, with a small number of direct parallels. This is not to say that these Brāhmī characters were necessarily borrowed from the Semitic script at the time for which we have attestations of its use. There is, in principle, no objection to believing that they were borrowed from a later, at present unattested, version of that script.

I would therefore support the view that the Brāhmī script was in use in India well before the time of Aśoka, and was probably brought from the West by traders, who had borrowed the idea of writing, and some of the characters, from Semitic traders.

There is, however, no certain evidence for the use of the Brāhmī script in India before the time of Aśoka - there is no agreement among scholars about the handful of inscriptions found in India which have been claimed to be preAśokan. It is, therefore, appropriate to ask why, if writing existed in India before the time of Aśoka, there are no records of it. The probable explanation is that writing was not at first used for religious or literary purposes, but exclusively for Administration and trading. Since there was no need for the documents of administrators and merchants to last for ever, they were written on ephemeral materials, and therefore perished.

Similarly, although the Gandhāra area had been a part of the Persian empire since the time of Darius, there are no remains of pre-Aśokan Aramaic inscriptions from that area, presumably because such records were similarly written on ephemeral material. Only kings had inscriptions carved on rock “that they might long endure” “as long as the sun and the moon”, as Aśoka says - and there appears to be no reason to reject the much-repeated suggestion that the wording of some of his edicts seems to contain echoes of the words of Darius, which suggests that Aśoka’s inscriptions were a direct result of his having heard about the inscriptions of Iranian kings at Behistun and elsewhere.

Even if writing was not used until the time of Aśoka, it is surprising that the Theravādin texts were not written down until 200 years after Aśoka’s time. Since many converts to Buddhism were kṣatriyas and vaiśyas, who would have been acquainted with the use of writing for accounts and lists of merchandise or military equipment, etc., it is inconceivable that not one of them made use of his earlier knowledge after he had become a bhikkhu. The opposition of the bhānakas would probably have prevented any attempt to transmit entire texts by writing, but it would be odd if sermon notes or aids to recitation ¹⁷⁴ were not written down. There could be no religious reason for not using writing. As I said in the third lecture, there is no mention of writing in the Vinaya rules, either to regulate or forbid its use.

Moreover, by the first century B.C.E., writing in a semi-religious context had been widely used for a century or more in Sri Lanka, for donative inscriptions, etc. It is hard to imagine that monks sitting in their caves with an inscription on the drip ledge did not think of the value which writing might have for the perpetuation of the Buddhavacana.

We must then ask if it is an incontrovertible fact that Buddhists did not make use of writing until the whole canon was written down in the first century B.C.E., or whether there is any evidence at all for the use of writing before the time of the so-called fourth sangiti. There are, in fact, hints which have led some scholars to suggest that writing was indeed used, if only to a limited extent. Following Lūders, ¹⁷⁵ John Brough ¹⁷⁶ compared a pāda in the Gāndhārī Dharmapada ¹⁷⁷ which ends in so vayadi “he goes” with the corresponding pāda in the Udāna-varga, ¹⁷⁸ which ends in sa vrajati, with the same meaning, and contrasted these forms with va sayati “he lies down” in the Pāli equivalent, ¹⁷⁹ and he pointed out that in the context the idea of moving is more likely to be correct than lying down. He explained the Pāli form as showing the metathesis of the two akṣaras va and sa, and noted that although such a metathesis might occur in a purely oral tradition, it would imply “an unbelievably slipshod paramparā”. In manuscript copying, however, this is a common and readily understandable error. The conclusion he arrived at was that the text was already being transmitted by manuscript copying, and not exclusively by oral tradition, at a date earlier than the redaction of the Pāli version.

It is not always possible to be certain whether such an error is the result of an oral or a written tradition. Even the belief that an error is due to writing does not, in itself, prove that the error was earlier than the writing down of the canon, since it could well have occurred at some later date when the manuscript was being copied. We should, however, note, that if this is the explanation, we have to assume that it happened at a time when manuscripts of any individual text were very few in number, and probably unique. It can be shown that this might well have been the case, for Alsdorf has pointed out that in one Jain text ¹⁸⁰ the whole tradition available to us, not only all extant manuscripts but even the oldest commentaries, go back to one single individual manuscript in which, through an ordinary clerical mistake, one akṣara had been omitted. ¹⁸¹

Similarly, although Brough wrote of a slipshod paramparā, we cannot rule out completely the suggestion that there might have been such an imperfect method of transmission. If all manuscripts of a particular Jain text and the commentaries upon that text go back to a single, unique manuscript, then, in theory, at least, all manuscripts of a particular Pāli text and the commentary upon it, may go back to a single, unique bhānaka, and, as I have said, we know that that was the situation with regard to at least one Pāli canonical text. In short, I agree with Brough that writing was already being used before the whole canon was written down, but I am not certain that the evidence he gave proves it.

Others too ¹⁸² believe that writing had already begun to be used on a small scale, and we may assume that its use increased, as it became more and more common in the secular world. In the absence of firm evidence, however, it is always possible for alternative suggestions to be made about the cause of errors in the transmission, and scholars rarely agree either about the probability of such errors being due to transmission by means of writing, or about the date when they might have occurred.

I mentioned in the third lecture some of the errors which, despite all the care which the bhānakas took in their reciting, might creep into the texts because of mispronunciations.

Writing, however, brought with it a whole new range of possible errors, of the sort with which we are familiar from the manuscript tradition of classical studies: writing the same syllable or group of syllables twice (dittography); omitting one of a pair of consecutive identical syllables or groups of syllables (haplography); allowing the eye to jump from words at the end of one line or verse to the identical words at the end of another line or verse, with the resultant loss of the words in between (homoioteleuton), etc., and a glance at the critical apparatus of any Pāli text will provide abundant examples of all these errors occurring in the manuscripts. Such errors are found from the very beginning of writing in India, and examples of all of them can be seen in the Aśokan inscriptions.

There was another type of error, however, which was due to deficiencies in the early Indian writing system. In the earliest form of the Brāhmī script, double consonants were not written, and the marks for long vowels were frequently omitted.

There are, in the Pāli canon, a number of textual variations showing alternative forms with single or double consonants, which are most easily explained as having been written at some time in such a script. When, in a later form of the Brāhmī script, the facility of writing double consonants was adopted, in places where the metre was of no help in determining the length of a vowel, e.g. in prose or in metrically doubtful positions, a copyist had to decide whether the single consonant in the exemplar, which he was copying, stood for a single or a double consonant. In most cases the context would make this clear, but in passages where both forms could be considered to yield some sort of sense, it was a matter of personal preference which form he chose to write or how he chose to interpret it. The tradition sometimes remained ambivalent, with both possibilities being handed down to us.

The writing of single consonants after vowels in metrically doubtful positions can explain how doublets of this kind could arise. ¹⁸³ In the Therīgāthā we find a verse which states, “There is no release from dukkha for you, even if you approach and run away”. This seems slightly odd. The word for “approach” is the absolutive upecca ¹⁸⁴ (< Sanskrit upetya), and the commentary sees the difficulty and explains that here it means sañcicca “having considered”, i.e. “after consideration” or, perhaps, “deliberately”. The commentary, however, also states that there is another reading, namely uppacca, which it explains as uppatitvā “having flown up”. This seems to make better sense in the context “there is no release from dukkha for you, even if you leap up and run away”, and it is very likely that this is the reading we should adopt, especially as it is the only reading found for the verse in some of the other places where it occurs. Here we can postulate the existence of an earlier form with single consonants, i.e. *upaca. Since the first syllable is in a metrically doubtful position at the beginning of a pāda, the scribe had no way of telling whether the word began with up- or upp-. In the absence of such knowledge, upaca developed to *upeca when -a- was palatalised to -e- before -c-. When the possibility of writing double consonants arose, the tradition interpreted *upaca as uppacca ( < Sanskrit utpatya) and *upeca as upecca ( < Sanskrit upetya).

In a prose passage in the Niddesa ¹⁸⁵ we find a long list of adjectives describing ascetic practices, e.g. hatthi-vattika “following the practice of elephants”. There are variant readings for each epithet, with a single -t-, e.g. hatthi-vatika “taking the elephant vow”. In Sanskrit we find -vratika in similar contexts, ¹⁸⁶ but since the difference in meaning between vratika and vrttika is very small, a scribe who was accustomed to a particular writing system, where single -t - was written for both vatika and vattika, might have difficulty in deciding which was the correct spelling, when the possibility of writing double consonants arose.

Sometimes we find that a variation for a word appears, not in the commentary, but in the scribal tradition, e.g. in a verse in the Theragāthā, ¹⁸⁷ where the Pali Text Society edition reads nimmissam, with double -mm-, there is a variant reading nimissam, with single -m-. The first syllable, being the first syllable of the pāda, is in a metrically doubtful position, so the metre cannot help us to decide between single and double -m-. In Pāli, nimināti “he exchanges” and nimmināti “he constructs” are sometimes confused, ¹⁸⁸ because in certain contexts their meanings converge. In this verse, however, it seems certain that we should adopt the variant reading nimissam with single -m-, since “exchange” makes better sense in the context, which talks of “exchanging the ageing for agelessness”. The word was presumably written as *nimisam in an earlier version, and because of the doubt about the metrical length of the first syllable, either nim- or nimm- was possible as an interpretation, and different branches of the scribal tradition transmitted different readings.

Not only did the early Brāhmī script write double akṣaras as single, but it also frequently omitted anusvāra, e.g. the active limpati “he smears” and the passive lippati (< lipyate) “he is smeared”, could both be written as lipati, since the passive ending -ate was often replaced by the active ending -ati in Middle IndoAryan. Exactly the same could happen with the active muñcati and the passive muccati < mucyate, both of which might be written as mucati. This would have caused no confusion as long as the oral and written traditions continued side by side, as they must have done for some time. But when the oral tradition ceased, scribes were then entirely dependent on what was written, which might be ambiguous, unless the context gave help. It is certain from the nature of some alternations that scribes were entirely dependent on written forms, with no help from any oral tradition.

Faced with a form mucati, a scribe who did write doubled consonants and anusvāra had to choose between writing muccati and mumcati. If the context did not make the choice obvious, then mistakes might occur. In a verse in the Patnā Dharmapada, ¹⁸⁹ for example, we find the passive mucceya, where other versions of the pāda have the active stem muñc-. There is considerable doubt about the meaning of this particular verse among modern scholars, and it is clear that some ancient copyists were equally perplexed, and were uncertain about the way in which to interpret the reading in their exemplars. It is noteworthy that this possible confusion of active and passive continued, to some extent, even after the use of double consonants and anusvāra became standard, and we find in Pāli that muñcati is sometimes used for muccati, ¹⁹⁰ and vice versa.

There is, similarly, doubt about the alternation between the active limpati and the passive lippati, to such an extent that limpati seems almost to acquire a passive sense. The root lip means “to smear”, and in the passive “to be smeared, to be defiled, to cling to”. Although the pattern is not entirely consistent in Pāli, it would appear that the case usage with this verb polarised, with the instrumental being used with the sense of “defiled by (something)”, and the locative with the sense of “clinging to (something)”, e.g. water clings to, or does not cling to, a leaf. In the second usage, where there was an absence of an obvious sense of passivity, a doubt possibly arose in the copyists’ minds as to whether the word lipati which they found in their exemplar was not perhaps an active form, and they consequently wrote the active limpati. We consequently find that the Pāli manuscript tradition is quite uncertain about this verb, and limpati and lippati are sometimes found as alternative readings.

We can get some idea of the difficulties which such a writing system caused if we examine a verse in the Gāndhārī Dharmapada, which is written in the Kharoṣthī script, which has similar deficiencies, in that it does not write double consonants or long vowels. When we try to interpret the words kamaradu and kamaramu in the Gāndhārī Dharmapada, ¹⁹¹ we are unable, because of the nature of the Kharoṣthī script, to decide whether the first part of these compounds stands for kāma- or kamma-. If we look for parallel or near-parallel passages in Pāli, to help us make a decision, we find that both kāmarata and kammarata occur there. ¹⁹² The occurrence of the two forms would seem to be a clear indication that at one time that portion, at least, of the Theravādin tradition was transmitted through a script which did not write either double consonants or long vowels. Consequently the first part of the compound was written as *kama-, which could be variously interpreted.

It has been pointed out ¹⁹³ that the so-called ablatives in -am in Middle IndoAryan may be due to a transmission through manuscripts which, like the script of the Gāndhārī Dharmapada, did not mark vowel length or nasalisation or, like the Brāhmī script of the Aśokan inscriptions, did not do so consistently, so that from the point of view of the written form an accusative of a short -a stem noun, which should be -am, and an ablative, which should be -ā, would be identical, i.e. -a. We should expect that, when a text written in such a manner was rendered into a full orthography, whether at the time of translation into Pāli or subsequently, this would for the most part be done correctly, since the scribe’s knowledge of the language would enable him to interpret from the context. Equally, however, it would not be surprising if, from time to time, either because his attention was wandering, or because the passage was genuinely difficult or ambiguous, the scribe misinterpreted his exemplar, and wrote -am where he should have written -ā.¹⁹⁴

We find similar problems arising from a type of orthographic ambiguity which occurs in some later manuscripts, where the copyists used dots to denote anusvāra and also to indicate doubled consonants. ¹⁹⁵ This is not, strictly speaking, the use of an anusvāra to indicate doubling, but the use of a dot which could be confused with the dot used for anusvāra. The end result of this scribal habit was similar to that which I have already mentioned. Without help from the context, it is difficult in a Middle Indo-Aryan text to tell whether, say, lipati, written with a dot above the -p - in such a manuscript, stands for lippati or limpati.

There is a similar potential ambiguity found in manuscripts written in Burmese script, where the i-mātrā is a circle above the consonant, and the anusvāra is a (smaller) circle above the consonant. ¹⁹⁶ This therefore sometimes leads to a confusion between -i and -am. This phenomenon sometimes gives us a very useful criterion when investigating the early history of a particular manuscript tradition. In Sinhalese manuscripts we sometimes find -am where we should expect -i, and vice versa. Such an error should not occur in manuscripts in the Sinhalese tradition, because in the Sinhalese script -i and -am do not resemble each other at all. If, therefore, we find in a Sinhalese manuscript an error which we can explain as being due to the confusion of the two aksaras, we can be certain that at some stage in its transmission the text was copied from a Burmese manuscript, ¹⁹⁷ and the copyist was mistaken in his identification of what he was copying.

I spoke in the third lecture of the value which repetition had in an oral tradition. It is likely that another result of writing the canon down was that scribes began to realise that there was no need to write out all the repetitions in full, since it was quite easy for a reader to look back to the place in the manuscript where the passages had occurred before, whereas it was not so good to say, when reciting, “I am going to leave out the next few paragraphs because I said them ten minutes ago”. There arose, then, the practice which we find in the manuscripts of writing the syllables pe or la which are abbreviations for peyyāla (< Skt paryāya), which means something like “formula”. In a text then, it means “(here comes a) formula, i.e. a repeated passage”, which is the equivalent of “and so on down to …”. Relying on such precedents, early European Pāli scholars also made abbreviations in their editions of Pāli texts, writing par” for pariyādiyati, etc. This was condemned by a number of Sinhalese scholars ¹⁹⁸ who pointed out that “to interfere, either with words or letters, otherwise than is done by the peyyālams made use of by the Arahats (who recited the texts at the various recitations), has frequently been declared to be not good”. Consequently, volume I of the Pali Text Society’s edition of the Añguttara-nikāya, which abounded in such abbreviations, had to be withdrawn and a new edition made.

Such indignation was perhaps an over-reaction to “western tampering with sacred texts”, because these pe and la aksaras are not always in the same place in different traditions, e.g. the oldest Pāli manuscript we know of - four folios of the Vinaya from Nepal, which are about 1,000 years old - sometimes agrees with the Pali Text Society edition in writing pe or la in certain places, ¹⁹⁹ but in other places gives no hint that there is an omission, even though the Pali Text Society edition does write la. ²⁰⁰

It is sometimes not made clear in western editions whether the omissions which editors have made are those authorised by this ancient abbreviation system, or whether they are the editors’ own innovations. Because the passages which have been abbreviated seem to vary, in this way, in the different manuscript traditions, someone who recently wished to make some calculations about the amount of repetition there was in a single sutta had to go to a great deal of trouble to compare different traditions of that sutta, to try to establish something approximating closely to the original unabbreviated version, with all its repetitions kept intact.

The etymology of the word peyyāla, with its eastern l for r, perhaps indicates that it came into use in Buddhism while the texts were still in an eastern dialect. That may mean that the texts had already been written down before the translation into a western dialect took place, or alternatively that this was a technical word, used in this sense in non-religious writing in both eastern and western dialects (as a borrowing from the East in the latter), and taken over into religion from that secular source. In just the same way we find āva (< Sanskrit *yāvat ) “(and so on) down to” used in the Aśokan inscriptions, and also the Jain Prakrit form jāva used in the same sense in Jain texts.

The fact that some characters were very similar to others in the Aśokan Brāhmī script led to errors being made by the scribes who perhaps misread a badly written aksara in their exemplar. So, for example, because of the close similarity of ca and va, we sometimes find that one version of an Aśokan Edict has va where the others have ca, and vice versa. The same is true of pa and sa, and ta and na, etc. Comparable errors are found in texts written in the various scripts of South and South-east Asia, which are derived from the Aśokan Brāhmī, and those working with manuscripts written in such scripts quickly learn the characters which, because of their similarity, may be confused by the copyists.

As the shapes of the aksaras developed, it sometimes happened that pairs of aksaras, which at an earlier date were capable of confusion because of their similarity, became dissimilar in shape, while pairs of aksaras which had originally been quite different, became more similar. Despite their common origin, the Sinhalese, Burmese, Thai and Cambodian scripts sometimes show widely divergent patterns of development in the shape of aksaras, so that two characters which are very similar in one script are perhaps quite different in another. This enables us to gain certain information from orthographical errors. If, for example, we find an error in script A which seems to have arisen from the confusion of two aksaras which only resemble each other in script B , then this strongly suggests that the tradition was dependent, at some stage of its transmission, upon a manuscript written in script B.

The change from oral to written transmission led to certain changes being made to features which were oral in origin, e.g. those which are dependent upon a variation, or ambiguity, in pronunciation. The introduction of writing meant that it was no longer possible to be ambiguous. In the Suttanipāta we find, in two successive verses, puns dependent upon an non-aspirated and an aspirated form: pakkodano “I have boiled my rice” says Dhaniya the herdsman, and akkodhano “I am free from anger”, says the Buddha punningly. Writing this pun causes no difficulty, but if the pun depended upon two pronunciations of the same word, then there are problems, because writing forced copyists to choose one spelling or the other, in the same way that Sanskritising copyists, as we shall see in the sixth lecture, had to choose between two Sanskrit forms when translating puns which depended upon Middle Indo-Aryan homonyms. So in a verse in the Dhammapada ²⁰¹ we find “For a bad deed done is not released immediately, like freshly extracted milk”, where “is released” is muccati. Although we might think that the verse has something to do with release from bad kamma, and milk being released from the cow, this really makes little sense.

The commentary, however, explains the verb as meaning “develops” (parinamati), and we can see that it must originally have been mucchati ²⁰² “it coagulates”, i.e. a bad deed does not develop (its result) immediately, just as new milk does not coagulate. In the parallel versions, the Patna Dharmapada ²⁰³ reads mucchati and the Udāna-varga ²⁰⁴ reads mūrchati, so it is clear that the idea was understood in those traditions. It would appear that the Pāli tradition thought that the context of the first part of the line, with the mention of kamma, needed the verb muccati, while the second part needed mucchati, i.e. there was a pun dependent upon a non-aspirated and an aspirated pronunciation. The copyist could, however, only write one of the two and he chose, wrongly, to adopt muccati, and this incorrect reading is found, as far as I know, in all manuscripts of both the Dhammapada and its commentary.

As is well known, the oral transmission of the Vedas continued for many centuries after the time of Aśoka. As far as we know, although writing was available it was not adopted for Vedic purposes. Similarly, although the Theravādin tradition states that the Pāli canon was written down in the first century B.C.E., oral recitation of Pāli texts continued and still continues to this day. We cannot tell how long the official bhāṇaka sort of oral tradition and the written tradition continued side by side, but it was perhaps not very long. The writing errors, which I mentioned earlier, and which seem to indicate that the oral tradition had been lost, must have occurred at an early date, because they could only happen as long as double consonants (and long vowels) were not written and we know that the facility for doing this was soon developed.

Writing down also had an effect upon the contents of the Theravādin canon. There is some doubt about the state of the canon when it was written down. We do not, for example, know whether it was complete, i.e. whether it was in the form in which we find it today, or not, and we may well wonder whether any texts contain anything which would enable us to identify it as material added after the canon had been committed to writing.

As far as I can judge, once the Theravādin canon had been written down, very little further change was made to it. From the point of view of its language, we should have expected anything added in Sri Lanka to show traces of the local Prakrit, but there are few signs of borrowings from Sinhalese Prakrit being inserted into it, and most of the borrowings which have been suspected can be explained otherwise. ²⁰⁵ The process of Sanskritisation remained incomplete, which suggests that nothing further was done after the canon had been written down. Nevertheless, the explanations given by the commentaries in later centuries show that the canon was not fixed absolutely by the process of writing it down. The system of reciting and approving the form of suttas at sangītis should have resulted in the elimination of all variations, but it is clear that the writing down of the canon did not, in itself, lead to the disappearance of readings which had not been preferred at the various sangītis, and this would support the view that the system of bhānakas continued, in some form or other, for some time, with some reciters still keeping alive readings not found in the written texts. It seems that such “unauthorised” readings, if still recited, could easily creep back into a text, if a scribe ²⁰⁶ heard them and reintroduced them while he was copying the text. This would account for the number of variant readings which we find mentioned in the commentaries.

Despite the separation of the Sinhalese Buddhists from North India, ²⁰⁷ it seems that literary material continued to reach Sri Lanka, but there is no evidence for the addition of any complete text to the Theravādin canon after the Ālokavihāra sangiti. An origin in North India is postulated for the Milindapañha, ²⁰⁸ the Pefakopadesa and the Nettippakarana. The fact that these texts are highly regarded by the commentators, but are not given canonical status, ²⁰⁹ suggests that they arrived in Sri Lanka after the closure of the canon, which presumably occurred at the time when it was committed to writing. This view is supported by the fact that these post-canonical works contain a number of verses and other utterances ascribed to the Buddha and various eminent theras, which are not found in the canon. Nevertheless, it seems that there was no attempt made to add such verses to the canon, even though it would have been a simple matter to insert them into the Dhammapada or the Theragāthā.

Although there are references to writing in the canon, there are only two which refer to religious texts being written. They are in the Vinaya, and both of them are in the Parivāra, the appendix to the Vinaya. At the end of the first eight (of 16) sections of the first chapter, there is a statement that “these eight sections are written in a manner for recitation”, ²¹⁰ while at the end of the book, after the words “the Parivāra is finished”, ²¹¹ there is a statement that Dīpanāma, having asked various questions about the ways of former teachers, thinking out this epitome of the details of the middle way of study, had it written down for the bringing of happiness to disciples. It seems, however, unlikely that if the whole of the Parivāra was composed after the time the canon was written down, i.e. if it was a late addition to the canon, it would have had those two references to writing inserted in the way they are. I can see no good reason for doubting that the two statements are interpolations, added when the Parivāra, together with the rest of the canon, was written down.

One result of the writing down of the canon was that the stranglehold of the bhānakas was broken, and the control which they had had over the contents of the nikāyas was relaxed. At the beginning of the oral tradition they must have observed the Buddha’s instruction about the four mahāpadesas, ²¹² the four references to authority: the Buddha, a community with elders, a group of elders, and a single elder, and checked whether any teaching said to have been obtained from these sources was consistent with the sutta and vinaya which the sangha already had. As time passed, however, and less and less hitherto unknown teaching was brought forward for acceptance, we can assume that the canon, or at least the various nikāyas which were in the hands of the bhānakas, was closed and no further additions could be made, except, perhaps, in such small ways as adding extra apadānas to the Apadāna or perhaps extra verses to the Thera- and Therī-gāthā. There was no way of inserting an entirely new sutta into a nikāya unless the bhānakas of that nikāya could be persuaded to accept it.

In the third lecture I spoke of the bhikkhu Arittha, who, you will remember, totally misunderstood the purport of the Buddha’s teachings about stumbling blocks. ²¹³ As long as the bhānakas recited the sutta which made the position clear, then any attempt which Arittha (or others who thought like him) might have made to pretend that his view was authorised by the Buddha, by inventing a sutta which authenticated it, would be unsuccessful, because the bhānakas would not have admitted a new sutta which included a view condemned by the Buddha and which was, therefore, not consistent with the rest of the canon.

We might also think of the bhikkhu Sāti, who, as we read in the Majjhimanikāya, ²¹⁴ so misunderstood the Buddha’s teaching that he thought it was “consciousness” (viñ̃̃ana) which continued in samsāra. ²¹⁵ This would appear to be a recollection by Sāti of a teaching similar to that found in the BṛhadĀranyaka Upaniṣad that vijñāna continues: ²¹⁶ “This great being, endless, unlimited, consisting of nothing but intelligence”. This view was refuted by the Buddha, who pointed out that he had frequently taught that “Without a cause there is no origination of consciousness”. ²¹⁷ Once again, Sāti had no chance of inserting his view into a nikāya.

Nor would it be possible for such views to be inserted into the nikayas, even after the introduction of the use of writing, as long as the bhānaka tradition continued alongside writing, as I have suggested it did, for some time, in the Theravādin tradition.

On the other hand, once the various schools of Buddhism had started to make use of writing, it was not difficult to produce a text and say that it was Buddhavacana, as long as there was no bhānaka tradition to refute the claim. You will know that the Buddha stated, as reported in Theravādin texts, that there was no “teacher’s fist”, ²¹⁸ as far as he was concerned, i.e. he was not keeping anything back for an élite, but was making his teaching known to all of his followers who wished to listen. There developed, however, a view in some Buddhist schools that what the Buddha had said openly was intended only for the masses. There was really another, hidden, meaning which the Buddha imparted only to a chosen few. For such schools, the adoption of writing enabled them to claim that their views were indeed Buddhavacana.

Richard Gombrich has dealt with this in the context of the rise of the Mahāyāna, [^5-55] but it is possible that not just Mahāyāna, but also dissident Hinayāna sects benefitted from the use of writing. Since no one ever accuses the Abhayagirivāsins, the opponents of the Mahāvihāravāsins, of transmitting their scriptures in some language other than Pāli, it is probable that their canon was in Pāli. Nevertheless, their version of the Buddhavamsa, as far as can be judged from the small portion which is preserved in Tibetan, does not agree with the Theravādin Buddhavamsa, and must therefore have been added to their body of scriptures at a time when the bhānaka system had been by-passed. It is quite possible that they were able to add to their scriptures in this way, by making use of writing. The way in which the so-called quasi-canonical texts came into existence in the Middle Ages in Sri Lanka and South-east Asia is another indication of the way in which the fact that texts were written down enabled them to be accepted by some Buddhists, at least, in a way which would have been impossible if the bhānakas were still transmitting all texts orally.

Paradoxically, then, if the writing down of the Theravādin canon may be presumed to have stopped the further Sanskritisation of Pāli and prevented the insertion of new suttas into the nikayas, it was writing which made possible the production and acceptance of Mahāyāna and other texts. We may suppose that, to a very large extent, the advent of writing meant that an already existent canon was fixed when it was written down, but writing allowed new canons to come into effect because the authors did not have to point to a long bhānaka tradition of the texts, which alone, before the use of writing, could prove that they were Buddhavacana.

In view of this connection which has been seen between writing and the Mahāyāna, it is not unreasonable to believe that the writing down of the Theravādin canon was not due simply to a threatened breakdown in the bhānaka system of transmission in Sri Lanka, and the social, political and economic conditions of the time, as the Pāli commentarial tradition suggests, and as I proposed earlier, but also to a need to give an authenticity and prestige to the Theravādin canon vis-ā-vis the written texts of other schools. If the beginnings of the Mahāyāna, and therefore religious writing, can be dated to the second century B.C.E., then it is likely that Hīnayāna texts were also being committed to writing, in North India if not in Sri Lanka, at that time.

Since the social and religious conditions which had existed at the time of the Buddha had by then changed greatly, and since Pāli and the literary forms of other Middle Indo-Aryan dialects were now almost as much out of touch with the languages of the common people as Sanskrit had seemed 400 years before, it is probable that many such texts were being written in Sanskrit - the language of culture - as opposed to being translated into it.

VI - Buddhism and Sanskritisation

In the fifth lecture I referred to the fact that the writing down of the Theravādin canon in the first century B.C.E. seems to have put an end to the Sanskritisation of that canon which had begun in the decades before the reign of Vattagamini Abhaya, although, paradoxically, writing seems to have led to the appearance of other texts, which, as far as we can tell now, were actually composed in Sanskrit.

In this lecture I want to consider the reasons for the process of Sanskritisation, the way in which it was effected and the result which it had.upon Buddhist texts and Buddhism itself.

I must start with a definition. What exactly do I mean by Sanskritisation? I use the word in two senses. In the first place, I talk, in a broad sense, about the Sanskritisation of Buddhism, when I am discussing a particular phenomenon, namely the way in which Buddhism, which had started as a revolt against the social and religious system which was exemplified by the use of Sanskrit for literary and religious purposes, now began itself to embrace Sanskrit as a medium for the propagation of the Buddhavacana.

In the second place, Sanskritisation means the use of Sanskrit in Buddhist texts as a replacement for the dialects of Middle Indo-Aryan in which the Buddha’s teachings had previously been transmitted for some hundreds of years. In this sense, the term is applicable to the whole range of Buddhist texts starting from those in a Prakrit which contains a very small amount of Sanskrit, or Sanskrit-like, forms in it, through a range of texts which are in a variety of languages which might be regarded as Sanskritised Prakrit or Prakritised Sanskrit, sometimes called Buddhist Hybrid Sanskrit, to texts which are in pure classical Sanskrit, in accordance with Pāninian grammar. ²¹⁹ The language of the last group is classified by some as Buddhist Sanskrit, because the texts are written by Buddhists about some aspect of Buddhism or Buddhist history, and perhaps contain items of vocabulary which are specifically Buddhist. We can therefore classify Sanskritised texts under three headings: (1) texts written in a Middle Indo-Aryan dialect into which some Sanskritisms have been inserted; (2) texts originally written in a Middle Indo-Aryan dialect which have been translated into Buddhist Hybrid Sanskrit or Buddhist Sanskrit; and finally (3) texts composed in Buddhist Hybrid Sanskrit or Buddhist Sanskrit.

In this lecture I want to concentrate upon those aspects of Sanskritisation which are exemplified in the first two of these classes. I shall be considering for the most part the problems which arose when texts in Middle Indo-Aryan dialects were converted into other dialects with a Sanskritic content, small or large, or into Sanskrit. I shall not be referring to those texts which were composed as original works in Buddhist Hybrid Sanskrit or Buddhist Sanskrit, except in so far as some of the vocabulary they employ perhaps owes something to Middle Indo-Aryan.

This pattern of Sanskritisation was not restricted to Buddhist texts. We find something very comparable in the language of the inscriptions found in North India. The earliest inscriptions, those of Aśoka, are written in a variety of Prakrit dialects. Those of the early centuries of the Common Era are in a mixture of Prakrit and Sanskrit. From the fifth century C.E. onwards inscriptions are written in classical Sanskrit. If we examine the inscriptions in the middle phase at Mathurā, we find that they do not show a steady progression from Prakrit to Sanskrit. What we know of the languages of North India, during the period covered by the inscriptions, indicates that the local population spoke some dialect of Middle Indo-Aryan. Donors would presumably dictate in Prakrit what they wanted to have carved on their donations, and as it became fashionable to write inscriptions in Sanskrit the scribes would “translate”, to the best of their ability, into that language. An inscription in bad Sanskrit is, therefore, not necessarily older than one in good Sanskrit. The difference in quality may simply represent the ability of one donor to employ a better educated scribe than the other. Similarly, the fact that one Buddhist text is in better Sanskrit than another does not necessarily mean that it was Sanskritised at a later date.

Why did Sanskritisation begin? If we consider the partial Sanskritisation of the language of the Theravādin canon, we must, I think, agree that there is no obvious reason why Sanskritisation should have started in Sri Lanka in the first century B.C.E. We have no information to make us believe that there was a strong pro-Sanskrit movement in the island then. There is, for example, no hint that the Abhayagirivihāra, the rival of the Mahāvihāra, was making use of Sanskrit at this time. The Sanskritisation of Pāli can hardly have started spontaneously, in the absence of any reason, and we must assume that it started under the stimulus of Sanskrit elsewhere, presumably in North India. There was no obvious reason why a language used in Sri Lanka should have been influenced by anything happening in North India, and so we must assume either that the process of Sanskritisation started in the Theravādin canon before it was transported to Sri Lanka, or, more probably, that there was still a strong connection between North India and Sri Lanka after the introduction of Buddhism to the island.

The Pāli chronicles report that Buddhism was taken to Sri Lanka during the reign of Aśoka. We know that, for secular purposes at least, Prakrit was preferred to Sanskrit at the time of Aśoka, and it was only after his death that Sanskrit began to regain its position of predominance, although we should perhaps note the Sanskritising tendency of the Aśokan scribe at Girnār, even if we are unable to decide whether the Sanskritisms there are features of the local language (which might have been rather archaic) or insertions by a Sanskritising scribe. In favour of the latter, we should note that there is a strong case for believing that the -r - forms at Girnār owe more to the scribe’s views of what was appropriate than to the actual vernacular spoken in the area. ²²⁰ If we believe that the second century B.C.E. saw the appearance of new Buddhist texts, setting out new teachings, composed by new sects, in Sanskrit, the language of culture and literature, then we might well believe that that was the time when, in competition with these new Sanskrit works, those schools of Buddhism which had hitherto used a Middle Indo-Aryan dialect, began to make changes in the languages of their teachings in order to rival the growing use of Sanskrit by other sects, and to make their teachings available to the same classes of readers.

There were various degrees of Sanskritisation: (1) in its simplest form it involved the partial restoration of Sanskrit phonology, perhaps the restoration of a number of Sanskrit forms to consonant groups, and possibly in addition the restoration of long vowels which had been shortened in Middle Indo-Aryan, e.g. before consonant groups; together with (2) the restoration of Sanskrit morphology, e.g. “correct” verbal and nominal endings; and perhaps (3) the substitution of Sanskrit vocabulary in place of Middle Indo-Aryan vocabulary. This might include the removal of Eastern forms which, as I pointed out in the fifth lecture, represented some of the oldest elements of Buddhist vocabulary, e.g. āvuso.

Further levels of Sanskritisation would include: (4) the insertion of the correct Sanskrit sandhi forms, i.e. those demanded by the rules governing word juncture. We find that hiatus is avoided by the insertion of particles, or by the rearrangement of the order of words, or by a change of vocabulary, (5) and, in metrical texts, there is the avoidance of irregular metre by changes of word order, or of vocabulary.

When we come to examine the Sanskritisations in Pāli, we have first of all to decide whether they are in fact Sanskritic features which were inserted by the redactors, or whether they might not be original features of the language, i.e. are the Sanskritisms in Pāli retentions or restorations of Sanskritic features? If we thought that such Sanskritisms were remnants of an earlier form of the language of the canon, then we could regard them as genuine archaic features in the language, and we might then define Pāli as one of the oldest (linguistically speaking) of the Middle Indo-Aryan dialects, in as much as it was a form of Sanskrit with some Middle Indo-Aryan developments in it.

An investigation, however, shows that some of these Sanskritisms in the Theravādin canon are incorrect back-formations, e.g. atraja, which is probably a misinterpretation of attaja (< ātmaja). ²²¹ Although some of the forms with the consonant group br - are correct, e.g. brāhmaṇa, others of them are non-historic, e.g. brūheti. ²²² Even some of the forms which are correct, e.g. the absolutive ending -tvā and br - in brāhmaṇa, can be shown to be unoriginal in certain contexts. There is evidence for the absolutive ending -ttā,²²³ which we can assume was the regular absolutive ending in the dialect before the restoration of the -tvā ending, and both the metre and the etymologies which are given for brāhmaṇa in the canonical texts ²²⁴ show that at an earlier time the word occurred with initial b-, not br-. We can therefore conclude that these forms, and probably all other Sanskritic features, are deliberate attempts at Sanskritisation, made at some time during the course of the transmission of the canon. It is therefore clear that it is not correct to speak of them as retentions. They are features which have been restored to the texts by scribes or reciters who were trying to change into Sanskrit the language which they had received in their exemplars.

A close examination of the Sanskritic features in the Theravādin canon suggests that the process took place in two phases, separated by some centuries.

To the early phase we can allot the Sanskritisation of some consonant groups, e.g. tv in the absolutive ending -tvā, which I have just mentioned, and tr in the suffix -tra of the locative pronouns, and the st in utrasta and bhasta, etc. ²²⁵ Some of these changes were probably made as early as the writing down of the canon, and they are certainly earlier than the commentaries, because the commentators refer to the variation between the suffixes -ttā and -tvā. It is noteworthy that r is not restored in groups with p. The fact that there are no forms with pr in Pāli, in contrast to the Girnār version of the Aśokan inscriptions, where pr - is the most common group containing -r-, suggests that, despite all claims to the contrary, there is probably no connection between Pāli and the Girnār dialect of the Aśokan inscriptions.

Other Sanskritic features include the retention or restoration of long -ā-, e.g. in certain compounds and derivatives of vāk “speech”, even when the resultant form of the word goes against the pattern of the dialect by producing a long vowel before a consonant group, e.g. vākya, which by the rules might have been expected to develop into *vakka, with a short -a-.²²⁶ The fact that vākya and comparable forms are back-formations is supported by the existence of other derivatives of vā k where the expected shortening of long -ā- does occur before a double consonant. We can deduce that the redactor did not recognise the existence of the word vāk in such compounds, and consequently did not restore the long vowel.

A later phase of Sanskritisation took place under the influence of the grammarians, centuries after the first phase, and probably for the most part after the appearance of the Saddanīti, Aggavamsa’s grammar of Pāli which was written in the twelfth century. Some of these Sanskritisations are in the field of vocabulary or morphology, and many are concerned with sandhi. Sandhi in Pāli is quite different from Sanskrit sandhi, in that it is much more flexible. Since final consonants have disappeared, sandhi in Pāli consists of the contraction of the final vowel of a word, including a nasalised vowel, with the initial vowel of the following word. Such contractions produce a wide range of crasis vowels, which are not always predictable.

The grammarians, with their extensive knowledge of Sanskrit, tried to bring Pāli more in line with Sanskrit grammar, and copyists inserted into the manuscripts the forms which the grammarians prescribed. Final -i and -u which had remained in hiatus were (sometimes) changed to -y and -v. If the final vowel had been -e or -o, then an indication of this was sometimes given by prefixing y or v to the vowel which had evolved by contraction. The fact that these changes are later additions to the texts, and are not the result of the earlier phase of Sanskritisation, is shown by the fact that the manuscripts are rarely consistent about such insertions. If the contraction of final and initial vowels led to a long vowel occurring before a consonant group, this was sometimes kept, even though it offended against the law in Middle Indo-Aryan which demanded a short vowel in this position.

It is clear that the reasons for these two phases of Sanskritisation of Pāli were quite different. It is probable that the reason for the first was the appearance of other schools using Sanskrit, as I have suggested, but the reason for the second is that grammarians who knew enough about Sanskrit to be able to model their grammars upon Sanskrit, and in fact to quote extensively from the Sanskrit grammarians, thought that they should try to make Pāli more like Sanskrit.

Sanskritisation caused problems of a different nature from those which arose in the period of oral tradition. The nature of the development of Middle IndoAryan meant that there was a increase in the number of homonyms. A glance at Sheth’s Prakrit-Hindi Dictionary (Pāiasaddamahannavo) shows, for example, that there are at least ten Sanskrit words which can develop into Prakrit saya-, and, if we look at compounds, there are no less than 48 Sanskrit equivalents for the Prakrit compound para-(v)vāya. Although a specific context would immediately rule out many of these, it is obvious that a redactor translating into Sanskrit might well have difficulties when trying to decide between homonyms.

Problems arose even in the limited amount of Sanskritisation which we find in Pāli. The problem for the redactor was to decide which of two or more possible forms to choose when Sanskritising. There is sufficient evidence for us to deduce that in the language of the Theravādin canon before it was Sanskritised the absolutive ending was - tā, as I have mentioned. This would be identical with the form in Ardha-Māgadhī, the language of the Jain canon. A form such as kattā < Sanskrit krttvā “having done” was therefore identical in form with, i.e. was a homonym of, kattā, the agent noun “a doer” from the same root kr - There is a pāda which occurs twice in Pāli which in its earlier form contained, we can deduce, the word kattā. In one context kattā was taken as a verb and Sanskritised as katva. In the other the tradition interpreted it as the noun kattā, and so retained it in that form and explained it as a noun. ²²⁷

Exactly the same happened with the word chettā, from the root chid- “to cut”. Here, in one and the same context, the tradition is ambivalent. In the edition of the canonical text we find the absolutive chetvā “having cut”, but the commentary reads the noun chettā “a cutter”, and explains it accordingly. ²²⁸

When a Sanskritising recensionist was faced with the word santa, he had the option of assuming that it was the present participle of the verb “to be”, which gets the meaning of “good” - you will know that the feminine of this (satī) is used of the good woman who is so devoted to her husband that she immolates herself upon his funeral pyre - or the past participle of the verb śam “to rest”, i.e. the equivalent of Sanskrit sānta “at rest, peaceful”. In some cases it is possible, even probable, that in a Middle Indo-Aryan dialect santa was intended to be ambiguous, but in Sanskrit one or other of the two meanings has to be preferred.

The same problem can arise with the Middle Indo-Aryan word dīpa, which may be the equivalent of Sanskrit dīpa “lamp” or dvīpa “island”. As a parallel to the verse in the Dhammapada ²²⁹ “a wise man should make a dīpa which the flood will not overwhelm”, the Sanskrit Udānavarga ²³⁰ has dvīpa “island, but the Chinese version ²³¹ has “lamp”, showing that it is based upon a Sanskrit version which had dīpa “lamp”.

A verse in the Dhammapada ²³² tells us that someone who has pīti in the dhamma sleeps happily. The word pīti is ambiguous since there are two words pīti in Pāli, one from Sanskrit pīti “drinking” from the root pā “to drink”, and the other from priti “joy” from the root prī “to please”. It seems very likely that both meanings are intended in this verse. The second pāda tells us that he sleeps with a clear mind, perhaps unfuddled by drinking the dhamma as opposed to the intoxication he would have experienced if he had drunk strong drink. In the last pāda of the verse, however, we read that the wise man delights (ramati) in the dhamma, ²³³ which suggests that pīti is also to be taken as “joy”. The Pāli commentator presumably did not see the possibility of the word play, and explains it only as drinking the dhamma. ²³⁴ The redactor of the Sanskrit Udānavarga had to choose between pīti and prīti, and perhaps because of the idea of “delight” in the last pāda, or perhaps because he was following a different commentarial tradition, he decided to read prīti. ²³⁵

There was some doubt about the Sanskritisation of the Pāli word nekkhamma. As the Pali-English Dictionary points out, although the word is derivable from a hypothetical Sanskrit form *naiskramya “going out”, from the root kram- “to go”, i.e. departing from the householder’s state to the houseless way of life, there are also word plays which seem to be based upon the meaning “the state of being without desire” (*naiskāmya) from the root kam- “to desire”. I see no reason to doubt that we are in fact dealing with two homonyms here. In Buddhist Sanskrit, however, we find only the form naiskramya, ²³⁶ even in conjunction with the word for desire (kāmeṣu). ²³⁷

We sometimes find that Pāli and Sanskrit technical terms do not correspond, and in such cases we may surmise that the lack of correspondence may be due to Sanskritisation from a homonym. The Buddhist Hybrid Sanskrit equivalent of Pāli sammappadhāna “right effort” is samyakprahāna “right abandoning”. It has been suggested ²³⁸ that the Sanskrit version is an incorrect backformation from an Middle Indo-Aryan form *samma-ppahāna which might stand for either -pradhāna or -prahāna. In view of the fact that such word-plays are very common in Indian literature, we cannot exclude the possibility that the authors of the passages which contain such words were making using of the ambiguity, and, if this was so in this particular case, we can see that neither the Pāli nor the Sanskrit form is capable of giving the double meaning which the author intended.

Sometimes commentators are aware of the fact that a word can have two meanings, and they incorporate both senses in their explanations. It is suggested by some ²³⁹ that the Buddhist tradition deliberately capitalised on, or even exploited, any ambiguity which a Middle Indo-Aryan form might have, but it seems to me that using such phrases as “deliberately capitalised” is perhaps the wrong way of looking at the matter. In my view it is rather taking advantage of the situation. The speakers of Middle Indo-Aryan dialects were not likely to know that a given item in their vocabulary had developed from two or more Sanskrit words. They merely knew that it had two (often quite different) meanings, and the commentaries tried to make use of this fact when giving their explanations. It was only when someone proficient in Sanskrit tried to Sanskritise that the need to choose between two possible Sanskrit antecedents arose. We should not be too surprised if a recensionist made the wrong decision, especially since, ²⁴⁰ in the example I have just mentioned, Sanskrit pradhāna does not have the meaning “endeavour”.

There are many other similar examples of wrong backformation in partially Sanskritised dialects, such as the language of the Theravādin canon and the Patna Dharmapada, and also in more highly Sanskritised languages. It should be noted that it is not always the Sanskrit version which is incorrect, and we can surmise that there was sometimes a commentarial tradition available to the translators, which gave information about the meaning and interpretation of the words they were concerned with.

The need to choose between homonyms quite often leads to a situation where puns and explanations which worked in Middle Indo-Aryan no longer worked in Sanskrit, or even in a Middle Indo-Aryan dialect with partial Sanskritisation. In the Dhammapada the word brāhmaṇa is explained by means of forms from the root bāh- “to remove”. A brahman is one “who has removed, got rid of, his sins”. ²⁴¹ The connection between brāh- and bāh- is not as close as one would like for an etymology, but the answer to the problem is to deduce that this explanation was originally formulated in a Middle Indo-Aryan dialect where the development from brāhmana was *bāhana, so that the pun on the meanings of the two verbs, which in their Sanskrit forms are brh - “to be strong” and brh - “to remove”, worked perfectly. The Sanskritisation to brāhmana which occurred in Pāli, as I have already noted, had already spoilt the pun. Another pun found in the Dhammapada which does still work in Pāli, is that found in the etymology which is given for the word samana “ascetic”. One is called a samana “because one’s sins have been put to rest”, ²⁴² where the sam- of samana is punningly linked with the root sam- “to be at rest”. In a Sanskritised form, however, samana becomes śramana, which is from the root śram- “to make an effort”, while the Sanskrit form of the root “to rest” is sam-. The pun which worked well in Pāli is lost when we have śramana explained by sam-. ²⁴³

The concept of the pratyeka-buddha is well known as the middle element in the triad: buddhas, pratyeka-buddhas and śrāvakas. The word is usually translated as “a buddha (awakened) for himself”, but this usage of pratyeka (pacceka in its Pāli form) is unusual, and it is not at all clear how it could acquire the meaning which is given for it. The concept of a similar type of Buddha is also found in Jainism, but there, in Jain Prakrit, the name is patteya-buddha. The variation in the name suggests that the concept of pratyeka-buddhas was borrowed into both Buddhism and Jainism, and we can therefore look for an etymology for the word outside Pāli. It seems most likely that the earlier form of the word was pacceya-buddha, where pacceya is to be derived from Sanskrit pratyaya “cause”. A pacceya-buddha was “one who is awakened by a specific cause, a specific occurrence (not by a Buddha’s teaching)”, and both the Buddhist and the Jain tradition give lists of the occurrences which caused the awakening of the four most famous persons in this class. The Pāli form pacceka is an incorrect translation, probably a hyper-form, of pacceya, and pratyeka is a Sanskritisation of pacceka.

Sanskritisation led to hyper-forms, as a result of misinterpretations. Many Buddhist Sanskrit texts are entitled sūtra. To anyone who comes to Buddhist studies from classical Sanskrit studies, this name comes as a surprise, because, in Sanskrit, sūtra literature is a specific genre of literature, composed in prose, usually of a very abbreviated and concise nature, while Buddhist sūtras have an entirely different character. This difference is due to the fact that the word sūtra in Buddhist Sanskrit is a Sanskritisation of the Middle Indo-Aryan word sutta, which is probably to be derived from Sanskrit sūkta, a compound of su and ukta, literally “well-spoken”. ²⁴⁴ It would be a synonym for subhāṣita, which is the word used of the Buddhavacana by the emperor Aśoka, as we shall see in the seventh lecture, when he said: “All that was spoken by the Lord Buddha was well-spoken”.

Another wrong back-formation centres around the same word sutta. There is a verse in the Pāli Dhammapada where we are exhorted to chinda sotam “cut off the stream” (of samsāra), and the Sanskritisation in the Udāna-varga gives the same sense: chindhi srotaḥ. The Patna Dharmapada, however, was doubtless dependent upon a Middle Indo-Aryan version which, through an orthographic variation, must have had the reading sutta, via sŏtta. This was then Sanskritised as sūtra, so the pāda in that version tells us to “cut off the thread”, presumably interpreted as the thread of rebirth, since we find in Pāli that craving (tanhā) is described as “the seamstress” (sibbanī), which joins us to samsāra by means of death and rebirth. ²⁴⁵

One of the best-known examples of Sanskritisation is the word bodhisattva, which is a back-formation from Middle Indo-Aryan bodhisatta. Anyone who knows anything about Sanskrit will realise that the translation which is commonly given for this word “one destined to be a Buddha”, or “one destined for awakening”, is, to say the least, unlikely, and we should note that MonierWilliams gives the basic meaning as “one whose essence is perfect knowledge”, ²⁴⁶ which would be more appropriate as an epithet for a Buddha than for a bodhisatta. It is noteworthy that the Pāli commentaries did not assume that the second element of the compound was the equivalent of Sanskrit sattva. They give derivations either from the root sañj- “to be attached to”, which should have given a Sanskrit form bodhiṣakta “one attached to bodhi”, or from the root śak “to be able” which should have given a Sanskrit form bodhiśakta “one capable of bodhi”. ²⁴⁷

We should note, incidentally, that wrong Sanskritisations are not restricted to Buddhist texts. One of the Sanskrit words for “pearl” is mukta, which literally means “released”. It is, however, more than likely that the Sanskrit form is a wrong Sanskritisation from a Middle Indo-Aryan form muttā, which occurs, for example, in Pāli, so that its Sanskritisation is an example of a folk etymology: a pearl is called “released” because it is released from the oyster. It has been suggested that Middle Indo-Aryan muttā is to be derived from Sanskrit mūrta²⁴⁸ “coagulated, shaped, formed”, but in view of the wide-spread existence of a word muttu in Dravidian, ²⁴⁹ with the meaning “pearl”, it is perhaps more likely that Middle Indo-Aryan muttā is a loan-word from Dravidian.

Resolved consonant groups, those consonant groups which have been separated into their constituent members by inserting an epenthetic (svarabhakti) vowel between them, were Sanskritised by the removal of the svarabhakti vowel, e.g. Sanskrit kriyā “action” became Middle Indo-Aryan kiriyā, and was then Sanskritised back to kriyā again. Sometimes misunderstanding on the part of the translators led to forms such as parṣat being Sanskritised ²⁵⁰ from Middle Indo-Aryan parisā, which was derived from Sanskrit parisat, on the assumption that -i - was an epenthetic (svarabhakti) vowel, and similarly nyāma “rule”, or “regulation” was, in a similar way, wrongly back-formed from Middle IndoAryan niyāma, which was identical with Sanskrit niyāma. This type of error can be seen in the oldest translations which we have in India, i.e. the Aśokan inscriptions, where one of the scribes wrote hveyu in place of the form huveyu, from the root bhū-, which he had received in his exemplar, presumably in the belief that -u - was a svarabhakti vowel.

Although I said earlier that the fact that one Buddhist text is in better Sanskrit than another does not necessarily mean that it was Sanskritised at a later date, nevertheless, in some cases we can see how Sanskritisation developed. We have, for example, an edition of an earlier version of the Udānavarga from Chinese Turkestan, ²⁵¹ which is closer in its form, especially with regard to word order, to its Middle Indo-Aryan original than the later version from Chinese Turkestan, ²⁵² which, in its attention to metre, its observance of sandhi rules, etc., represents a more correctly Sanskritised version of the text. ²⁵³

It is sometimes thought that because a more Sanskritised version of a text is in better, i.e. more correct, Sanskrit, it must also give a better, i.e. more correct, interpretation of the underlying Middle Indo-Aryan text, and we should therefore interpret a Middle Indo-Aryan text in the light of a later Sanskritised version. This, however, is not necessarily a good principle to follow. Since all the versions we have of Hīnayāna texts are translations of some earlier version, the correctness or otherwise of a Sanskrit version depends upon the aids which the Sanskrit translator had available to him when he was making his translation. Where there was doubt about the way to interpret, and hence to translate, a word or passage, a Sanskrit translator who had better guides to interpretation than someone translating into a Middle Indo-Aryan dialect, was likely to produce a Sanskrit version which was more correct than the Middle Indo-Aryan version. If the Middle Indo-Aryan redactor had better guides, then it is likely that the Middle Indo-Aryan version would be more correct. If neither of them had any help, then their translations would be based upon their ability to guess the meaning, and neither translator was likely to be consistently better than the other at guessing.

Reviewers and others sometimes point out that in my translations or studies of Middle Indo-Aryan words and works I have not taken account of the Chinese or Tibetan or Khotanese, or whatever other, version they think is important. This is perfectly true, for a number of reasons: (1) I am deliberately confining myself to translating or dealing with the Middle Indo-Aryan material; (2) I was probably ignorant of the Chinese, Tibetan or Khotanese version’s existence; (3) I could not handle it even if I knew about it; (4) such versions are translations, usually from Sanskrit, which is itself a translation, from some Middle Indo-Aryan version. It may therefore be interesting, as an interpretation which one tradition has put upon the material it has inherited, but it is in no way an authority, to be followed slavishly. It very often happens that such versions are incorrect, because of errors or misunderstandings in the tradition.

A study of the metre of the Buddhist Hybrid Sanskrit version of Upāli’s verses has recently been made. ²⁵⁴ A Pāli version of these verses occurs in the Majjhimanikāya, and portions of a Buddhist Hybrid Sanskrit version of the verses were discovered in Chinese Turkestan and published in 1916. Since then other fragments have been identified and published. The verses are of great interest, for a number of reasons. They give a list of 100 epithets of the Buddha in a metre which occurs only very rarely in Pāli - the old ārya or old gīti metre. An examination of the structure of the verses of the Buddhist Hybrid Sanskrit Upālisūtra shows clearly that the transforming into Sanskrit was made from a Middle Indo-Aryan exemplar in a way which was almost entirely mechanical. We can see that of the features of Sanskritisation which I mentioned earlier, the redactor restored consonant groups, replaced non-historic case endings with more correct endings, and made some changes of vocabulary. He did all this without any regard whatsoever for the metre, and although he made some changes to the sandhi, the frequent examples of hiatus have been reproduced almost without exception. It is obvious that the verses of the Buddhist Hybrid Sanskrit version were originally composed in a Middle Indo-Aryan dialect, but the differences in word order, and indeed in the order of the verses, show that it was a tradition which differed somewhat from the Pāli.

There is also a Chinese version of these verses, which seems not to be based on the Pāli version or any of the Sanskrit versions which we possess. Some of the epithets in the list differ from those we find in the other versions, although it is not clear whether the differences are due to variations in the tradition or to the Chinese redactor’s inability to understand his exemplar. It is very probable that the Chinese translation was made from a Buddhist Hybrid Sanskrit version, which in turn was made from a Middle Indo-Aryan version, probably in a dialect of the North-western Gāndhārī variety, which must in turn have come from another Middle Indo-Aryan dialect, which may have been the dialect used by the Buddha, but was probably not. And yet, despite the long history of translations in the tradition, each of which would have given opportunity for errors to arise, people are very surprised when I say that the Chinese, Tibetan or Khotanese, or whatever version it was, might not have been all that useful to me, if I had been able to read it. I must make it clear that I am talking about Hīnayāna texts whose origin may be presumed to have been in India.

Sanskritisations from the Gāndhārī dialect are always likely to be of doubtful value, because of the wide discrepancy between orthography and pronunciation in that dialect. It is clear that sometimes, at least, the orthography was carried over into translations, irrespective of whether it made sense or nonsense. There is a story found in the Chinese version of the Vinaya of the Mūlasarvāstivādins, where Ānanda heard a monk reciting a Dharmapada-verse which ended with the words, “It were better that a man live only for one day, and see a water-heron”. Ānanda’s efforts to persuade the monk that the verse should have ended with the words “and see the principle of coming into existence and passing away”, were unsuccessful. This Chinese version was obviously following a tradition based upon a Sanskrit form *udaka-baka, which could only come from a Gāndhārítype dialect ²⁵⁵ which inserted a non-historic -k - in place of a glide -y-, in the compound udaya-vyaya ²⁵⁶ “arising and passing away”.

A recent study of the language of the earliest Chinese version of the Saddhārmapundarīka ²⁵⁷ suggests that a number of the idiosyncrasies of the language of that text owe something to the fact that the text seems to have been based not upon the Sanskrit versions which we know now, but upon some other version in a dialect which was very similar to the Gāndhārī dialect. The frequent confusion between jñāna “knowledge” and yāna “vehicle” found in the text seems to indicate that there must have been a variant jāna in place of jñāna in the donor dialect, and a development of this with -y- in intervocalic positions in compounds would have produced forms such as Buddha-yāna in place of Buddha-jñāna. ²⁵⁸ Similarly, the fact that bhoti, the contracted form of bhavati “he becomes”, appears to have been Sanskritised as bodhi “awakening” implies a translation from a dialect which, like Gāndhārī, articulated the aspirate so weakly that it could be written or omitted at will.

Among the features which are very common in Middle Indo-Aryan metrical texts is the way in which verses, although hypermetric, can be made to scan by assuming the very common phenomenon of the resolution of syllables, under the musical influence which allows, in certain positions in many metres, the replacement of a long syllable by two short syllables. When pādas showing resolution are Sanskritised without any further change, and Sanskrit forms, etc., are restored, the redactor cannot compensate for such resolution, and a verse results which has to be categorised by a modern editor as hyper-metric, ²⁵⁹ or irregular. ²⁶⁰ We can sometimes see that the first stage of Sanskritisation produced verses which barely approximated to Sanskrit in form and metre, and a second stage of Sanskritisation was required to produce something which was more correct, by changes of vocabulary or word order.

Another common feature in Middle Indo-Aryan metrical texts is the way in which forms are evolved metri causa, by the lengthening or, less commonly, shortening of syllables, e.g. we find nirūpadhi instead of nirupadhi “without sub-strate”, or anūdaka as a replacement for anudaka “without water” in the cadence of a śloka verse, where the metre requires the short-long-short-long pattern of syllables. When such metrical adjustments are Sanskritised, the earlier Sanskritised Prakrit version writes forms which are still not correct Sanskrit, e.g. niropadhi and anodaka. In a more correct Sanskritised version we find that a different way of dealing with the problem is evolved, often involving the use of vrddhi vowels, e.g. niraupadhi. ²⁶¹

What effect did all these various aspects of Sanskritisation have on Buddhist texts? It is obvious that, by definition, the Sanskritisation of Buddhist texts made them more like Sanskrit, which meant that ambiguous Middle Indo-Aryan phonology and morphology was replaced by precise Sanskrit forms, ambiguities of sandhi and grammar were removed, and the metre was repaired, where it was recognised. This also had a negative side. Where a Middle Indo-Aryan form was ambiguous, the Sanskrit redactor may have chosen the wrong alternative. Where the ambiguity was intentional, perhaps to make a deliberate play upon words, e.g. when giving an etymology based upon a homonym, the Sanskrit redactor could only translate the homonyms into Sanskrit, where they were not homonyms, thus losing the point of the pun. If the redactor did not understand his exemplar, and was forced to translate mechanically or by guesswork, then the results were unpredictable, but were almost certainly wrong.

In the third lecture I spoke of the Waxing Syllable Principle, the way in which words in stock phrases are often arranged in order according to the number of syllables in each word, and I considered the help which this principle might be considered to have given in the memorisation and recitation of texts. The Waxing Syllable Principle was sometimes violated in Sanskritisation, because Sanskritisation changed the syllable count by removing svarabhakti vowels, or by changing the vocabulary, i.e. by replacing Middle Indo-Aryan words by Sanskrit synonyms. ²⁶² The Waxing Syllable Principle was a feature of oral literature, intended as an aid to memorisation and recitation. Since Sanskrit and writing went hand in hand, the fact that Sanskritisation made memorisation more difficult was perhaps of less importance than it might have been at an earlier time.

What effect did Sanskritisation have upon Buddhism? I mentioned in the fourth lecture the suggestion made by some scholars that in the often-quoted passage from the Vinaya, where bhikkhus ask for permission to translate the Buddha’s teachings, the word chandaso means “into Sanskrit”. If that is correct, we must note that the Buddha forbade translation in this way. Centuries later, however, that instruction (if it was indeed his instruction) was ignored, and Buddhists works were indeed translated into Sanskrit. Whether those who made the translation were aware that they were ignoring the Buddha’s command, is an interesting question.

The Sanskritisation of Buddhist texts meant that their readership or intelligibility was not limited to monks who were acquainted with the dialect in which the texts had hitherto been transmitted, and it must, therefore, have made them more accessible to a wider range of readers. If a monk of any individual school could read and understand the (more or less) Sanskritised version of the teachings of his own school, he would be able to read the teachings of other texts too, if in his travels he came to a vihāra belonging to another school, or met a monk of another persuasion. The effect which such increased communication between sects and schools had is a matter for speculation. It has been shown, for example, that the Sarvāstivādins and Mūlasarvāstivādins possessed a number of different versions of the same text, e.g. the Udānavarga, and also remained in close contact, so that a Sarvāstivādin version of a text might be influenced by a Mūlasarvāstivādin version of the same text. ²⁶³ Such influence could more easily be effected if both versions were in Sanskrit, than if they were in two different dialects of Middle Indo-Aryan.

This takes us back to the reason for Buddhism making use of the vernacular languages in the first place. This must have been two-fold: as a rebellion against the brahmanical caste and their language, and to make the Buddha’s teaching available to the common people. The brahmans were the communicating channel between men and gods. They alone knew the rituals and the sacrificial methods to intercede with the gods on behalf of their patrons. The Buddha, by denying the power of the gods, removed the need for a priestly caste to intercede, and taught a system whereby every man could gain his own salvation, either immediately or in a future existence. His teaching was available in the language of the people. As the teaching became codified and its language became ossified, and as the vernacular languages continued to develop, a gap arose between the language of the teachings and the language of the people. Already by the second century B.C.E. we can see that there was a considerable difference between Pāli and Sinhalese Prakrit, as it appears in inscriptions, and a century or two later the difference was so great that, apart from Pāli loan words in Sinhalese Prakrit, one would be hard pressed to see that the two Middle Indo-Aryan dialects were related. If, therefore, the Buddhavacana was to be propagated in a form which the common people could understand, then bhikkhus would have to translate the texts which they had memorised and recited in Pāli into the vernacular languages of the people. Ultimately, as we know, the Pāli canon had to be translated into Sinhalese (Prakrit). Something similar must have been the experience of all sects and schools of Buddhism.

Although the language of the Gāndhārī Dharmapada was archaic, and approximated to the spoken language of two centuries before, nevertheless we may assume that it was still intelligible, for the most part, to the people living in the Gandhāra area and along the southern Silk Road in Chinese Turkestan where it was found. Texts written in this language, or something very similar, were, however, then translated into a Prakritised Sanskrit and then into Buddhist Sanskrit by the monks living in the Turfan area on the northern Silk Road. We may wonder whether anyone, other than those monks, could read the scriptures in that form.

If the teachings were no longer intelligible to the masses, then it did not matter much which language they were in, and it is clear that Sanskrit, the language of culture and literature, was an obvious choice. It was the literary language of North India, at least, and translation into that language meant that educated people - mainly monks, but probably some laymen as well - could read their own scriptures and also those of any other school, if they were so minded. The result was, however, not hard to predict. Buddhism started to become an academic study, where only the educated, who had learned Sanskrit, had access to the teachings in their written form. Bhikkhus, studying in their vihāras, became more and more remote from the people. We know of only one Buddhist work devoted to the life and duties of a layman, the Upāsakajanālanikāra. When Buddhism came under attack from the iconoclastic Muslim invaders, and the vihāras and their libraries were destroyed, Buddhism in the land of its founder virtually disappeared.

There is a very obvious contrast with Jainism. There the monks and the laymen were more closely integrated. Texts dealing with the duties of laymen are numerous. Many translations of canonical texts were made, but into vernacular languages, not Sanskrit, and there were also commentaries written in vernacular languages. Although the Jains did use a form of Sanskrit, which resembled the Sanskrit used by Buddhists in its dependence upon Middle IndoAryan forms, Sanskrit never gained the importance for Jains which it had for Buddhists. Although both Buddhism and Jainism had royal patrons, Jainism also had close links with, and was less remote from, the people, and we find, for example, that the Jains made great use of popular literature, with Jain versions of the Mahābhārata and the Rāmāyana. It has been argued that it was as a consequence of its support among the common people that Jainism, although hit no less than Buddhism by the invaders, survived in North India. Buddhism did not.

VII - Buddhism and Aśoka

While re-reading recently a book on Buddhism by an eminent Buddhist scholar, I noticed the following statement: “This notion of establishing the sāsana or Buddhism in a particular country or a place was perhaps first conceived by Asoka himself. He was the first king to adopt Buddhism as a state religion, and to start a great spiritual conquest which was called dharma-vijaya … . Like a conqueror and ruler who would establish governments in countries politically conquered by him, so Asoka probably thought of establishing the sāsana in countries spiritually conquered by him”. ²⁶⁴

In other publications I have seen such claims made as: “Aśoka was the first Buddhist Emperor”, “Aśoka was connected with the popularisation of Buddhism, and with the enthusiastic promotion of religious activities such as pilgrimage and the veneration of relics through his involvement in the construction of stūpas and shrines”, ²⁶⁵ and “Aśoka was the greatest political and spiritual figure of ancient India”. ²⁶⁶

Such comments are typical of the way in which Aśoka is described in books about early Buddhism. As I stated in the first lecture, I have spent a large portion of my academic life studying Aśoka’s inscriptions, and I do not find that the picture of the man which emerges from his edicts coincides entirely with what we find written about him. So in this lecture I want to consider the part which Aśoka played in the history of Buddhism, and I shall compare what we learn about him from Buddhist texts with the information which we can get from his own inscriptions.

It is probable that most people know about Aśoka from the information given about him in the Pāli chronicles, and in particular the Mahāvamsa, although many of the same stories are told in greater detail in Sanskrit and Chinese sources.

In the Mahāvamsa we read how, after the death of his father Bindusāra, Aśoka killed 99 of his 100 brothers, sparing only Tissa, and became the sole ruler of Jambudvīpa. ²⁶⁷ After hearing the Buddhist novice Nigrodha preach, he was established in the three refuges and the five precepts of duty, i.e. he became a Buddhist layman (upāsaka). ²⁶⁸ We are told of Aśoka being a firm supporter of Buddhism, to the exclusion of other religions. He stopped giving food every day to the 60,000 brahmans whom his father had fed, and in their place gave food to 60,000 bhikkhus. He gave orders for 84,000 vihāras to be built in 84,000 towns, ²⁶⁹ of which the most famous was the one which he himself founded in Pātaliputra - the Asokārāma - and he built stūpas in the places which the Buddha had visited. ²⁷⁰ He was persuaded, by the thought of becoming an heir of the doctrine, ²⁷¹ to allow his son Mahinda and his daughter Samghamittā to join the order in the sixth year of his reign, ²⁷² and Mahinda was subsequently sent as a missionary to Sri Lanka. At the time of the schism in the order he personally listened to the bhikkhus expounding their views and was able to decide who were orthodox and who were heretics. After the schism had been settled, the third sangiti was held under his patronage. ²⁷³

The Mahāvaṃsa ²⁷⁴ says that because of his wicked deeds, he was known as Candāssoka “fierce, or violent Aśoka” in his early days, but later because of his pious deeds he was called Dhammāsoka. The change of name is, of course, intended to emphasise the difference between Aśoka as a non-Buddhist and as a Buddhist.

We get rather different information about Aśoka from reading his inscriptions.

For example, the story of his conversion in the chronicles is somewhat at variance with his own statements. The early history of Aśoka’s involvement with Buddhism is told in the first Minor Rock Edict. Whereas, according to the Pāli sources, as I have just stated, he had already been converted to Buddhism, had 84,000 vihāras built, and had given permission for his son and daughter to join the order within six years of his consecration as king, nevertheless, we can calculate from Aśoka’s own words that his conversion to Buddhism occurred fairly soon after the war in Kalinga, which he states in the thirteenth Rock Edict [RE XIII(A)] took place when he had been consecrated eight years. His conversion was presumably because of his remorse, not for his fratricide which seems to be disproved by the references which he makes to his brothers and sisters in the fifth Rock Edict [RE V(M)], but for the transportation of 150,000 persons, the killing of 100,000 , and the death of almost that many, in Kalinga. The chronicles show no knowledge whatsoever of this carnage. At the time at which he promulgated the edict, he had been a layman for more than two and a half years, including a year when he was not very zealous - I assume that this means that after he had been converted to Buddhism, he had not been a very energetic Buddhist for a while - and then more than one year when he was zealous, after he had “approached the sangha” (which, perhaps, means that he went on a refresher course) with good results - “I have made good progress”, he says. When he issued the first Minor Rock Edict he was, therefore, at a point just short of the eleventh year of his reign. He issued the third Rock Edict in the twelfth year, so the first and second Rock Edicts, which are not dated, were issued either in the same year, or in his eleventh year.

As the story is related in the first Minor Rock Edict, however, there is no direct evidence that it was the Buddhist sangha he went to. Although in one version of the edict (the one at Maski) he is described, as a Budhaśake, which Hultzsch translates as “(I am) a Buddha-Śākya”, it seems fairly certain that the insertion of the word Budh(a) was done by a single local scribe to “correct” the word upāsake which he had already written. It is noteworthy that Budhaśake or its equivalent does not occur in any other version of the first Minor Rock Edict. All the other versions have the word upāsaka. ²⁷⁵

It is probable that the scribe, realising that, as I have just said, there is no indication of the sect in which Aśoka was an upāsaka, and wishing to make the situation clear to all readers of the edict, tried to insert the word “Buddha” before the word upāsaka, which he had just carved on the rock, with only partial success.

Nevertheless, confirmation that Aśoka had become a Buddhist is provided by the reference to his visit to the bodhi tree, which is described in the eighth Rock Edict [RE VIII(C)] as happening when Aśoka had been consecrated ten years. This must have been one of the first consequences of his conversion to Buddhism, and it perhaps coincided with his visit to the sangha which improved the quality of his religious life.

We can, in fact, be certain that, for Aśoka, sangha means the Buddhist sangha, because in the seventh Pillar Edict [PE 7(Z)], when he summarises all his achievements, he states that he has set up mahāmātras “ministers” of morality to look after the affairs of the sangha, the brāhmanas, the Ājivikas, the Jains, and various other religious sects. In the context, with the other sects specified by name, the sangha, by a process of elimination, must be the Buddhist sangha.

There is, then, no doubt that Aśoka was a Buddhist layman, but there is no reason to believe that his mind was closed to other religions, and we read in the sixth Pillar Edict [PE 6(E-F)] that he had honoured all sects with various forms of honour, and the best of these, in his opinion, was a personal visit to them. As we shall see, his sort of dhamma was moral and ethical, as opposed to spiritual, so that he could equally well have been a Jain layman.

The edicts give a great deal of information about the way in which Aśoka propagated his own dhamma. He writes, among other things, of dhammathambhas “dhamma pillars”, dhammalipi “dhamma writings”, dhammamamgalas “dhamma ceremonies”, dhammadāna “dhamma giving”, dhammānuggaha “dhamma benefit”, dhammayātrās “dhamma journeys”, dhammasavana “hearing the dhamma”, dhammamahāmātras “dhamma ministers”, dhammavijaya “dhamma victory”.

The problem is to know if Aśoka’s dhamma was the same as the Buddhadhamma. He makes it clear that there is a difference between ordinary practices and institutions, and the dhamma version of them. A pillar is a thambha. It becomes a dhamma-thambha if Aśoka’s dhamma is carved on it. There were mahāmātras before Aśoka’s time. He was the first to institute dhammamahāmātras to propagate his dhamma. Before his time kings went on yātrās. He instituted dhamma-yātrās, so that he could practise his dhamma while on journeys. People performed all sorts of ceremonies (mangalas) - in case of illness, and at weddings, to get children, before going on journeys, etc. The dhammamangala, however, is the proper treatment of slaves and servants, honouring teachers, self-restraint with regard to living creatures, generosity to śramanas and brāhmanas, etc. [RE IX(G); cf. RE XI(C)].

Aśoka’s dhamma is set out clearly in several inscriptions, e.g. in a concise form in the second Minor Rock Edict: “Obey one’s parents; obey one’s elders; be kind to living creatures; tell the truth”. All this is said to be in accordance with ancient usage (porānā pakatt) - a third century B.C.E. version of “back to basics”. Elsewhere, in the third Rock Edict, a slightly expanded version of this is given: “Obedience to mother and father is good; liberality to friends, acquaintances and relatives, to brāhmanas and śramanas is good; abstention from killing animals is good; moderation in expenditure and moderation in possessions are good” [RE III(D)].

The series of seven edicts on pillars, which we call the Pillar Edicts, is devoted to an explanation of Aśoka’s dhamma, with an account of how he himself has complied with it, by planting trees for shade by the road-side and digging wells and building watering-places for men and animals. Pillar Edict 1 tells of government by dhamma. Pillar Edict 2 states that dhamma consists of doing little sin, doing much good, showing compassion, making donations, telling the truth, and purity. Aśoka has done much good by not killing. Pillar Edict 3 tells of good and evil, and identified the latter as fierceness, cruelty, anger, pride, and envy. Pillar Edict 4 emphasises the need for equality of justice and the rehabilitation of prisoners. Pillar Edict 5 prohibits the killing of a number of animals which are specified by name. Pillar Edict 6 states that the aim is to bring happiness to all. All sects are to be honoured, especially by personal visits. Pillar Edict 7 seems to be a summary of all that Aśoka has done. He explains how kings in the past had sought to increase dhamma. Aśoka had decided to do it by preaching and instruction, and had instituted dhamma-pillars (dhammathambhas) and dhamma-ministers (dhammamahāmātras) to put this decision into effect. The dhammamahāmātras were concerned with all sects. Dhamma is defined again as: obedience to parents, obedience to teachers, respect to the old, and proper behaviour towards brāhmanas and śramanas, to the poor and to slaves and servants. There had been an increase of dhamma as a result of Aśoka’s legislation, e.g. about killing animals, but also because of an attitude of mind, i.e. personal conscience (nijhati). In this way the next world is gained.

Elsewhere, in the series of major Rock Edicts, we read that one must obey the dhamma and conform to it [RE X(A)]. The gift of the dhamma is defined as the proper treatment of slaves, obedience to parents, etc., generosity to brāhmanas and śramanas, and non-killing. The dhamma gives endless merit [RE XI(E)].

Aśoka calls his edicts dhamma-writings (dhammalipis), and we can read them and see exactly what dhamma each lipi contains. We can see that his dhamma is exclusively a moral one, which is why we often translate dhammalipi as “rescript on morality”. Aśoka promoted his dhamma widely, and instituted dhammamahāmātras to supervise it, dhammathambhas to carve it on, and had messengers (dūtas) to carry it all over India and even to the Greek kings to the West. Except in so far as the moral ideas are quite in conformity with Buddhist moral teachings, there is no hint of anything exclusively Buddhist in them, and in the insistence on non-killing (ahimsā) his thought closely resembles the Jain emphasis on this, and in fact parallels have been noted between the lists of animals declared inviolable in the fifth Pillar Edict and lists of animals in Jain texts. ²⁷⁶

In the Bairāt edict we find that, under his personal name of Priyadassi, Aśoka greets the Samgha, and wishes it well. He goes on to say that it is known how great is his faith in the Buddha, the Dhamma, and the Samgha. It is obvious that in this context Dhamma has its usual Buddhist meaning as one element of the Triratna, but other than this one reference, it is, in fact, very clear that Aśoka’s references to dhamma do not refer to the Buddha’s dhamma, and Aśoka’s dhamma was not the same as the Buddha’s dhamma. ²⁷⁷ The quotation I gave at the beginning of this lecture about establishing the sāsana in countries which he had spiritually conquered seems, therefore, to be based upon a misunderstanding of the nature of Aśoka’s dhamma.

That same quotation also seems to be wrong when it talks about countries being spiritually conquered by Aśoka, at least if, by spiritual conquest, it is referring to Aśoka’s dhamma-vijaya. As I understand the situation, Aśoka expanded his empire by force, but thereafter devised the principle of victory by morality, and commended it to his successors.

It seems to me that the guiding principle of Aśoka’s teaching was non-killing (ahimsā), presumably as a result of his remorse over the killing in Kalinga. His very first Rock Edict is concerned almost entirely with the prohibition of killing in daily life, including the killing of animals for sacrifice and for food.

In the thirteenth Rock Edict he tells the story of how, after the victory in Kalinga, with all the death and suffering this involved, he had a great desire for morality (dhamma) [RE XIII(C)]. It pained Aśoka that those who did obey his dhamma, including brāhmanas and śramanas (J), suffered nevertheless (G). His hope for the forest dwellers is that they too (like him) may repent (of past killing ?) and not kill (in the future) (M-N). What he now wished for was a dhamma-victory, and this consisted of security for all creatures, self-restraint, equanimity and gentleness (O). Messengers were sent to preach this everywhere, including the Greek kingdoms to the West (Q). His conquest by morality (dhamma) is promulgated in the hope that his successors will not think of another (military) victory, by force of arms which would entail slaughter, similar to that in Kalinga, and that in their own victory there will be mercy (khanti) and light punishment (lahudandatā) (X).

And so it seems to me quite certain that the messengers (dūtas) who were sent to the Greek kings were not charged with the propagation of Buddhism. It would seem clear that they were sent in an attempt to persuade the rulers, probably despotic rulers, of the neighbouring states that they too should give up their desire for conquest by war, and should try to institute the reign of peace and tranquillity, based upon the principles of Aśoka’s dhamma. In these circumstances, to talk, as some do, about the Aśokan missionary expansion of Buddhism among the Greeks, seems to me to be a mistake. Certainly we have no evidence from the Greek side which indicates that any Buddhist missionaries had arrived among them c. 250 B.C.E.

It will be clear from what I have said so far about Aśoka’s dhamma that those who talk of him making Buddhism the state religion are very wide of the mark.

In his edicts, Aśoka says little or nothing about Buddhism. There is no reference to any of the basic tenets of Buddhism, e.g. samsāra, mokkha, nibbāna, anattā, the eight-fold path or the four Noble Truths. In the Separate Edicts he stated that his aim was the happiness of all (SepE I), and a number of inscriptions include the statement that his aim was that his people may attain happiness in this world, and heaven in the other world. His boast was that he had mixed men with gods, a statement which has been variously interpreted. I take it to mean that he had succeeded in bringing men to heaven, where of course they will be reborn as gods, i.e. mixed with other gods. This seems to me to be far from the idea of an endless series of rebirths which we normally associate with Buddhism.

His general failure to mention Buddhism has been variously explained. It was perhaps due to ignorance, i.e. although he was nominally an upāsaka, he had very little knowledge about Buddhist doctrines. Alternatively, perhaps he knew about Buddhism, but he thought it would be favouring one sect unduly, and thus destroying the impartiality which he aimed to show elsewhere, if he referred to it in detail. It is also possible that he thought that it was irrelevant to his purpose in publishing his edicts, namely to spread knowledge of his own personal dhamma, which was intended to bring peace to his empire and enable all his subjects to live in harmony with each other.

One of the more bizarre explanations I have come across is the view that what Aśoka conveys in his edicts is the state in which Buddhism was in his time. Hultzsch says: “Aśoka’s dharma is in thorough agreement with the picture of Buddhist morality which is preserved in the … Dhammapada. Here we find Buddhism in statu nascendi” - perhaps, ‘in its infancy’.²⁷⁸ He goes on to say, “In one important point Aśoka’s inscriptions differ from, and reflect an earlier stage in the development of Buddhist theology or metaphysics than, the Dhammapada: they do not yet know anything of the doctrine of Nirvāna, but presuppose the general Hindu belief that the rewards of the practice of dhamma are happiness in this world and merit in the next world”. ²⁷⁹ Hultzsch’s statement raises an interesting question about the nature of Buddhist theology. Is it possible that the doctrine of nirvāna is a later stage in its development? I cannot believe that that is so, and I therefore think that Hultzsch’s reason for Aśoka’s failure to mention nirvāna cannot be correct.

On the face of it, the situation appears to be very similar to that found in Sri Lanka and described by Richard Gombrich. ²⁸⁰ He reports, “But most Sinhalese villagers do not want nirvāna … . They say that they want to be born in heaven.” They do, at least, know about nirvāna. Aśoka gives no hint of ever having heard about it. Although the villagers’ statement may seem strange, it is in fact quite in conformity with the statement which we find at the end of the Alagaddūpamasutta of the Majjhima-nikāya, ²⁸¹ where the Buddha states that those bhikkhus who act in accordance with the dhamma and with faith, will gain awakening (sambodhi), while those who merely have faith in, and love for, the Buddha will attain heaven. The villagers know about nibbāna, but prefer the heaven which they will gain because of their love of the Buddha. Aśoka makes no mention of loving the Buddha. For him, heaven is gained by doing the things which he specifies in his dhamma. The statement is not even put in the context of acquiring good karma. He makes it clear that this is, in itself, the summum bonum: “What is more important than gaining heaven ?” he asks in the ninth Rock Edict [RE IX(L)].

The strongest indication of his connection with Buddhism is the edict at Bairāt, which I have already mentioned, and we should note that the one place where he actually refers to the Buddha’s teaching is in this edict addressed to the sangha. In it he says that everything said by the Buddha was well said, and he commends seven texts by name to the sangha. He refers to the Buddha’s teaching as the saddhamma, which perhaps was an intentional action to distinguish the Buddha’s dhamma from his own. We must hope that Aśoka was preaching to the converted. If his exhortation had been intended for the common people, it would presumably have been in a Rock Edict. There is the problem that we cannot be certain of the identity of some of the texts, but since we sometimes find that the commentaries, e.g. that on the Sutta-nipāta, know of some texts under other names, ²⁸² it is perhaps not altogether surprising that we cannot recognise all of Aśoka’s choices.

In his statement in the eighth Rock Edict [RE VIII(C)] that dhamma journeys (dhammayātrās) have replaced the pleasure trips that kings used to take, Aśoka says that he went to the bodhi tree, but he says nothing about the need for others to go on pilgrimage to sacred places, or about pilgrimage as a religious activity. As a result of his visit, he defines a dhammayātrā as an opportunity to travel around in order to put his dhamma into effect. He says precisely what it entails: giving audiences and making distributions to śramaṇas and brāhmaṇas, giving audiences and distributing money to old people, giving audiences to the country people and preaching to them and answering questions about his dhamma.

What we read of Aśoka’s own visits to sacred places does not encourage us to think that he thought that they were very important. It is interesting to note that of the four places which are sacred to Buddhists - where the Buddha was born, achieved sambodhi, gave his first sermon and died - we have only first-hand evidence of Aśoka’s visit to two of them. Although the Aśokāvadāna tells of Aśoka being taken on a conducted tour of these four places and others, Aśoka himself states that he went to the bodhi-tree in his 10th year, and to the Buddha’s birth-place at Lumbinī, where he had a pillar erected, ten years later, in his 20th year. To some extent, of course, this is an argument from silence. Perhaps Aśoka did visit the other places, as the texts say, and did raise pillars there. If so, they are now lost. The discrepancy in the dates, however, is less easily explained.

We should note that the visit to sambodhi has been interpreted by some as meaning that Aśoka was so proficient a practitioner of Buddhism that he actually gained bodhi himself. Such an interpretation would imply that meditation, bodhi and nirvāna were, in fact, known to Aśoka, even though there is no trace of that knowledge in the edicts. Even as recent a writer as Jules Bloch toyed with this idea, but then decided that “go to sambodhi” is philologically unlikely as a way of saying “gained awakening, became a Buddha”, and he concluded that it means “visited the site of the bodhi tree”. 20

We have an inscription which tells us that in his 14th year Aśoka enlarged the stūpa of the previous Buddha Konākamana. It is not clear who built it. The inscription has been interpreted as meaning that Aśoka enlarged it for the second time, but it is perhaps more likely that it means “enlarged it to twice its former size”. We have no more information from the edicts about Aśoka erecting or enlarging stūpas, although the Chinese pilgrims record the existence of a number of stūpas which were attributed, in their day, to Aśoka, including one built at the place where the Mahāsānghikas held their assembly. ²⁸³

Both the Northern and the later Southern Buddhist sources give the information that Aśoka broke into caityas which he thought might hold relics of the Buddha, and when he eventually found one with relics in it he re-interred them in 84,000 caityas in the 84,000 vihāras which he had had built. The words budhasa salīle “relics of the Buddha” which occur at the end of the version of the first Minor Rock Edict found at Ahraurā in 1961, might be thought to refer to this redistribution of the relics, ²⁸⁴ but since the words are found in only one of the seventeen versions of that edict so far discovered, it seems more likely that the words are an invention of the scribe at Ahraurā, based upon a misunderstanding of something which he found in the version of the edict sent to him. ²⁸⁵

Aśoka devotes the whole of the twelfth Rock Edict to making it clear that he is equally concerned with adherents of all religions, and he honours them all with gifts and other sorts of honours. All sects must listen to each others’ dhamma, so that there may be an increase of sālā (which I take to mean “communication”) between them [RE XII(I)]. Then there will be an increase in each individual sect and an illumination of dhamma [RE XII(N)]. Aśoka wishes them all to live in harmony together, without self-aggrandizement or disparagement of other sects.

Aśoka seems to use the compounds brāhmaṇa-śramana and śramaṇabrāhmaṇa (in their various Prakrit forms) to mean all members of religious orders, orthodox and heterodox. In most cases he puts the word śramana- first, but in two places he reverses the order of the words, and in the fourth Rock Edict he has both forms of the compound. The scribes at some sites change brāhmaṇaśramana to śramana-brāhmana, perhaps thinking that to have brāhmana first was wrong, or perhaps believing that they ought to correct what appeared to be Aśoka’s inconsistency. In the thirteenth Rock Edict he writes of brāhmanas and śramanas and members of other sects (pāsamdaā ). He would surely not have referred to the sects in this way if he had in fact rejected the brāhmanas completely, as the Mahāvamsa story suggests. The variations of this word order are probably due to the regional scribes, who (depending on their personal feelings) put one or the other first. This applies especially to the scribe at Girnār, who always prefers to have brāhmana- first. This may be connected with the Sanskritisations we find at Girnār, and suggests that the scribe was perhaps a brahman. In Pāli texts, it seems to be conventional to have the words in the order samana-brāhmana, which is not surprising. In the seventh Pillar Edict [PE 7(HH)], where we have only one version of the edict, we find the compound in the order bābhana-samana.

His encouragement of all sects must mean that he did not stop feeding brāhmanas, and, as I have said, his dhamma in fact specifically includes giving to śramanas and brāhmanas. His donation of caves to the Ājīvikas in his 12th year[^7-24] is additional evidence that he was not devoted exclusively to Buddhism.

The item which most merits our attention, however, is the set of three versions of the so-called Schism Edict, because these have been interpreted not only as showing that Aśoka was sufficiently involved in the affairs of the sangha to be able to settle the schism in the order (sangha-bheda), as the Mahāvamsa says, but also to intervene to the extent of removing dissenting heretics by force. The question is how far the Schism Edict reflects the actual events of the schism which the chronicles tell us occurred in the time of Aśoka and led to the holding of the third sangiti.

It is not always realised that there are five accounts of the third sangiti and the events leading up to it given in the early Pāli chronicles and commentaries: there are two in the Dipavamsa, two by Buddhaghosa in his commentaries upon the Vinaya (Sp) and the Kathāvatthu (Kv-a), and one in the Mahāvamsa. The accounts they give are not identical, but differ in various details.

Most people know the version told in the Mahāvamsa, which in fact is the latest and most developed version of the story. It states (Mhv 5.229-70) that the heretics who had lost honour, when Aśoka started feeding 60,000 bhikkhus, put on the yellow robe and joined the bhikkhus. They went on proclaiming their own doctrines and performing their old practices. The bhikkhus could not restrain them, and for seven years the bhikkhus in Jambudīpa did not hold an uposatha ceremony or the ceremony of pavāranā in all the ārāmas. When Aśoka tried to make the bhikkhus in the Asokārāmavihāra perform the uposatha, his minister killed several bhikkhus. The king received a week’s instruction in the Buddha’s teaching, and then listened to all the bhikkhus’ doctrines, and caused all the adherents of false doctrines to be expelled from the Order. They numbered 60,000 (Mhv 5.270). The Order, now in harmony (samagga), ²⁸⁶ assembled and performed the uposatha. They then held the third sangiti at which Moggaliputta Tissa recited the Kathāvatthu. The end of the sangiti is dated to the 17th year after Aśoka’s consecration.

The two versions by Buddhaghosa are earlier than the Mahāvamsa. They are very similar to each other, and also to the Mahāvamsa version, with the detail, omitted in the Mahāvamsa, that Aśoka gave the heretics white robes when he expelled them from the Order. The heretics’ practices are said to include the tending of the sacrificial fire, from which we can deduce that some of them were brāhmanas.

The versions in the Dipavamsa are earlier than Buddhaghosa. We should note that the Dipavamsa is a strangely undisciplined text. It obviously represents a conglomeration of source material bundled together uncritically, so that there are often two versions of the same event, and sometimes three. There are, in fact, two versions of the schism story:

(I) The first version ²⁸⁷ states that the schismatics and heretics, among whom the Jains and Ājīvikas are specifically mentioned, had lost gain and honour, and consequently infiltrated the Order. For seven years the uposatha ceremony was carried out by incomplete groups (vagguposatha), ²⁸⁸ since the noble ones did not attend the ceremonies. By the time 236 years had passed since the death of the Buddha, 60,000 bhikkhus lived in the Asokārāma. The various sectarians ruined the doctrine, wearing yellow robes. Moggaliputta convened a recitation, and having destroyed the different doctrines and expelled the shameless intruders, he recited the text known as the Kathāvatthu.

(2) The second version ²⁸⁹ says there was a dreadful schism (bheda) among the Theravādins 236 years after the death of the Buddha. The heretics (numbering 60,000 ), seeing the honour being given to the sangha, furtively attached themselves to it. The Pātimokkha ceremonies in the Asokārāmavihāra were interrupted. A minister, who ordered the Pātimokkha ceremony to be performed, killed some of the bhikkhus, which led to the king consulting the elders about the killings. Moggaliputta presided over a gathering of 60,000 Buddhists, which had assembled to destroy the sectarians. Aśoka learned the doctrine from the thera, and is said ²⁹⁰ to have destroyed the (bhikkhu)-emblems of the intruders (rājā … theyyasamvāzabhikkhuno ²⁹¹ … nāseti linganāsanam). The heretics, performing the pabbajjā rite according to their own doctrine, damaged the Buddha’s utterances. To annihilate them Moggaliputta recited the Kathāvatthu. After that recitation he held the Third Recitation.

If we examine all these versions, we can probably trace the way in which additions were made to the basic version of the story. It is likely that the first account in the Dipavamsa is the earliest version. It dates the occurrence, and states that sectarians whose honour and gain had been reduced because of the growing prestige of the Buddhist Order infiltrated the order and wore the yellow robe. For seven years the true Buddhists would not perform the uposatha in their presence. Moggaliputta destroyed the various doctrines and removed the shameless ones. There is no mention of Aśoka, nor of the giving of white robes. The second version in Dipavamsa adds the statement that there was a schism (bheda) in the Theravāda. It does not specifically mention the uposatha, but states that the Pātimokkha ceremony in the Asokārāmavihāra was interrupted, although it does not say for how long. A minister tried to settle the matter, but his intervention caused bloodshed. The king asked about the bloodshed, received religious instruction, and destroyed the sectarians’ (bhikkhu)-emblems.

Buddhaghosa introduces the story of Aśoka becoming so involved in the matter that he sends a minister, rather than the minister acting on his own responsibility. That minister tries to settle the matter by forcing the bhikkhus to perform the uposatha, and killed a number of them in the process. After a week’s training in the doctrine, Aśoka was able to discern that the intruders had heretical views, and he consequently made them wear white robes and expelled them from the Order. The Order is then said to be in harmony (samagga). The Mahāvamsa version adds the detail that no uposatha ceremony was held in Jambudīpa for seven years, nor the pavāranā ceremony in all the ārāmas.

We can probably reconstruct the account of the matter in the following way. Sectarians (probably those who had fallen out of favour when Aśoka began to show a preference for Buddhism) infiltrated the Asokārāma, and the true bhikkhus refused to celebrate the uposatha ceremony, while they were there. There was therefore bheda in the Asokārāma sangha. It has been suggested that this sanghabheda must have been a very serious event, which carried a heavier penalty than that laid down for sanghabheda in the Vinaya-pitaka, ²⁹² namely, expulsion from the order, which is what wearing the householder’s white robes implies. I would suggest, however, that it was not a question of bhikkhus being forced to wear the householder’s white robes, but of infiltrators being forced to give up the emblems to which they were not entitled, and being made to depart from the vihāra, where they had no right to be. The Vinaya penalties would not be appropriate for those who were not genuine bhikkhus.

I see no reason to believe that Aśoka himself carried out the expulsion. The earlier version in the Dipavamsa states that Moggaliputta removed the heretics, and makes no mention of Aśoka. It is, however, not unlikely that, as the chronicles say that the bhikkhus were unable to restrain the sectarians by the rules of discipline, Moggaliputta was unable to enforce the order of expulsion from the vihāra. In this case, recourse to the civil power was perhaps inevitable, and a minister had to deal with the matter. This action would not be a case of one of the king’s ministers intruding into a religious matter, since those to be evicted were not true bhikkhus.

There seems to be no reason to doubt that this part of the story is historically true. The Mahāvamsa version, however, has Aśoka himself becoming involved, doubtless because it was “his” arāma. According to this version, he personally sent his minister, and became further involved after the bloodshed which was caused. Aśoka’s commitment to the Theravādin cause is emphasised by the story that he personally decided who held the heretical views, and expelled them from the Order. The story is, however, given a slightly unreal element by the insertion of a detail whereby Aśoka, after recognising the heresies of the dissidents, and the correct views of the orthodox bhikkhus, then asked the thera Moggaliputta Tissa what the Buddha actually taught (Mhv 5.271). When the sectarians had been removed, the sangha in the Asokārāmavihāra became samagga “united, in harmony”. The final expansion of the story adds the detail that no uposatha ceremony was held in Jambudīpa for seven years, nor any pavāranā ceremony in all the arāmas. These additional details presumably represent an attempt to make the matter appear far more widespread than it really was.

The shortest version of the three versions of Aśoka’s Schism Edict, which we may assume gives the gist of the edict, states that the sangha had been made united (samagga), and that any monks and nuns, who caused schism in the future, should be made to live outside the dwelling (avāsa), i.e. the vihāra, and to wear white robes. There is no information about where the sangha had been made samagga or by whom, and the order to remove schismatics refers to the future and does not say that any had already been removed. Consequently, it is by no means obvious that Aśoka’s edict and the story in the chronicles refer to the same event. Nevertheless, as we have seen, the removal of the (bhikkhu)emblems or the wearing of white robes, the expulsion and the sangha being made united (samagga) are mentioned in some of the Pāli accounts. I believe that it is too much of a coincidence for there to be no connection whatsoever between the edict and the Pāli accounts. I conclude, then, that the references in the Pāli texts must go back to a very early tradition, brought from India and preserved in the Mahāvihāra, that Aśoka did, or at least wrote of doing, these things. It is interesting to note that the references to white robes and the sangha being samagga do not occur before Buddhaghosa’s account of the matter, which implies either that these details were not available to the author of the Dipavamsa, perhaps because they did not yet exist, or else that he chose to omit them for some reason.

It is obvious that some of the statements made about Aśoka by modern writers can be verified by reference to his inscriptions. ²⁹³ Aśoka was a Buddhist and he was emperor of India, or at least the sole ruler of a large proportion of it. We know of no Buddhist ruler of this, or any other comparable, territory before Aśoka, so it is not incorrect to call him the first Buddhist Emperor. On the other hand, some of the statements made about him seem to be rather extravagant, and not capable of being verified. A case can perhaps be made for saying that “Aśoka was the greatest political figure of ancient India”, but, in the absence of any first-hand information about his spirituality, it seems unjustified to say that he was also the greatest spiritual figure. Aśoka’s own words about the function of dhamma journeys (dhammayātrās) seem to make his alleged enthusiastic promotion of religious activities such as pilgrimage and the veneration of relics less than certain.

There are three questions to answer. The first question is: why did the Buddhists claim that Aśoka was exclusively pro-Buddhist? The second question is: why did the Theravādins claim that he favoured them at the time of the schism? The third question is: why do modern writers make claims about him which are not supported by his own words? The last is easily answered. Modern writers say what they do because they have either read only the Buddhist sources or been misled by other modern writers. In either case they have not actually read the Aśokan inscriptions themselves. Those who have heard that Aśoka recommended certain suttas, i.e. portions of the Buddhist dhamma, and know that Aśoka set up dhamma-writings and sent out messengers, have put the two pieces of information together, and have assumed that Aśoka set up, i.e. popularised, Buddhist teachings and sent out Buddhist missionaries. This incorrect view may owe something to the editor of the Dictionary of Pāli Proper Names, ²⁹⁴ who had perhaps confused Aśoka’s messengers with Moggaliputta’s missionaries, and actually states that Aśoka sent out the missionaries after the third sangiti, although the texts clearly state that they were sent out by Moggaliputta.

It is clear that the Buddhists appropriated Aśoka for their own use. It is possible that the earlier version of the story in the Dipavamsa goes back to a very early form of the tradition, when there was no need to “invent” Aśoka’s involvement, whereas the second Dipavamsa story was formulated later, when the political situation had changed. There was undoubtedly rivalry with the brahmanical caste, when the power of the brāhmanas grew again after Aśoka’s death, as Pusyamitra supported a brahmanical reaction. Samprati, Aśoka’s grandson, is said in Jain sources to have been a great supporter of Jainism and, in face of the royal support for both those religions, it was perhaps inevitable that the Buddhists, building upon the undoubtedly correct fact that he was a Buddhist layman, would maintain that the great king Aśoka favoured only the Buddhists, to the exclusion of other sects, and in fact had 18,000 Jains put to death in a single day, according to the Aśokāvadāna. ²⁹⁵ In later times the story was embroidered even more, and I-tsing reports that an image of the Buddha was dressed in a monk’s robes of a particular pattern, ²⁹⁶ implying that Aśoka was more than just a layman.

It is probable that it was only after the schism in the Buddhist sangha, and indeed because of the schism, when the inclusion of the word mahā - in the name Mahāsānghika seemed to imply that the Theravādins were only a minority sect of Buddhism, that it became necessary to prove that the Theravādin view of the Buddha’s teaching was the correct one. The Theravādin school consequently began to make statements about Aśoka himself favouring their view of the Buddha’s teachings, in order to legitimise their claims to be the true exponents of the Buddhist tradition.

It has been said that Aśoka’s patronage was responsible for establishing Buddhism over a far wider area than could have been imagined before the founding of the Mauryan empire. ²⁹⁷ We must note that we can find no evidence in the edicts that there was any greater patronage of Buddhism than of any other sect. It is probably pure chance that we have little or no information about the patronage which Aśoka bestowed upon other sects, although I have mentioned the caves which he gave to the Ājīvikas. It is also probable that some of the things which the Buddhist texts claim Aśoka did were not done by him personally, but by the mahāmātras appointed to look after the affairs of the sangha. There is no reference to Asoka quelling schism in the order (sañghabheda) in the seventh Pillar Edict, which is dated to Aśoka’s 27th year, and seems to be a summary of Aśoka’s activities as a ruler, nor is there any mention there of the third sangiti. The fact that Aśoka says nothing about the sangiti being held under his patronage suggests that he did not include among his achievements any detailed information about the mahāmātras’ activities. It would have been the mahāmātras who made whatever arrangements were needed to ensure that the sangiti could be held. Since the Buddhists recognised that they did so on behalf of the king, they were able to claim that he was their patron.

Similarly, the instructions in the covering letter, which is attached to the version of the so-called Schism Edict at Sarnath, about mahāmātras coming on every uposatha day to read and understand the edict, probably refer to the mahāmātras whose duty it was to look after the sangha. We may not be too wide of the mark if we also assume that it was not Aśoka himself who officiated at the voting, recorded in Chinese sources, ²⁹⁸ after the dispute which resulted in the arising of the Mahāsānghikas, but the mahāmātras, acting on Aśoka’s behalf.

If it is true that Buddhism expanded during the reign of Aśoka, then it seems to me that, rather than this being the result of his patronage, or to his deliberate attempts to propagate it (for which, as we have seen, there is little or no direct evidence), it is more likely that it was a result of the peace which he established, leading to greater prosperity and the expansion of trade. As I said in the second lecture, Buddhism followed the trade routes. It was undoubtedly those same trade routes which were followed by Aśoka’s emissary dhamma-mahāmātras, who, as we read in the fifth Rock Edict, were instituted in Aśoka’s 13th year to look after members of all sects, and to spread Aśoka’s dhamma through all parts of his territories, and we may be sure that the religious missionaries sent out by Moggaliputta followed in the footsteps of the mahāmātras. ²⁹⁹

I mentioned the doubt about the identity of some of the seven texts which Aśoka recommended to the sangha. Beside that doubt, there is also a dispute about the significance of the list, some believing that Aśoka’s ability to mention texts by name implies that a canon was already in existence in his day. I will deal with this problem in the eighth lecture.

VIII - Buddhism and Canonicity

We are accustomed to talking loosely of Buddhist texts being canonical or noncanonical. Perhaps we ought to try to define what precisely we mean by “canonical”, because when discussing the period of early Buddhism to which these lectures have been mainly devoted, scholars tend to use the terms in different, and therefore confusing, ways. For example, Étienne Lamotte asserted that there was no question of there being either a canon or a Tripitaka before the end of the Mauryan period, and he stated categorically that there was no canon, Magadhan or otherwise, before the period of Aśoka. ³⁰⁰ Since it has been pointed out that a western term such as “canonical”, although convenient, must nevertheless be used with circumspection, ³⁰¹ we must recognise, when considering Lamotte’s statement, that much hinges upon the definition of “canon”, because there is a danger that we may be perpetrating an anachronism by trying to read the wrong concept of the word “canonical” into the period immediately preceding and following the time of the Buddha’s death.

I want in this lecture to restrict myself to the canons of the Indian schools of Buddhism, and specifically to the Theravādin canon, although there is no reason to think that, in principle, the canons of other Hīnayāna schools were formulated in a different way. We simply do not have enough information to be certain about it.

To define the word “canon” briefly, we may say that it is a collection of scriptures (oral or written), which gives a certain authority to those texts included in it. The collection may be open, giving the possibility of other texts being added to it, or closed, which implies that the texts listed in it, and no others, are documents fundamental to the religion concerned.

For the most part, except where the founder of a religion produces a closed body of inspired scripture, many religions start with a body of texts which have been collected together over a period of time during the life and after the death of the founder, and it is only later, if at all, that the list is closed, sometimes after the weeding out of texts which were at one time thought of as authoritative, but about whose nature doubts have arisen. So in the case of the Christian scriptures, it was only some centuries after the foundation of the church that the present contents of the New Testament were given canonical status, in the second, closed, sense. The situation in Jainism is confusing. One section, the Digambaras, believe that all the early teachings of the founder Mahāvīra have been lost, and what they regard as authoritive is a number of texts written by eminent Jain authors from the second century C.E. onwards. The Śvetāmbara section of the Jain community, however, believe that not all the early scriptures were lost, but even they are not in total agreement about those which remain. One sect rejects a number of texts which are accepted by another, and even the latter cannot agree about the texts which are to be included in the prakīrṇaka “minor” class, with the manuscript tradition rarely agreeing about the texts in this category.

If we speak of the Theravādin canon, in the form in which we have it now, it clearly comes into the second category. It is closed. The concept of canon for Mahāyānists and in Tibet and China differs. Some traditions include texts which are canonical and non-canonical by Theravādin standards, and there was the possibility of making additions at a late date, so that it was, for example, possible for a translation, and a bad translation at that, of the Jātakamālā to be added to the Chinese canon at the end of the eleventh century C.E.

Perhaps we should begin by considering the categories which do not necessarily constitute a canon. There are in the Theravādin canonical texts traces of a different division of the dhamma from the one we have now - one into nine aṅgas. The number is increased in some traditions, with a list of twelve añgas being found in the Mahāvyutpatti. ³⁰² Not all the divisions in the list (suttam geyyam veyyākaranaṃ gātham udānam itivuttakam jātakam abbhutadhammam vedallam)³⁰³ are easily explicable. According to the Dipavamsa, ³⁰⁴ the change from the old ninefold classification into the later one took place at the first joint recitation (sañgiti). This is therefore taken by some as implying that there was a nine-fold canon already in existence during the Buddha’s lifetime. The añgas seem, however, not to refer to precise portions of the canon, but are rather a classification of types of texts, at least if we follow Buddhaghosa’s interpretation of their meanings, ³⁰⁵ and it does not seem possible to take them as evidence for the existence of a canon in the second sense of the word.

In the quotation which I gave at the beginning of this lecture, Lamotte made a distinction between canon and tipitaka. He was right to do so, since it is clear that the word “tipitaka” is not synonymous with “canon”. Tipitaka is a division of texts rather than an assessment of their authority, and this and other terms, such as nikāya, only came into use at some time after the Buddha’s death.

The word tipetakin “having three pitakas” occurs in the canon in the Parivāra to the Vinaya, ³⁰⁶ and tepitaka and tipetaka occur in the Milindapañha. ³⁰⁷ The Vinaya reference occurs in the list of theras prefixed to the Parivāra, which is probably an addition made in Ceylon as late as the 1st century C.E. ³⁰⁸ It is difficult to date the Milindapañha, but it must have been composed long enough before the time of Buddhaghosa for it to be regarded as authoritative, for he seems to quote it on the question of tradition, authority, etc. ³⁰⁹ Nevertheless, these texts, however they are dated, must be some centuries before Buddhaghosa, and the word tipitaka, and the idea of pitakas, must therefore be old.

We are accustomed to refer to the Buddha’s teaching as the t(r)ipitaka “the three baskets”, but this is not the only way in which the Buddhavacana was divided up. Hsūan tsang records the fact ³¹⁰ that the Mahāsānghikas collected the Sūtra-pitaka, the Vinaya-pitaka, the Abhidharma-pitaka, the miscellaneous pitaka (Khuddaka-nikāya), and the Dhāranī-pitaka, and thus they distinguished five pitakas. It is noteworthy that they call the Khuddaka-nikāya a pitaka, but this arises from the fact that the word pitaka is not only used of the baskets of the Buddha’s teachings. There is a Theravādin canonical text called the Cariyāpitaka, and pitaka is used in the Theravādin canon itself to refer to the collections of other teachers, including brahmans. ³¹¹

It is clear that the words pitaka and tipitaka simply denote a type of text or an arrangement of texts, and do not, in themselves, imply any sort of canon, open or closed, although of course the words can be, and are, applied to a body of scriptures which is regarded as canonical.

The use of the word “pāli” in this connection is both confused and confusing. Although it is often translated as “canon”, it is more correctly translated as “text”, since it is the complement to atthakathā (“commentary”), being the text on which the commentary is written.

For the most part, the atthakathās we have are on canonical texts, in the sense of the closed canon which we know of from Buddhaghosa’s list of texts in the three pitakas. There is, however, an aṭthakathā upon the Netti-ppakaraṇa, and in it the author Dhammapāla refers to the Netti as a pāli, and the tīkā also refers to Netti-pāli. ³¹² The bibliographical text Gandhavamsa also calls the commentary upon the Netti an aṭthakathā. ³¹³

This same criterion applies to certain portions of the Jātakas “the birth stories of the Buddha”, which are for the most part in verse, the prose being later and commentarial. There are, however, certain portions in prose upon which there is a commentary, implying that these prose passages are pāli. We in fact can accept that this is so, because almost without exception ³¹⁴ such prose passages occur in the Kunāla-jātaka, and they are, for the most part, composed in a very archaic form of rhythmical prose known as vedha, quite different in style from the ordinary commentarial narrative type prose.

Nevertheless, the whole discussion about pāli and aṭthakathā is made uncertain by the fact that there are certain passages where pāli means nothing more than text, with no implication whatsoever about canonical or non-canonical status. ³¹⁵

The fact of the matter is that there is no single Pāli word which exactly coincides with our word “canon”. It is probable that what we call a canon in this context is what early Buddhists would describe as Buddhavacana “the words of the Buddha”, since they were probably concerned only with what the Buddha taught.

I mentioned in the fifth lecture the four mahāpadesas, ³¹⁶ the four references to authority: the Buddha, a community with elders, a group of elders, and a single elder, mentioned in the Digha-nikāya and elsewhere in the Pāli canon, ³¹⁷ which were concerned with finding out whether a particular teaching was Bhagavato vacanam. The Buddha’s criteria were clearly intended to be applied after his death, when the Buddha himself was no longer available as an authority. In effect, anything that was said to be Buddhavacana and was consistent with the sutta and the vinaya was acceptable. The importance of obtaining such a sutta is shown by the existence of a rule in the Vinaya stating that a monk may interrupt his rains retreat for up to seven days to get a text which may otherwise be lost. ³¹⁸

If, then, by canon we mean a body of texts regarded as having a specific kind of authority, ³¹⁹ there seems to be no reason to doubt that such a body of texts began to come into existence very soon after the Buddha’s death, and indeed might have already existed in embryonic form during his lifetime.

Buddhaghosa talks of bringing together in the Vinaya-pitaka what had been both recited and not recited at the first sangiti, recognising that many sayings of the Buddha had escaped the attention of the sangitikārakas, and also that many additions to the collections were made at a date subsequent to that of the first sangiti. ³²⁰

If, however, the canon is what is Buddhavacana, then we have problems with the various texts which are ascribed to various persons other than the Buddha. We find that notice is taken of this in some of the early canonical texts, where an enquirer received his reply, not from the Buddha, but from one of the Buddha’s close followers, and we are told that this is in effect the word of the Buddha because the follower has answered in exactly the same way as the Buddha would have answered had he been present, and it was therefore approved of by him. ³²¹

The summaries of the story which we find in some texts, e.g. the Mahāparinibbānasutta, including the information about the distribution of relics, the portions of verses which we find in, say, the Suttanipāta, which are ascribed to the sangitikāras, ³²² all go to show that the concept of a closed canon or a canon consisting entirely of Buddhavacana cannot be entirely correct.

The very fact that the Kathāvatthu was recited at the third sangiti, in the reign of Aśoka, somewhere around 250 B.C.E., i.e. a century and a half or so after the death of the Buddha, however, produced a controversy about the propriety of regarding it as Buddha-bhāsita, ³²³ and the Mahāvihāra tradition, reported by Buddhaghosa, had to explain that it was Buddha-bhāsita because the Buddha had drawn up the table of contents, foreseeing that it would be elaborated by Moggaliputta Tissa at a future date. ³²⁴ The reciter was therefore was merely giving words which the Buddha had already put into his mind, so to speak. The Sarvāstivādins faced a similar problem by saying that the Buddha composed the Jñānaprasthāna, but Kātyāyanīputra edited it after the Buddha’s death.

Buddhaghosa, in his account of the first sangiti, states that the fifth nikāya, the Khuddaka-nikāya, included whatever sayings of the Buddha were not included in the first four nikāyas. ³²⁵ Since he lists the texts included in that nikāya, he must have been aware of the fact that many of them, e.g. the Niddesa, Theragāthā, Therīgāthā, Vimānavatthu, Petavatthu, and portions of the Apadāna, are attributed to others. It is, however, noteworthy, that, in the Theragāthā and Therīgāthā, many of the verses are said to have been uttered by the Buddha in the first place. They only became the verses of the thera or theri after they had repeated them.

According to the Saddhammapajotikā, ³²⁶ the commentary upon the Niddesa, the Niddesa was composed by the Buddha’s disciple Sāriputta, but if this was so, then it is strange that he gives three different explanations of his own words as recorded in the Suttanipāta. This would seem to indicate that although some of the Niddesa material perhaps goes back to the time of the Buddha, the work as a whole was later. ³²⁷

What, however, are we to make of a text like the Apadāna, where we find that some individuals are actually acquainted with the Kathāvatthu, ³²⁸ therefore making (those portions at least of) the text even later than Aśoka?

It becomes clear that Buddhavacana was interpreted in a broader sense. Buddhavacana, in this broad sense, means the sayings of the Buddha, but sometimes it means what the Buddha would have said, had he been there, or sayings about the Buddha, or sayings in accordance with the Buddha’s teaching.

How did the Theravādin canon develop from being an open collection of Buddhavacana into a closed canon?

As far as the contents of the Theravādin canon are concerned, we have the information which Buddhaghosa gives us. He lists three pitakas: Vinaya-, Suttaand Abhidhamma-; the Sutta-piṭaka has five nikāyas, the fifth of which, the Khuddaka-, contains fifteen texts, although he also mentions a division into fourteen parts, ³²⁹ without specifying which text is omitted under this scheme; the Abhidhamma-piṭaka contains seven texts. He tells us, however, that some bhānakas had different ideas, and also put texts into different categories. ³³⁰ He was giving information about the situation in his own time, but the situation in earlier times must have been rather different.

In the third lecture I spoke about the tradition that a collection of teachings of some sort existed and was recited immediately after the Buddha’s death, and I dealt with the institution of the bhānakas, the reciters of various parts of the Theravādin canon. I suggested that this event was not as early as the tradition suggests. Although we can point to the references to the foundation of the bhāṇaka system as support for the existence of the nikāyas - the constituent part of the Sutta-piṭaka - at an early date, it is not in the canon itself but only in the commentaries that we find mention of these reciters, ³³¹ although there are references to them in early inscriptions. ³³² There are also inscriptional references to nikāyas, although we have no idea of the contents of those early nikāyas. We must accept, then, that the suggestion that the bhānakas were instituted at the first recitation creates difficulties because the Sutta-piṭaka was probably not in its present form at that time. It is obvious that if the Digha-nikāya, Majjhimanikāya, etc., had not yet been formulated and named, there could hardly have been Digha-bhānakas and Majjhima-bhānakas, etc.

There is also some doubt in the tradition as to when the Abhidhamma-piṭaka came into existence. In place of the expected abhidhamma-dhara “remembering, i.e. an expert in, the abhidhamma” to go with vinaya- and sutta-dhara “expert in the Vinaya and the Sutta-piṭaka”, in a number of places in the canon we find mātikā-dhara, implying that only the mātikās “the indexes” to the Abhidhamma texts existed at an earlier time.

It is quite clear that, as long as the canon was open for any text to be added to those texts, by following the principles enunciated in association with the four mahāpadesas, then the canon in so far as it consisted of the word of the Buddha was not closed. At some stage, however, each of the pitakas must have become closed and the question is to decide whether they all became closed at the same time, or whether at different times, and if at the same time, what that time was.

Ever since the time when the discovery of different sets of canonical texts belonging to different schools of Buddhism undermined the claim of any one version to be the authentic word of the Buddha, some scholars have nurtured a hope that, by delving behind the texts which we have available to us, we can find the texts upon which they are all based, perhaps the very words of the Buddha himself. As part of the task of discovering that underlying canon, scholars have subjected the various versions of the Buddhist canon we possess to careful scrutiny, and by seizing upon the various anomalous factors they have discovered therein they have been able to find traces of earlier languages.

I mentioned, for example, in the fourth lecture that there were different views taken of the non-Pāli features of the Theravādin canon, which were obviously remnants from earlier dialects through which the texts had passed. S. Lévi, ³³³ when he discussed them, spoke of them as “pré-canonique”. Lūders, ³³⁴ however, decided that they were remnants of an Ur-kanon “a primitive or original canon”, the language of which he called “Old Buddhist Ardhamāgadhī”. He regarded this as the language of the original canon of writings. Both scholars had the idea that there was something earlier than the Pāli canon, but as can be seen from the way in which they described the material, it is uncertain whether the earlier form should be described as a canon, or whether it should be described as precanonical. Most scholars now, I think, would be very wary about this whole matter. These forms, in themselves, show nothing more than that the material had existed in another dialect before it was taken over into the Theravādin canon. It is, of course, possible for both Lévi and Lūders to be correct, if they were following different definitions of “canon”.

The problem is to decide when the open canon became a closed canon. When did the body of the Buddha’s teaching gain the status of canonicity? In general, scholars are far from unanimous about the time when the Theravādin canon became closed, and we are not likely to be able to come to any definite answer about this in this lecture. I would hope, however, that we will be able to narrow things down a little.

There are those those who say that the Pāli canon, as we know it, was not fixed until the time of Buddhaghosa who, by listing the texts which he regarded as forming the various constituent parts of the tipitaka, in effect defined and limited the scope of the tipitaka. There are, on the other hand, others who believe that the writing down of the canon in the first century B.C.E. had effectively done that already.

Lamotte seems to have held both views. He stated that the Pāli canon was open until the fifth century C.E., but this is at variance with another statement which he made with reference to the writing down of the Theravādin canon in the first century B.C.E.: “From that moment”, he said, “the text of the Tipitaka in Māgadhabhāsā was drawn up in its final form”. ³³⁵ Although it is possible that he made this statement as a summary of the tradition in the Pāli chronicles, and not as his own point of view, this precise form of words does not occur in any text known to me. I mentioned in the third lecture the fact that Buddhaghosa used commentarial material which was perhaps centuries old, and it is possible that the number and names of texts which make up the canon were indeed fixed at the time of writing the canon down.

Nevertheless, when Lamotte said that, with regard to the Pāli canon, the collections were not closed before the time of Buddhaghosa in the fifth century of the Christian era, ³³⁶ we might think that, while this might be true of the collections as a whole, it was probably not true of the greater part of them. It is hard, for example, to think of the Vinaya-pitaka being open until so late a date. On the other hand, it could well have been true of the Abhidhamma-pitaka, since, as we have seen, the Kathāvatthu was not composed until the time of Aśoka.

We can give tentative dates to the closure of some parts of the Theravādin canon. The fact that the Theravādin version of the Vinaya and the Vinayas of other schools of Buddhism include accounts of the events which led up to the second sangiti, and an account of the second sangiti itself, shows that those accounts were composed after that event. Traditionally it took place 100 years after the death of the Buddha, but in view of the fact that 100 is a round figure, not intended to be accurate, we can probably assume that the second sangiti was somewhere around 60 years after the death of the Buddha. There is, however, no reference in the Vinaya to the third sangiti, held during the reign of Aśoka, i.e. about 150 years after the Buddha’s death. This would suggest that the Vinaya had not been closed by the time of the second sangiti, but was closed before the third sangiti.

The fact that the Theravādin canon includes the Kathāvatthu shows that the Abhidhamma-pitaka was not closed before the third sangiti. Similarly, since the Niddesa is a commentary upon a canonical text, it must be later than that text. If the Kathāvatthu was the last text added to the Theravādin canon, then the Niddesa is earlier than this. The Dipavamsa states that the Niddesa was one of the texts rejected by the Mahāsānghikas after the second sangiti. ³³⁷ This perhaps is not to be taken literally, but it means that the Niddesa was not included in the Mahāsānghika canon, either because it did not exist at the time of the schism, or because it had not yet attained canonical status at the time the Mahāsānghikas split from the Theravādins, at some time before the third sangiti.

With these exceptions, we probably have to agree in principle with those who date the closure of the canon to the time of Buddhaghosa. There is, however, some evidence which suggests that the canon had been closed at an earlier time. As I said in the fifth lecture, despite the separation of the Sinhalese Buddhists from North India, ³³⁸ it seems that literary material continued to reach Sri Lanka, but there is no evidence for the addition of any complete text to the Theravādin canon after the sangiti at the Ālokavihāra in the first century B.C.E. An origin in North India is postulated for the Milindapañha, ³³⁹ the Petakopadesa and the Nettippakarana. The fact that these texts are highly regarded by the commentators, but are not given canonical status, suggests that they arrived in Sri Lanka after the closure of the canon. This view is supported by the fact that these post-canonical works contain a number of verses and other utterances ascribed to the Buddha and various eminent theras, which are not found in the canon. It was possibly such material that Buddhaghosa had in mind when he referred to some of the Buddha’s utterances being pāli-muttaka “not included in the texts”. ³⁴⁰ Nevertheless, there was no attempt made to add such verses to the canon, even though it would have been a simple matter to insert them into the Dhammapada or the Theragāthā. This perhaps supports the suggestion I have made that Buddhaghosa was sometimes giving the views of commentators who lived long before him, so that the canon had effectively been closed before his time.

Another point to consider, and I will return to this in the ninth lecture, is the fact that some of the best known stories in Buddhism, e.g. the story of the kṣatriya maiden Kisāgotamī, ³⁴¹ are known in the Theravādin tradition only in the commentaries, although they are found in texts which are regarded as canonical in other traditions. It would make no sense to say that these stories are inventions of the Northern Buddhists which reached Sri Lanka too late to be included in the canon, if it was really open until the time of Buddhaghosa, who translated many of the Mahāvihāra commentaries into Pāli. We have, therefore, to assume that at least the Vinaya- and Sutta-piṭaka had been closed at an earlier date.

There is some doubt about the status of four texts (the Petakopadesa, Suttasangaha, Nettippakarana and the Milindapañha) in Burma. In a survey of Pāli literature in Burma they are said to be recognised as canonical in that country, ³⁴² and certainly the Nettipakarana, Petakopadesa and Milindapañha are added to the usual fifteen texts of the Khuddaka-nikāya in the edition of the canon which was carved on stone by King Mindon after the fifth sangiti in the nineteenth century, and also in the printed sixth sangiti (Chaṭthasangāyana) edition in this century. A reviewer of the survey, ³⁴³ however, said that this statement was mistaken, and commented: “No educated Burman, lay or monk, ever included these four works among the pitaka books of the Khuddakanikāya. They are placed after the books of the Khuddaka-nikāya only because of their intrinsic value (this applies to the first and the last two, the second [the Sutta-sangaha] being merely a collection of suttas) as a help to the study of the Scriptures”.

The Nettippakarana itself includes the statement that it was recited at the first sangiti, in which case we must ask why Buddhaghosa did not include it in his lists, and it is perhaps the designation as pāli, which I have mentioned, which has led to the Netti being added at the end of the Khuddaka-nikāya in Burma. We can also see why the Sutta-sangaha might be included. It is in fact a selection of the Buddha’s suttas and is, therefore, Buddhavacana. We are back to our original question of what constitutes a canon. If these texts are published with the canon, how are we to decide whether they are regarded as canonical or not?

I have just said that the fact that the Kathāvatthu could be added to the Abhidhammapitaka at the time of the third sangiti shows that that pitaka at least had not been closed before the reign of Aśoka. Do we have any firm evidence that there was a canon of any sort in existence at that time?

Aśoka’s Bairāt inscription, in which he recommends seven texts to the sangha, ³⁴⁴ has been variously interpreted. There is the problem that we cannot be certain of the identity of some of the texts which Aśoka recommends, but since we sometimes find that the commentaries, e.g. that on the Sutta-nipāta, ³⁴⁵ know of some texts under other names, ³⁴⁶ it is perhaps not altogether surprising that we cannot recognise all of Aśoka’s choices. It has been suggested ³⁴⁷ that, as Aśoka specified the “Exhortation to Rāghula in regard to lying”49 as a text to be studied, he must have known of the existence of the other exhortation to Rāghula, ³⁴⁸ but this does not follow - it may have been his teacher who specified the sutta in this way. The question is whether this reference to texts indicates that the Buddhist canon was already in existence in Aśoka’s time, as some have suggested.

He clearly had knowledge of some individual texts. In what language were they written? The titles are in an Eastern Prakrit, but that may simply be as a result of Aśoka using such a Prakrit. It tells us nothing about the dialect of the texts - presumably an Easterner could read a Western dialect - but on balance it seems more likely that he knew the texts in an Eastern form. And it was therefore not the Theravādin canon as we know it. It seems to me that all that Aśoka’s inscription proves is that texts were still thought of as individual suttas, not as parts of nikāyas or other large collections, and that these seven texts were already thought of by some, including Aśoka (and presumably his teacher who recited them to him), as of particular value.

The fact that Aśoka specified suttas by name, rather than referring to nikāyas or pitakas need cause no surprise. Although there are very few references in the Theravādin canon to portions of the canon by name, what references there are, are to individual suttas, and there is evidence that texts were often spoken of as individual suttas, by name, rather than as parts of nikāyas. ³⁴⁹ We should remember that the Niddesa is a commentary on two vaggas and one sutta of the Sutta-nipāta, which suggests that those vaggas were still thought of as individual vaggas, when the Niddesa was composed. As can be seen from catalogues of manuscripts, what was transmitted in manuscripts was frequently individual suttas rather than the whole sutta-pitaka, or even a single nikāya. ³⁵⁰ Similarly, it is interesting to note that when texts were exported to other countries, e.g. Burma, it would seem that they were taken as individual texts or as groups of texts, rather than as nikāyas or pitakas. ³⁵¹

When Lamotte wrote of the impossibility of there being “a canon or a Tripitaka” before the end of the Mauryan period, it is not clear what distinction he was making between the two, but if by “canon” he meant a closed canon, then there can be no doubt that he was correct. Nevertheless, the fact that Aśoka could quote titles of suttas means that those texts were available for recitation to laymen, and in that sense we can say that some sort of collection of the Buddhavacana was in existence in Aśoka’s time.

We know little about the closure of the canons of other schools of Hīnayāna Buddhism. As I said earlier, there is no reason to think that, in principle, the canons of other Hīnayāna schools were formulated in a different way. We can surmise that, like the Theravādin canon, they were at first open, but were closed at some later date.

The Dīpavamsa tells us something about the contents of the Mahāsānghika canon: “They rejected the Parivāra, Abhidhamma, Patisambhidā, Niddesa, and a portion of Jātaka”. ³⁵² This probably means, not that they actually rejected these texts at the time of the split from the Theravādins, but that they were not in the Mahāsānghika canon. Since all these texts, with the exception of the Jātakas, are accepted as late compositions, on stylistic and other grounds, we can surmise that they were not in the Mahāsānghika canon because they did not exist when the split took place. We may therefore date their composition some time after the second sangiti. ³⁵³ In the case of the Jātakas, the collection available to the Mahāsānghikas would have represented the size of the Jātaka collection at that time. Further additions must have been made by the Theravādins before the canon was finally closed, ³⁵⁴ just as the presence of Jātaka stories in the Mahāvastu ³⁵⁵ (which is described as being based on the redaction of the Vinayapitaka made by the noble Mahāsānghikas, the Lokottaravādins of the Middle Country) ³⁵⁶ which have no counterpart in the Pāli collection shows that the Mahāsānghikas also continued to add Jātaka stories to their canon. ³⁵⁷

The Dīpavamsa gives a description of the Mahāsānghikas’ canon: “Altering the original redaction, they made another redaction. ³⁵⁸ They transposed suttas which belonged to one place (in the collection) to another place; they destroyed the meaning and the doctrine which were in the five nikāyas. The monks, not knowing what was taught with exposition or without exposition, neither the natural meaning nor the recondite meaning, placed in one place the meaning which was spoken with reference to another. Those monks destroyed much (true) meaning under the shadow of the words”. ³⁵⁹ These words might have been written with the Mahāvastu in mind, for the inclusion in it of suttas which were in the Theravādin Sutta-nipāta, e.g. Sabhiya-sutta, Nālaka-sutta, and Khaggavisāna-sutta, and Jātakas which were in the Theravādin Jātaka collection, e.g. Kusa-jātaka and Sarabhañga-jātaka, would inevitably give the impression to any Therāvadin who examined the Mahāsānghika canon that the Theravādin collections had been re-arranged. ³⁶⁰ The phrase “placed in one place the meaning which was spoken with reference to another” describes very accurately the displacement of some of the pādas in the Sabhiya-sutta from their original position, so that they are no longer in close contact with the words they are defining.

An example of the uncertainty about the meaning of the word “canon”, which I mentioned earlier, can be seen in Lamotte’s use of the word when referring to the Gāndhārī dialect. He stated ³⁶¹ that the mere existence of the Gāndhārī [Dutreuil de Rhins] Dharmapada “does not allow us to infer the existence of a canon in North-Western Prakrit”. The fact that one text does not prove the existence of a closed Tripitaka is so self-evident that it seems unlikely that Lamotte could have been using the word in that sense. If he was using the word in the other sense of “canon”, then we must remember that he was writing before the publication of John Brough’s edition of the Gāndhārī Dharmapada, in which the idea of such a canon was considered in detail. Brough concluded that it was difficult to believe that a group of monks, whom he identified tentatively as Dharmaguptakas, might have possessed a Dharmapada, without at the same time possessing at least some stock of Sūtra and Vinaya works. ³⁶² It has also been shown that some Chinese translations of canonical texts appear to have been made from the Gāndhārī dialect, or something very similar to it. This does, however, raise the question of whether such a collection merits the name of “canon”, and clearly this depends on which definition of “canon” we adopt.

If we accept the latest date proposed for the closing of the canon, i.e. the time of Buddhaghosa, ³⁶³ then we should note that, in some cases, this may have been effective only so far as the titles of texts were concerned, not necessarily the contents.

The form of some texts, including the Kathāvatthu, Dhammapada, Apadāna, Theragāthā, and Therīgāthā, is such that even when the texts themselves had been accepted as canonical and included in an individual nikāya, or pitaka of the Buddhavacana, additions could still be made. As I mentioned in the third lecture, the uddānas to the Theragāthā and Therīgāthā, the index verses found at the end of the text, give a list of the number of theras and theris and their verses which does not accord with the texts as we have them now, showing that changes have been made since the uddānas were formulated, doubtless at one of the sangitis, by those who were holding the sangiti, the sangiti-kāras.

It is probable that the Kathāvatthu which was composed at the time of the third sangiti, was in fact only the core of the text as we have it, and its form was not finally fixed until the early commentators, whose work Buddhaghosa made use of, wrote their commentaries upon it. Even after that time we might believe that minor additions could be made. The first part of the Kathāvatthu deals with items which we can identify with the five points of Mahādeva, which the other Buddhist traditions state caused the schism in the Buddhist sangha which led up to the third sangiti, at which the Kathāvatthu was recited. Because of the form of the text, there would have been no problem whatsoever in adding further material dealing with other points of doctrinal interest, over the next few centuries, until the text was finally fixed.

As a general principle, we may note that once a portion of a text had been commented upon, then the presence of canonical words in the lemmata, and in the explanations of those words, meant that changes were very unlikely to be made. We must, however, note that not every word in every text is commented upon, and we must not assume that because something is not commented upon it must be a later addition. All we can say is that a word or phrase, which is commented upon, was in the text in front of the commentator. There are, however, in the Suttanipāta verses which have no old commentary upon them, ³⁶⁴ and it might be argued that they were not commented upon because they did not exist at the time when the commentaries were being compiled.

We can conclude that the form of the Theravādin canon, and the texts it comprises, are fixed by the information Buddhaghosa gives. He records other views about this, but clearly his version prevailed. The latest date for the closing of the canon, i.e. the date of the establishing of its canonicity, is, then, the fifth century C.E. We can, however, to a certain extent, narrow down the date of the closing of the canon. We have seen that the Vinaya includes an account of the events which led up to the second sangiti, but not to the third sangiti, and it must therefore have been closed at some time between the two. The Abhidhammapitaka includes the Kathāvatthu, and cannot therefore have been closed until the time of the third sangiti in the reign of Aśoka. The canon was in all probability closed some time before the time of Buddhaghosa because stories and texts and verses from North India did not find their way into the canon.

I suggested in the fifth lecture that the writing down of the Theravādin texts prevented any further changes in the language of the texts, and it is highly likely that the same event also prevented any changes in the contents of the canon, with the exception of additions to those texts whose structure permitted them. Such additions were finally stopped by the comments made upon them by the commentaries, mainly in the fifth and sixth centuries, although we must note that these commentaries were mainly based upon earlier material, and so in effect the prevention of additions dated from an earlier period. A text like the TherigāthāApadāna, having no commentary, was only partially limited by the quotations from it in other texts, especially the Therigāthā-atṭhakathā.

If these guesses are correct, then the canonicity of the Theravādin texts was effected some time after the writing down of the canon in the first century B.C.E. - perhaps in the first century C.E. - and we should perhaps see writing and canonicity together as two parts of an attempt by the Theravādins of the Mahāvihāra to legitimise their claims to be the true recorders of the Buddhist tradition. It seems clear that the Theravādin school have used the canon for their own purposes. Their aim was doubtless to make it clear that their texts, and their texts only, represented the true Buddhavacana. ³⁶⁵

There is, however, an interesting post-script to this discussion of canonicity. As we have seen, the Buddha allowed for texts being added to the collection of suttas, presumably after his death, and Buddhaghosa accepted that not all texts had been recited at the first sangiti. Although it is clear that the sources which Buddhaghosa was following had defined exactly the texts which were, according to their belief, included in the tipitaka, it is not certain whether Buddhists at a later time thought that their canon was closed, and that nothing beyond what had been listed could be regarded as canonical. The fact that the Buddha had given criteria for determining whether a newly found sutta was authentic or not, meant that it was always possible for another text to be brought forward and added to the body of material. If something looks as though it is Buddhavacana, i.e. it follows the pattern of earlier accepted texts, contains the typical formulae, includes a reference to the Buddha preaching, and has in it nothing which is at variance with the sutta and the vinaya, then why should this not be Buddhavacana, in the broadest of the senses I have mentioned?

It is probable that by medieval times, once the Mahāvihāra had triumphed over the Abhayagirivihāra, the need to maintain the closed nature of the canon became less important, and the distinction between canonical and non-canonical became blurred. This had already received an impetus from the fact which I have already mentioned, that some of the best known stories about the Buddha were in non-canonical texts. It is certainly the case that canonical and non-canonical texts were often written in the same manuscripts, ³⁶⁶ and preserved by the same tradition, in the same place, at the same time. ³⁶⁷

Consequently we find individual Jātaka stories and collections of such stories in south-east Asia which are not identical with the collection of stories in the Theravādin canon, but are clearly modelled upon that collection in their formal structure. ³⁶⁸ They are sometimes called “Apocryphal”. ³⁶⁹ There are also individual texts found in the same area. By title they are suttas; they have the standard canonical opening evam me sutaṃ ekam samayam …; the narrative attributes their contents to the Buddha; they meet the requirements of the Buddha’s criteria, in as much as they contain nothing which is not in conformity with the Theravādin version of the Buddha’s teaching, but they are not included in any edition of the Pāli canon. These are sometimes called “quasi-canonical” or “allegedly non-canonical”. ³⁷⁰

It is probable that the purpose behind such works was merely the desire to produce teaching material, or sermons. Having composed something which was suitable for these purposes, based upon scriptural material, there perhaps seemed to be no reason why the usual scriptural trappings - the opening formula and standard phraseology, etc., should not be added. There was no intention to deceive.

I know of no evidence that would make me think that those who composed such works were deliberately forging material which they hoped to pass off as Buddhavacana. There is nothing which would make such a thing desirable. The fact that the newly found sutta had to be compatible with the sutta and the vinaya would mean that the possibility of anyone forging a document to support a new heresy he had just thought up was ruled out. It was not like the situation in Chinese Turkestan at the turn of the century when the presence of European explorers searching for ancient documents meant that there was a market for forged documents. Nor is it like the Chinese translation of the Jātaka-mālā, which is not so much forged (as its description as a “pseudo-translation”73 implies), as made by incompetent scholars.

I have already mentioned the fact that one of the canonical texts, the Niddesa, is a commentary upon parts of the Suttanipāta, made - it would seem - so early that it could be accepted as part of the canon. I will deal with this at greater length in the ninth lecture.

IX - Buddhism and the Commentarial Tradition

Commentaries were probably needed from the very beginning of Buddhism. We can assume that the earliest form of commentary was simply the explanation of one word by another, and we need not doubt that the Buddha on occasion replaced words by synonyms, to make his teachings more easily understood. We find some sort of evidence of this in the words of the Buddha himself, where he gives, for example, seven different words all meaning “bowl”, ³⁷¹ which he says are “country language” (janapadanirutti). This is presumably something which he found from experience, i.e. as he moved from one place to another, he found that the word used for a particular object in one area was quite different from that used in another area for one and the same thing. Anyone who has ever looked at Sir George Grierson’s Bihar Peasant Life, ³⁷² will know exactly the problems which the Buddha faced. Grierson moved around the area for which he had responsibility as a civil servant, recording the wide range of different names which the villagers of Bihar used for all the items of everyday domestic and village life. The Buddha must have found a situation which was very similar to this 2½ millennia ago.

The earliest extensive portion of Theravādin commentary material we have is that found in the Vinaya-pitaka. We know that, at the suggestion of King Seniya Bimbisāra of Magadha, ³⁷³ the Buddha instituted a fortnightly recitation of the Pātimokkha (following the practice of other sects which recited texts). We are therefore certain that the Pātimokkha existed from the very early days of Buddhism. The Pātimokkha does not have a separate existence in the Theravādin tradition, but is embedded in the Vinaya-pitaka. If it is the earliest text for which we have direct evidence of recitation, it is not surprising if it was the first text that needed explanation. And so we find a simple type of commentary upon it, whereby words are for the most part explained by synonyms or clarification of terms, e.g. the first word of the first pārājika rule is yo “whoever”, and it is explained: “‘Whoever’ means he who on account of his relations, his social standing, his name, his clan, his morals, his dwelling, his field of activity, an elder, or a novice, or one of middle standing, this is ‘whoever’”. ³⁷⁴ The nature of that commentary is made clear by its name: pada-bhājanīya “analysis of words”. The commentary is so early that, like the Pātimokkha itself, it is embedded in the Vinaya-pitaka after each of the Pātimokkha rules, and therefore the text and the commentary have been handed down together.

The earliest Theravādin commentarial text is one which is actually given separate canonical status. It is the Niddesa, a commentary upon two vaggas and one sutta of the Sutta-nipāta. Here again we find a great deal of explanation by means of synonyms. We also find another characteristic feature of this type of commentary. Whenever a particular word occurs in the text, the same explanation is given verbatim, e.g. whenever there is a reference to kappa “figment”, or a verb based upon the root kapp-, we are told that there are two types of kappa, i.e. tanhā-kappa and ditthi-kappa, and the identical explanation is given, even if the words recur in successive verses. In the third lecture, I pointed out that this kind of fixed repetition is typical of the oral nature of an early text, giving support, if it were needed, for the assumption that the Niddesa is very old and was composed well before the writing down of the canon.

Even the early commentarial passages in the canon itself are not merely lists of synonyms. Commentaries have two functions, although not all commentaries perform both of them all the time: they explain the meaning of the words and they explain the meaning of the phrase or sentence in which the words occur. In the Aranavibhangasutta “The discourse on the undefiled” of the Majjhimamikāya, ³⁷⁵ the arana “undefiled person” is defined in one paragraph by the Buddha. The remainder of the sutta is devoted to explaining the meaning of each sentence of that paragraph. Similarly, the Niddesa is not merely a list of synonyms. It also includes exegetical passages, giving us some idea of what the early commentarial tradition thought was the meaning of the relevant portions of the Sutta-nipāta. It is, however, not a organic structure of exegesis, but a series of disconnected phrases, which serve as explanations of the individual words, not in the particular context of the Sutta-nipāta, but in any setting.

The chronicles tell us that, at the time of Aśoka, Mahinda took Buddhism to Sri Lanka. ³⁷⁶ The story, as related in the Dipavamsa, the Samantapāsādikā, and the Mahāvamsa, tells how King Tissa met Mahinda, exchanged greetings with him and had a discussion. All three sources agree that the first sermon which Mahinda preached to the king was the Cūlahatthipadūpamasuttanta “The short sermon on the elephant’s footprint”. ³⁷⁷ The Mahāvamsa ³⁷⁸ states that Mahinda preached the true doctrine in two places “in the speech of the island” (dvīsu thānesu dhammam bhāsitvā dīpabhāsāya), implying that in the other places he did not “translate”. 9 The Dipavamsa and the Samantapāsādikā make no mention of this “translation” process, and although there are instances of the Mahāvamsa including authentic historical material which was either unknown to, or consciously omitted by, earlier chroniclers, it is possible that the need for “translation” was deduced by the author of the Mahāvamsa from the statement repeated at the beginning of Buddhaghosa’s commentaries, that they were translations into Pāli of the commentaries which were brought to Sri Lanka by Mahinda and put into Sinhalese for the benefit of the islanders. ³⁷⁹ The author of the Mahāvamsa may well have thought that if Mahinda needed to “translate” the commentaries then he must have “translated” the canonical sermons too. The fact that the “translation” process is said to have happened in only two places suggests, however, that the statement is based upon tradition. If it is true, we may deduce that in some places Mahinda preached the suttas in their canonical form, whereas in two places he added an exegesis, in Sinhalese Prakrit.

There is no evidence about the language of the commentaries which Mahinda is said to have brought with him to Sri Lanka. They could have been in the same language as the canonical texts, whatever that was, or they might have been in the language of the missionaries, in so far as they may have spoken a different Prakrit. We have no evidence earlier than Buddhaghosa’s statement which I have just mentioned, ³⁸⁰ that Mahinda did bring commentaries with him, but we have no reason to doubt that commentaries were brought from India either by Mahinda or other missionaries coming after him. The commentaries seem to have contained much information about India, and particularly North India, which could only have been brought from there. As we shall see, the Pāli commentaries sometimes employ a form of words in their explanations which is closer to the versions found in Sanskrit or the Gāndhārī Prakrit than the Pāli version which is being commented upon, ³⁸¹ and this can only be explained by assuming that the Pāli commentary and the Sanskrit and Gāndhārī versions go back to a common source, which must have been in North India.

There is evidence for dialect differences in the early commentarial material. We know from the Aśokan inscriptions that in the third century B.C.E. there was a dialect variation in the treatment of consonant groups. In the East the dialects tended to resolve consonant groups by inserting an epenthetic (svarabhakti) vowel. In the West such consonant groups were simplified by assimilation. We occasionally find that a word is found in the Pāli canon which shows one regional variation, while the explanation of the word which is preserved in the commentaries shows a cognate form from another dialect, e.g. we find in the Dhammapada ³⁸² the phrase vissam dhammaṃ samādāya, usually translated “having adopted the whole law”, assuming that vissa is from Sanskrit viśva “all”. ³⁸³ The commentary, however, explains this as “vissa means taking up a dhamma that is uneven (visama) or a dhamma that is foul-smelling, concerned with physical activities, etc.”. ³⁸⁴ The alternative explanation signified by “or” (vā) shows that the commentator was uncertain about the meaning of the word, but nevertheless the tradition of the Sinhalese aṭthakathā he was following had preserved an alternative dialect form.

One Sanskrit version of the verse has veśma-dharma ³⁸⁵ and it is probable that the Gāndhārī ³⁸⁶ version also has veśma-. The meaning therefore is “domestic manner (of life)”. ³⁸⁷ In the Pāli version vissa stands for vēssa, where -ss- results from the assimilation of -śm- > -ss-. The first explanation in the commentary shows the alternative development of an epenthetic vowel being inserted, i.e. veśma > *viśma > *visma > visama, which the commentator took to be the equivalent of Sanskrit visama “uneven”. Another Sanskrit version of the same verse actually has visama, ³⁸⁸ which is clearly a backformation from the same tradition which preserved the Pāli commentarial reading visama. We may presume that the verse was uttered at least twice, once in an area where consonant groups were assimilated, and once where they were resolved. These explanations in different dialects must therefore antedate the introduction of the canon into Sri Lanka, and they show that the dialect variation already existed at an early date, perhaps at the time of the Buddha.

Other explanations which depend upon non-Pāli and non-Sinhalese Prakrit forms must also antedate the introduction of the canon into Sri Lanka, e.g. the explanation of cola ³⁸⁹ “top-knot” by cora “thief” must depend upon a source which, like Māgadhī, confused l (ḷ) and r. Similarly, abhibhāsana ³⁹⁰ “illuminating” is explained as tosana “delighting”, ³⁹¹ but this is really the explanation of abhihāsana, ³⁹² and shows that the explanation must have been made in a dialect where both abhibhāsana and abhihāsana appeared as abhihāsana.

The Critical Pāli Dictionary ³⁹³ refers to nominative singular forms in -e in two commentarial passages ³⁹⁴ and claims that these are from the <Sīhala→Atthakathā, because Proto-Sinhalese, like Māgadhī, was an -e dialect. These two passages are virtually identical, and it seems unlikely that, of all the hundreds of Sinhalese Prakrit forms in -e which must have been in the Sinhalese commentaries, Buddhaghosa not only forgot to “translate” into Pāli the first time he gave the explanation, but also repeated his error a second time. We may therefore deduce that Buddhaghosa deliberately quoted the passage with the -e forms (attributing it to the Atthakathā the second time), perhaps because it was invested with such authority in the Sinhalese commentary that he felt obliged to quote it verbatim. We may note that something very similar to the passage occurs in the Kathāvatthu, ³⁹⁵ and there is evidence ³⁹⁶ for believing that it was a common phrase in the early Buddhist tradition in Magadha. I conclude, therefore, that these forms are not remnants of the Sīhala-atthakathā.

Childers was of the opinion ³⁹⁷ that the commentaries brought by Mahinda were in Pāli. If Childers was right, we have to assume a progression from Pāli to Sinhalese Prakrit and back to Pāli again. While this is not impossible, it does raise the question of why the Pāli of the canon was sacrosanct, while the Pāli of the commentary could be translated. It also raises the question of the nature of Pāli. If it means the language of the Theravādin canon, then it is unlikely that at the time of Aśoka, when Mahinda came to Sri Lanka, the language of that canon was already a western Prakrit with a certain amount of Sanskritisation introduced into it. I think that the Sanskritisation came later, perhaps because of the rise of the Mahāyāna. It would seem more likely that the commentaries already represented a heterogeneous mass of material, in various dialects, and probably including comments on readings which differed from those in the canon as established in Sri Lanka.

In the fifth lecture, on Buddhism and writing, I quoted the passage in the Dipavamsa which stated that prior to the reign of Vattagāmini Abhaya the tipitaka and the aṭthakathā were transmitted orally, but in his reign they were written down. We tend to concentrate on the fact that it was the tipitaka which was written down, without thinking much about the commentaries because, to us, the word “commentaries” implies the works composed, in particular, by Buddhaghosa and Dhammapāla in the fifth and sixth centuries C.E. respectively. The reference in the Dipavamsa must, however, be to whatever commentarial material was available in the first century B.C.E., and which had hitherto been transmitted orally in conjunction with the canonical texts to which it referred.

The fact that the commentarial material was already of a disparate nature would probably have led to an attempt to impose homogeneity upon it, and also to make it more intelligible to the Sinhalese bhikkhus by translating it into the vernacular language. Because it was in the vernacular it would have been easy for additions to be made to it. That there was a continuing commentarial tradition in Sri Lanka itself is shown by the fact that Buddhaghosa quotes (from the Sinhalese commentaries he was using) the names of individual Sinhalese theras whose views he is accepting or rejecting.

According to the Cūlavamsa, there were no commentaries available in India in the fifth century C.E., and so the thera Revata suggested to Buddhaghosa that he went to Sri Lanka and translated the Sīhala commentaries into Māgadhi. ³⁹⁸ At the beginning of the Samantapāsādikā, Buddhaghosa states that his work will be based on the Mahā-atṭhakathā and the Mahāpaccariya, while also taking into account such commentaries as the Kurundi. He also quotes from the Andhaka and the Sankhepa, and the Paccarī, but it is not clear whether this last named is the same as the Mahāpaccariya.

There has been much speculation about the Andhaka-atthakathā. It is said ³⁹⁹ that it was handed down at Kāñcipura in South India. It is further stated ⁴⁰⁰ that it was very likely written in the Andhaka language. Probability then becomes certainty, and we than find statements such as “the references relate to … the Sinhalese and some of the Dravidian commentaries”, ⁴⁰¹ and “Buddhaghosa drew his material not only from Sinhalese and Dravidian but also from … Pāli”. ⁴⁰² Comparable statements are made about Dhammapāla: “Dhammapāla … very probably was a Dravidian by birth. It is also likely that he made use of Dravidian commentaries”. ⁴⁰³

There was, in all probability, some connection between the Andhaka commentary and either the Andhaka country or the Andhaka sect, and it is very likely that the sect was so called because it came from the country. Probably the Andhaka commentary came from the Andhaka country originally, for Buddhaghosa quotes it ⁴⁰⁴ as referring to conditions in that country. It must, however, like the other commentaries, have had a core of material which came from North India, for Buddhaghosa quotes the Andhaka-atthakathā when he is talking about the Magadha-nāli ⁴⁰⁵ “a particular measure” but quotes the Mahāatthakathā for the Damila-nāli in the same passage. The fact that the Andhaka commentary is usually quoted only to be refuted ⁴⁰⁶ tends to support the view that the basis, at least, of the Andhaka commentary belonged to the Andhaka sect.

About the language of the Andhaka commentary nothing can be said definitely, but it would seem clear that by Buddaghosa’s time it was no longer available in South India. The statement that no commentary was available in India, ⁴⁰⁷ which is given as the reason for Buddhaghosa to go to Sri Lanka, was not likely to have been included in the Sinhalese commentaries upon which the Mahāvamsa was based while there were South Indians such as Dhammapāla available to refute it, if it was false. It is in any case clear, from Buddhaghosa’s statement, ⁴⁰⁸ that the view of the thera Mahāsumma was regarded as authoritative in the Andhaka-atthakathā, that the commentary must have been introduced to Sri Lanka some time before Buddhaghosa, for Mahāsumma is a Sinhalese thera who is datable to the first century C.E. Since, however, Buddhaghosa stated plainly that he translated Sinhalese commentaries into Māgadhī, we can be fairly certain that if the Andhaka commentary was originally composed in a Dravidian language, it had already been translated into Sinhalese Prakrit by Buddhaghosa’s time.

It is usually stated ⁴⁰⁹ that no material was added to the Sinhalese commentaries after the first century C.E., but a very careful study ⁴¹⁰ of the individuals mentioned in the commentaries has shown that King Mahāsena is mentioned by name in the Samantapāsādikā. ⁴¹¹ Since this king is datable to 276-303 C.E., this shows that additions to the Sihala-atthakathās continued to be made until the very end of the third century or the beginning of the fourth.

The author of the commentary upon the Sutta-nipāta notes that the Sinhalese commentary did not comment upon two verses in the Sutta-nipāta, ⁴¹² and he therefore excludes those verses from the number of verses in the original sutta. ⁴¹³ This statement has been taken ⁴¹⁴ as pointing with more or less certainty to an addition made to the canon in Sri Lanka. The verses could equally well, however, have belonged to a different recension from that being used by the Mahā-atṭhakathā, and been either unknown to the author of the Sihala commentary or ignored as being less authentic. It is clear that different recensions of some texts did find their way to Sri Lanka. Buddhaghosa and the other commentators frequently record the existence of variant readings, and it is questionable whether these would all have been in one and the same version of the canon. It is interesting that the commentator on the Sutta-nipāta did not invent a commentary upon these verses. It would seem that the commentary was, as far as he was concerned, closed.

If the Buddha himself explained X by Y in one place, but Y by X in another, then we can see that X and Y were both Buddhavacana, and this situation might lead to a different sort of commentarial tradition, whereby the commentarial explanation in one tradition was the canonical reading in another tradition, and vice versa. I have just mentioned the comparable situation with regard to vissa, veśma and visama.

It has been shown ⁴¹⁵ that in a number of cases other (non-Pāli) traditions preserve as canonical forms which are akin to those in the Pāli commentary, showing that those traditions thought that they were Buddhavacana. For example, where a verse in the Dhammapada ⁴¹⁶ includes the words adhisessati, chuddho, and apeta-viñ̃̃āno, the Gāndhārī Dharmapada equivalent ⁴¹⁷ has three different words in their place (vari śa’isadi, tuchu, and avakada-viñana). The glosses in the Pāli commentary, however, are the equivalents of the Gāndhārī forms (upari sayissati, tuccha, and apagata-viñ̃̃āna, Dhp-a I 320-21), which clearly shows acquaintance with the same tradition as underlies the Gāndhārī version. A possible explanation of this is that the Buddha uttered the same sermon on more than one occasion, sometimes making changes, as the occasion or locality demanded. When tradition preserved more than one version, it sometimes kept one as an explanation of the other.

Another example of this parallelism between canonical and commentarial traditions, which seems to cross the boundaries of sect, can be seen in Buddhaghosa’s commentary on the incident where bhikkhus ask for permission to change the Buddha’s words chandaso. He glosses: “Let us translate chandaso means, let us translate into Sanskrit (or into the refined language) as (we translate) the Veda”. ⁴¹⁸ Here “into the refined language” (sakkata-bhāsāya) seems to correspond to phrases in two of the Chinese versions of the same incident. The Dharmaguptaka version includes the request in the form “in accordance with the fine language (or perhaps texts) of the world”, while in the Vinaya-mātrka version the Buddha states that his doctrine “is not concerned with beautiful language”. It seems very likely that both these canonical phrases are based upon originals which included the equivalent of samskrta or satkrta. ⁴¹⁹

It is an interesting fact that in his commentary on the heretics’ views which are quoted in the Sāmaññaphala-sutta, ⁴²⁰ Buddhaghosa seems to include phrases which are akin to those in the Tibetan version of the Pravrajyā-vastu.⁴²¹ He glosses: “cooking means beating others with a stick”, ⁴²² which seems to be a combination of two phrases in the Tibetan: ‘Who grills and lets grill, who beats and lets beat’. ⁴²³ At the end of his explanation of Ajita’s doctrines, he comments: “fools give, the wise take”. ⁴²⁴ The Tibetan version states: ‘Thus (only) the fool accords instruction, the sage receives instruction’. ⁴²⁵

These parallels suggest that the commentarial and canonical traditions which underly them are equally old, and the material upon which they are based must pre-date the separation of the sects. It seems probable that some of this material, since it is accepted as canonical by some sects, must go back to the earliest days of Buddhism, perhaps to the time of the Buddha himself. A close comparison of the Theravādin commentaries with non-Theravādin canonical texts might well bring to light other parallels of this nature. This might help us to date the various strata of the commentarial tradition, which would, in turn, enable us to estimate its value more accurately.

Sometimes there is mention of an event in a commentary, with no reference at all to it in the canon, although it appears in the canon of another school. As I mentioned in the eighth lecture, a problem arises from the fact that some of the best known stories in Buddhism are known in Pāli only in the commentaries, e.g. the story of the kzatriya maiden Kisāgotamī. ⁴²⁶

In some cases, portions of Buddhist history seem not to have reached Sri Lanka at all, e.g. the early history of Devadatta’s hostility to the Buddha, and the reason why the Buddha called him “lick-spittle”, ⁴²⁷ khelāsika. ⁴²⁸ Buddhaghosa does not know the reason, and the incorrect explanation he gives ⁴²⁹ is followed by modern translators and commentators, so that, for example, Miss Horner translates it “to be vomitted like spittle”. Some of the Chinese versions of the Vinaya, however, do know the correct meaning of khelāsika, and tell a story, perhaps invented to explain the word, of Devadatta eating Ajātaśatru’s spittle.

Sometimes the commentators give explanations based upon individual syllables of words, e.g. bhikkhu: samsāre bhayam ikkhanatāya, “because of seeing danger in samsāra” Vism 16,21; bhagavat: vanta-gamano bhavesu, “he has rejected going in existences” Pj I 107,26*; King Pasenadi is so called because: paccāmittam parasenam jināti, “he conquers the hostile army of another” Ud-a 104,27. Since the Sanskrit form of his name was Prasenajit [Pkt Pasenai], the commentarial explanation is correct to include “he conquers”, and this makes it clear that when this explanation was composed (which might have been long before Dhammapāla) the name still had the spelling with -j-, or the tradition at least knew that the correct form was -j-.⁴³⁰

The fact that the commentary explains the syllable -di by jināti, shows that the text and the cty were transmitted separately, ⁴³¹ with neither having an effect upon the other. The reason for the -d - of the text form is not clear. It is claimed by some as a Sinhalesism, because j>d in Sinhalese. If this were the explanation, I should have expected the commentary to show the change, not the text. The disagreement between the canonical text and the commentary is probably due to the fact that they were in the keeping of different groups, i.e. the bhānakas responsible for the canonical text were not also responsible for the safe keeping of the commentary upon that text.

Other explanations are given in accordance with the type of etymologising in Sanskrit literature known as nirukti or nirvacana, which was taken over into

Buddhist literature. It often resembles folk etymology in its mode of operation, but this is to misunderstand its purpose. Quite frequently words are explained by means of others which are similar in appearance but are, in fact, in no way related to them. Such etymologies are found in the canonical texts from an early period, and are then taken over into the commentarial tradition, e.g. brāhmaṇa, based upon a root brah- meaning “to be strong”, is explained in a Dhammapada verse by means of a different root brah- “to remove”. ⁴³² A brāhmana was a remover - of his evil deeds, of course. Similarly a samana is so called because he has put to rest his evil deeds, explaining sam- (Skt śram- “to strive”) by sam( Skt sam- “to be quiet”). ⁴³³ The word rājā “king” is explained in the Dighanikāya by “he pleases others by righteousness”, explaining rāj- “to rule” by means of rañj- “to be delighted”. ⁴³⁴ The fact that a comparable etymology is given, or alluded to, several times in the Mahābhārata and Rāmāyaṇa ⁴³⁵ shows that this is not a Buddhist invention.

Some of these explanations seem ludicrous in western.eyes, and this has led to some modern commentators trying to see the joking nature of the Buddha at work in them. It has been claimed that the Buddha is here parodying the etymologies (nirukti) found in brahmanical texts. ⁴³⁶ I cannot rule out the possibility of the Buddha joking, although it is noteworthy that joking is not a well-known or highly approved Buddhist pastime, but I can see no evidence that these etymologies were really intended as jokes. Certainly in a series of etymologies of this type in the Sabhiya-sutta of the Sutta-nipāta, dealing with the words kusala, pandita, muni, etc., no joke seems to be intended, and I find it hard to understand why we should think that some etymologies are meant to be humorous, and others not.

I am reminded of talking to various people about Sanskrit, and being asked by one if Sans-script was so called because it had no script and could not be written down, and by another if Sand-script was so called because the early Indians had no writing materials and had to scratch the characters in the sand. These folketymologies seemed comical to me, but both questioners were very serious about them.

These nirvacana etymologies are sometimes of value (beside the etymological theorising they contain) because they give information about the dialect(s) in which they were first given, and such information about the earlier history of the text may not be available in other ways. ⁴³⁷ For example, to explain brāhmana by using the form bāhita shows that in the dialect in which the etymology was invented the word brāhmana had the form *bāhana, while some of the etymologies in the Sabbiya-sutta make sense only if we assume that they were first composed in a dialect in which -t - and -j - became -y-.

The fact that, despite the displacement of pādas, etc., which makes the etymologies in the Sabbiya-sutta very hard to understand, the commentary still contains the correct traditional explanation of some of the words included in the sutta, e.g. kusala and ājāniya, shows that the sutta tradition, which is very corrupt, and the commentary tradition, were quite separate, as in the case of the name Pasenadi which I have just mentioned, and the commentary went on transmitting the correct explanations even though the text which was being commented upon no longer had the correct readings. ⁴³⁸

The growth of homonyms in MIA often meant that two or more explanations were possible for a word, depending upon which etymology was being considered. This sometimes enables a Pāli commentary to give a richer and fuller exegesis than would be possible in Sanskrit, where a redactor had to decided upon one form or another. The BHS equivalent of sammappadhāna “right effort” is samyakprahāna “right abandoning”. I mentioned in the sixth lecture the suggestion ⁴³⁹ that the Sanskrit version is an incorrect backformation from an MIA form sammappahāna which might have either meaning. This view is supported by the fact that Buddhaghosa’s explanation of the term actually includes the idea of abandoning, ⁴⁴⁰ and this suggests that the commentarial material was first composed in a dialect where the form of the word was sammappahāna.

Where the commentators no longer knew the meanings of words, they sometimes had to deduce them by devising etymologies which we can see are incorrect. In Pāli the consonant cluster in the word -jña “knowing” becomes -ññ-. The consonant cluster -ny- also becomes -ññ-. When dealing with the word vadaññu, which is derived from Skt vadānya “bountiful, liberal, a munificent giver” or “eloquent, speaking kindly or agreeably, affable”, ⁴⁴¹ the commentarial tradition wrongly took the final element -ññu to be from -jña, and the word is explained as vacana-vidu, ⁴⁴² i.e. understanding vadaññu as the equivalent of vada-jña “knowing the utterance (of the Buddha)”.

Another development from Sanskrit -jña is Pāli -jina, with an epenthetic (svarabhakti) vowel evolved between -j - and -ñ-. The resultant form is identical with the word jina “conqueror”, and the commentaries consequently explain compounds ending in -jina as “conqueror”, e.g. magga-jina ⁴⁴³ “conqueror (of the defilements) by means of the road”, ⁴⁴⁴ whereas we are probably to see its meaning as “knowing the road”. The same type of explanation is given for khettajina “conquering the field”. Here we can see that -jina is certainly from -jña, because there is a Sanskrit equivalent ksetra-jña “knowing the field”. ⁴⁴⁵

We are inclined, in western philology, to believe that there is only one correct answer to a question of etymology. In India, however, there was a custom of seeing more than one meaning in any word or phrase - the so-called ślesa. So, instead of saying the meaning is either this or that, as we would do, commentators very often say that the meaning is this and that. Sometimes the meanings they give include what we would regard as the correct etymology, but sometimes they are all, from our point of view, incorrect. These are not, however, intended as western-style etymologies, but they have rather a religious, or even mystical, purpose.

So the word bhagavat is explained with the words: “The reverend one (guru) has blessings (bhagi), is a frequenter (bhaji), a partaker (bhāgi), a possessor of what has been analysed (vibhattavā), has caused abolishing (akāsi bhaggaṃ), is fortunate (bhāgyavā), has well maintained himself in being (subhāvitattano) in many ways, has gone to the end of being (bhav-anta-go), thus he is called bhagavat”. ⁴⁴⁶ These multiple explanations would not be regarded in the tradition as specifying the meaning of the word bhagavat, but as attributes or epithets of the bhagavat himself, and would form the basis of a sort of litany “the bhagavat has blessings, etc.”, i.e. they would be taken as facts, not as explanations of the term bhagavat.

Similarly with arahat, which is explained: “One is an arahat because one’s defilement are far from one (āra-kilesa)”. ⁴⁴⁷ “The bhagavat is arahanta because of remoteness (āraka), because of the destruction of his enemies (ari) and of the spokes (ara = rāga, etc.), because of his being worthy of requisites, and because of the absence of secret ill-doing (pāpakarane rahābhāvā). ⁴⁴⁸ The inclusion of an explanation based on ari is interesting because the Jain form is ari-hant, i.e. with a svarabhakti vowel -i - instead of -a-.

A number of non-Buddhist terms were taken over and applied in a Buddhist sense, e.g. the Buddha is called a yakṣa and a nāga. Since it was thought inappropriate to call the Buddha by the name of a minor supernormal being, i.e. a “snake-god”, other explanations were invented. One is a nāga because one does not perform (na + āgu), or one will not go to rebirth again (na + gam “to go”), or will not come back again (na + āgam “to come”).

Sometimes the explanations which the commentators give include forms which suggest that they were first given in some other dialect. So, among the explanations of tathāgata, we find some based upon forms with gada instead of gata, i.e. they are taken from, or through, a dialect which voiced -t - to -d- Tathāgata is explained as: “He is thus gone (tathā gata), thus come (tathā āgata), he has thus known (tathā ājānana [verbs meaning ‘go’ also mean ‘know’]), has uttered what is real (tathā gadanato)”. ⁴⁴⁹ We are also told that just as a physician overcomes snake-bite by an antidote (agada), so the tathāgata overcomes the world with the antidote of truth. ⁴⁵⁰

Commentators not always well informed, and we can see that they or the sources they were following were sometimes mistaken. For example, in the commentary on the Sutta-nipāta, ⁴⁵¹ Ādiccabandhu is said to be the name of a pratyeka-buddha, presumably because it occurs in the verses ascribed to the pratyeka-buddhas. The word is, however, used so commonly as an epithet of the Buddha that I find it hard to believe that it does not apply to him in this verse. ⁴⁵²

Ignorance about rare grammatical forms also led commentators astray. The word phalesin is a future active participle, ⁴⁵³ and means “about to fruit”. It occurs in a verse in the Theragāthā ⁴⁵⁴ where the trees are described as being in blossom and about to fruit. The commentator, unacquainted with the form, analyses phalesin as coming from phala “fruit” and esin “seeking”, and describes the trees as “seeking their fruit”. The word occurs again in a verse ⁴⁵⁵ where a foolish action is likened to a man who cuts a tree down just when it is about to fruit. Here the error in the translation “a man seeking fruit cuts a tree down” is masked by the fact that this too would be a foolish action.

Although some of the commentarial material must go back to very early times in Buddhism, perhaps to the time of the Buddha himself, nevertheless, as we know well from other religions, all texts are interpreted in the light of the age and culture of the reader, which may differ considerably from the age and culture of the original composer of the text. We have, therefore, to decide how we should interpret a canonical text. Do we interpret it in the light of our knowledge of the culture of the time, the culture of the first commentary upon it, the culture of the second commentary on it, the culture of present-day Buddhists in South or South-east Asia, or our own western culture?

An examination of the Niddesa and the explanations it gives enables us to see that the composer of that text was already interpreting the ideas of the Suttanipāta in a different way from the original hearers of the text, as far as we can judge. ⁴⁵⁶ We also have a commentary upon the Sutta-nipāta attributed to Buddhaghosa, and he too interprets the Sutta-nipāta in the light of his own culture, or at least the culture of those who wrote the material in the Mahāvihāra which he used in his work. Again there are differences of interpretation, as the commentaries tried to interpret things in their own way, although, of course, this does not necessarily mean that their interpretation differed from that of the Buddha’s contemporary audiences.

The way in which interpretations changed can be seen easily from an examination of the meaning of the phrase tasā vā thāvarā vā. ⁴⁵⁷ If we base our translation on the meaning of the related Sanskrit words, then Sanskrit trasa means “moving” = “the collective body of moving or living beings” (as opposed to sthāvara), ⁴⁵⁸ and Sanskrit sthāvara means “standing still, not moving, fixed, stable, immovable”. ⁴⁵⁹ The phrase, therefore, means “(all creatures) moving or unmoving”. By the time the Pāli commentary was written, however, the interpretation of the words had been changed. The commentator states: ⁴⁶⁰ “tasa is a synonym for those who have cravings or fears: thāvara means that they are standing still. This is a synonym for those who are rid of their cravings, i.e. arahats”. To the commentator, therefore, the phrase had taken on a Buddhist flavour: “Ordinary individuals who still had cravings, and arahats who were freed from them”. This explains why Miss Horner translated the compound tasathāvara as “those who have craving and those who have none”, ⁴⁶¹ although it is very likely that to the original hearers the meaning was far less restricted, and more on the lines of the meaning in Sanskrit.

There are many other comparable examples of development of meaning. In the Uraga-sutta of the Sutta-nipāta we find the refrain: “That monk leaves behind this shore and the far shore as a snake leaves behind its old skin”. ⁴⁶² What does the phrase “this shore and the far shore” mean in this context? In some Pāli texts “this shore” means “this existence” and “the far shore” means “nibbāna”. By the time the commentator wrote in the fifth century C.E., the idea of leaving behind the far shore in the form of nibbāna was a Mahāyāna idea, which as a Theravādin he was very reluctant to accept. To understand the phrase therefore entails the discussion of the question of whether the Mahāyāna idea could have been in existence at the time when the sutta was composed, and, if not, what “far shore” could mean in this context. My own belief is that the references is not to samsāra and to the far shore of samsāra, i.e. nibbāna, but to this world and the next, and I believe that the verse was first formulated in a situation where the author was considering this world and the afterlife, rather than the endless stream of samsāra. When the Buddhists took the verse over, however, they had to make it fit into the samsāra system. In an attempt to avoid the Mahāyāna idea, the commentary on the Sutta-nipāta gives seven explanations, ⁴⁶³ including the lower fetters and the higher fetters, and the world of men and the world of gods. ⁴⁶⁴ It is interesting to note that, in the Sanskrit Udāna-varga, where the editions of Chakravarti and Nakatani read orapāram “near and far shore”, ⁴⁶⁵ Bernhard’s edition reads apāram “this shore”, which suggests that the redactor of that version also thought that the abandoning of pāra was unacceptable.

The Buddhaghosuppatti ⁴⁶⁶ states that the commentaries given by Mahinda, and the additions thereto, were burned in a great bonfire when Buddhaghosa had finished writing his commentaries. Although this has long been recognised as being merely an exaggerated way of saying that the Sinhalese commentaries fell into disuse, there is evidence to show that the earlier commentaries were not, in fact, completely superseded. It is clear that the commentaries which Buddhaghosa used to write the historical introduction to the Samantapāsādikā were not destroyed, for they were available to the author of the Mahāvamsa and even to the writer of the Mahāvamsa-ṭikā some centuries later. ⁴⁶⁷ The same applies to the commentaries on canonical texts. The Mahāvamsa-ṭikā refers to the Sīhala-atthakathā on the Majjhima-nikāya and includes information not given in the Papañcasūdanī. ⁴⁶⁸ The tīkā on the Sāratthappakāsinī includes a variant reading for a passage in the Majjhima-nikāya, ⁴⁶⁹ which is not only not given by Buddhaghosa, ⁴⁷⁰ but which shows by its form ahunā (< adhunā) = idān’ eva that it was taken from a dialect other than Sinhalese Prakrit and Pāli, probably a North Indian Prakrit, and perhaps even from a commentary brought over by Mahinda. Nevertheless, at some unknown date the Sinhalese atthakathās must have fallen completely into disuse, and they seem now to be irretrievably lost. Only the Pāli commentaries which were based upon them are now extant.

In time, of course, the commentaries themselves became less intelligible, and required explanation. This led to the appearance of tīkā, which were in effect commentaries upon the atthakathās. Sometimes, as I have just said, these commentaries, although late, nevertheless preserved readings and explanations which are superior to anything found in earlier commentaries. In time came a further expansion of commentarial material, with anutikās, and other commentaries being written, and the production of such works has continued right up until the present time, as Buddhist commentators have used all the facilities of Indian philology to explain the meaning of the Buddhavacana, as it has been re-interpreted again and again in the light of ever-changing cultural patterns.

How are we to evaluate the commentarial tradition? It is clear that sometimes it is very useful:

(1) The commentaries sometimes explain something we could not otherwise understand. Mrs Rhys Davids, for example, paid tribute to the indispensibility of the commentary on the Therīgāthā, ⁴⁷¹ when she was translating that text. Despite all its errors, it nevertheless gives correct explanations of many passages which would otherwise be quite unintelligible. The same is probably true of almost all commentaries.

(2) Commentaries sometimes contain readings which are better than those in the canonical texts we possess. So in the descriptions of famine which occur in the Vinaya and the Samyutta-nikāya, there is a word dvīhitikā, which is explained as being made up from du “two” or “difficult” + iñita “effort” or “activity”, i.e. the word means either “there is two-fold effort, i.e. begging will be either successful or unsuccessful” or “life will be difficult”. The commentary upon the Samyutta-nikāya, however, gives a variant reading duhitikā, which it says has the same meaning as dvîhitikā. Nevertheless, if duhita is taken as the opposite of suhita “satiated”, then we can see that *duhita would mean “hungry”, and duhitika “stricken by hunger, or connected with hunger”. ⁴⁷²

(3) As I have stated, commentaries show us how Buddhist thought has developed since the time when the canonical text they are commenting on was composed.

They were not always quite so useful:

(1) Sometimes the tradition has lost the meaning, and the commentator resorts to giving several explanations, all unlikely, and all very difficult to understand. This produces the familiar cry: “I can understand the text, but not the commentary”.

(2) Sometimes the explanations are of a circular nature, e.g. “wise means possessed of wisdom”, ⁴⁷³ and are only intelligible if the meaning is already known.

(3) Sometimes the explanations are wrong, as I have already pointed out.

(4) Sometimes the commentary explanation has had an insidious effect upon the canonical text, i.e. what was originally written in the commentary was sometimes included in the text (as “glosses”), or had an effect upon the words in the text, in that the text was changed to fit the meaning given by the commentary. It has been pointed out ⁴⁷⁴ that the threefold categorisation of kamma (karma), which is found in some Theravādin canonical texts, is due to the misunderstanding of the absolutive upapajja or upapajjam. ⁴⁷⁵ This was thought to be incorrect, and it was consequentially “corrected” to the “locative” upapajje. As a result of this, what had originally been a two-fold classification, i.e. “one feels the result [of a bad deed] in the here and now or, having been reborn, in some future period” became ” … in the here and now, or in (a future) rebirth, or in some future period”. This misinterpretation seems to have come into existence in a 15 th century tīkā on the Nettippakarana, from which it was introduced into manuscripts of the Netti itself, and then into manuscripts of the Majjhima-nikāya and the Añguttara-nikāya, on which the Netti passage was based.

X - Philology and Buddhism

This is the last of the series of lectures on “A Philological Approach to Buddhism”. A number of people have asked me what the difference is between “Buddhism and Philology”, which was the title of my first lecture, and “Philology and Buddhism”, which is the title of this lecture. My answer is “In principle, none”, but that does not mean that I am going to repeat the first lecture.

If I had entitled the two lectures “Buddhism and the achievements of Philology” and “Buddhism and the future aims of Philology”, respectively, it would perhaps would have given the best indication of my intention, because that really is the point of the reversal of the order of the two words. In the first lecture I looked back at some of the achievements of philology in the field of Buddhist Studies, and looked forward to the subjects which I was going to cover in this course of lectures. In this last lecture I want to consider the subject from the opposite point of view. I want to look back over the series of lectures, to summarise what I have said, and to look forward to see what contribution philology may be expected to make in the future.

In these lectures, I have tried to show some of the results which a philological approach to Buddhism has helped to produce. One such result is a better understanding of texts, as, of course, is to be expected if there is a better understanding of the language in which the texts are written, or if it is possible to make a comparison with other versions of the texts in different languages or dialects. This, however, is not the only result. A philological approach can also lead to a better understanding of the way in which the Buddhist tradition was transmitted, how the various versions of the Buddhist canon were developed, how cultural developments had an influence on Buddhist texts, and consequently how Buddhism itself developed.

In what I have said in these lectures I have been deliberately vague, avoiding as far as possible the presentation of a plethora of Sanskrit, Pāli and Prakrit words, although, even so, some of you may think that I have given far too many. I have very rarely quoted authorities for the statements I have made, and I have never given specific references to publications. I thought that such details were out of place in a course of lectures which was intended to be of general interest, and which, I hoped, would be intelligible to those who knew little or nothing about languages or philology. They will be added when the lectures are published.

In this final lecture I want to say something about my hopes and fears for the future. I want to suggest what further contributions I think philology can make to the study of Buddhist texts and to the study of Buddhism itself, and I want to point to some of the difficulties which lie in our path, and some of the problems which must be overcome if a philological approach is to achieve the goals of which I think it is capable.

(1) To do this, I must summarise what I have said in previous lectures. I started these lectures by explaining how I personally had become involved in philology and in the study of Buddhist texts, and I said something about the light which a philological approach had already thrown on some of the major problems of vocabulary and interpretation which occur in Sanskrit, Pāli and Prakrit versions of Buddhist texts.

(2) In the next lecture I spoke about the political, economic, social and religious background to the Buddha’s teaching. I commented briefly upon the date of the Buddha, and said something about his kpatriya background. I described the evolution of his teaching from the śramana religious movement which had arisen in opposition to the brāhmana caste, and I discussed the relationship between Buddhism, Jainism and brahmanical Hinduism, and pointed to some of the vocabulary which they held in common, although frequently with changed meanings.

The work of identifying traces of other brahmanical terminology in Buddhism continues, and scholars are currently devoting time to considering the echoes of the Brāhmanas and the Upanisads which can be heard in some of the Buddha’s teachings, e.g. in the Aggaññasutta of the Dīgha-nikāya. There are certainly more features common to Buddhism and Jainism which await detection. They may be small points, such as the explanation I mentioned in the ninth lecture of the word arahant as “destroyer of enemies”, which suits the Jain spelling of the word (i.e. ari-hant), but not the form which we find in the Theravādin canon. Nevertheless, collected together, lots of small points of this nature may help to establish a significant body of evidence.

(3) In my assessment of the oral aspects of the Buddha’s teaching I mentioned some of the work which has been done to identify the similarities, and also the differences, between the oral nature of the Buddhavacana in the fifth century B.C.E. and the recitation of oral epics in modern times. Investigation is continuing into the features in the Theravādin canon which seem to be mnemonic devices, designed to help in the task of remembering canonical texts and reciting them accurately. If we can assess the significance of such devices, and if we can understand the way in which modern oral literature is recited, we shall be better able to understand the transmission system by means of which the early Buddhavacana was handed down. Consequently we shall be able to see what effect that system had on it, and it will, perhaps, be possible to deduce something about the form of the Buddhavacana before the editorial process which we associate with the bhāṇaka system was instituted.

(4) As long as our knowledge of Middle Indo-Aryan dialects is limited to the languages of the inscriptions, starting with those of Aśoka, and the literary languages, there will be gaps in our knowledge of the dialects in which the Buddhavacana was first promulgated and through which it was transmitted. Nevertheless, as I suggested in the fourth lecture, on regional dialects, careful examination of what evidence there is in the Theravādin canon, e.g. anomalous forms with unusual sound changes, will enable us to make deductions about what lies behind our earliest tangible evidence, and to postulate something about the characteristics of the dialects which were in use at a very early stage in the development of Buddhism, and also about the development of translation techniques which were used when the Buddha’s teaching was being taken to areas where different dialects were spoken.

A linguistic problem which awaits solution is the nature of the early terminology of Buddhism, and the way in which the various schools make use of terms which seem similar enough to suggest that they are based upon common ancestors, and yet different enough to defy an easy solution to the problem, e.g. apadāna/avadāna; ekabijin/ekavicika; anupādisesa/anupadhiśesa; sañghādisesa/sañghāvaśesa; patisambhidā/pratisamvid; sammappadhāna/samyakprahāna. ⁴⁷⁶ Further investigation may enable us to decide whether these variants are genuine dialect forms, incorrect back-formations, or whether they have been modified to comply with the interpretations which were handed down in the various traditions.

(5) In the lecture about Buddhism and writing, I said something about the theories which are held about the introduction of writing to India, and the effect which early Indian writing systems had on Buddhist oral literature when it was first written down. I pointed out how the deficiencies of the writing system had led to ambiguous forms, which were then interpreted in varying ways. Confirmation of the provisional dating of the recently-found material from Anuradhapura in Sri Lanka is eagerly awaited, because, if it is as old as it is claimed to be, then, providing there is enough of it, we should be able to confirm whether our conjectures about the nature of early writing are correct. If they are, then we should be better able to assess its effect upon texts.

(6) In the lecture on Sanskritisation, I gave some examples of the effect which the introduction of Sanskrit forms into the Middle Indo-Aryan dialects had upon the structure of the languages of Buddhism. I pointed out how redactors sometimes chose the wrong one of a pair of homonyms when translating from Middle Indo-Aryan into Sanskrit, and also how the necessity of making a choice deprived them of the word-plays which homonyms had made possible. As more Sanskrit and Sanskritised texts from Turfan and elsewhere are published, and as comparative studies of them are made, then we shall be better able to assess the way in which Sanskritisation was carried out. This should enable us to identify, more accurately, the precise form of the Middle Indo-Aryan versions which underlie the Sanskrit texts.

(7) When considering Aśoka and the contribution which he made to Buddhism, I gave information about what the Theravādin chronicles said about him, and also what he had said about himself in his edicts, and I suggested that some of the claims made about him were erroneous or at least overstated, in some cases because researchers had not made the distinction between Aśoka’s dhamma and the Buddha’s dhamma. The time has perhaps arrived when a new assessment of Aśoka should be made, in which philology and archaeology are linked together to assess more carefully the area over which Aśokan influence was felt, and the nature and effect of that influence.

(8) In the lecture on canonicity, I noted that this is essentially a western concept, and that it is still not clear whether there was anything which could be called a canon in early Buddhism or, if there was, when the various canons were closed. There is uncertainty about the contents of the canons of some schools, and there is a continuing debate about the status of some Theravādin texts. There is need for more work of the sort which has been done recently, i.e. the extraction from Sanskrit, Tibetan, and other sources, ⁴⁷⁷ of material which is alleged to belong to various schools of Buddhism, to see what we can deduce about the texts which non-Theravādin traditions regarded as canonical.

(9) In the ninth lecture I spoke about the commentarial tradition, and showed that it was often transmitted separately from the text being commented upon, so that the commentary sometimes kept correct readings and explanations, even when the text was corrupt. I pointed to the evidence that material, which occurred in commentaries in one tradition, was regarded as canonical in another tradition, and I suggested that at an earlier time the distinction between text and commentary was not always so clear-cut, and that at one time text and commentary were possibly of equal importance. More comparative work of this nature between commentarial and canonical texts of different schools may be expected to unearth more examples of this phenomenon, and thus shed light on the way in which the Buddha taught, and the means by which his teachings were transmitted and explained.

In particular a thorough investigation of the commentaries should be made, because in the past the need for such an investigation has tended to be overlooked. Since their purpose is interpretive and exegetical, their language is often difficult to understand. Consequently, the scribal tradition often found them problematic, with the result that the text of commentaries is frequently corrupt. Some of the Pali Text Society’s editions are, in fact, so bad that the Society is unwilling to reprint them. New editions are needed, and also translations, and critical studies, in an attempt to establish what the relationship is to the texts they comment on, and to versions in other traditions and languages. What little study has been done has centered mainly upon the explanatory portions, the etymologies, etc., but the commentaries also contain long exegetical passages, which sometimes seem irrelevant to the subject matter being commented upon. We may presume that the reason for the inclusion of these passages perhaps varied depending on the type of text, i.e. material in a Vinaya commentary would have had a different purpose from that in a commentary on a Sutta-pitaka text, and that would have differed from anything in the commentary on an Abhidhamma-pitaka text. A careful study might help to reveal just what purpose the commentators had in mind when introducing such material.

The individual subjects of these lectures were not, of course, clear-cut and separate. They were designed to be inter-dependent to some extent, at least, and they were intended to blend together to make a whole. My hope is that any future work done along the lines I have suggested will lead to a situation where:

(a) A careful investigation of the Theravādin texts, and an assessment of the philological and cultural information contained therein, will enable us to unravel the problems of the development of the Middle Indo-Aryan dialects, and as a result of this, to date, if only relatively, the composition of certain portions of early Buddhist texts. ⁴⁷⁸ This will also require us to take into consideration philological evidence available from other, non-Theravādin and non-Buddhist, sources, and to co-ordinate it with archaeological and any other evidence, which sheds light on the development and spread of Buddhism.

(b) And even more ambitiously, any conclusions to which we come as a result of this philological work will not be restricted to Buddhism, but will enable us to define more accurately the development of the religious milieu as a whole during the period between the fifth century B.C.E., and the fifth century C.E., i.e. between the time of the Buddha and the time of Buddhaghosa, and also the literature, in the very broadest sense, which was composed in that religious environment, whether Buddhist, Jain, brahmanical or secular. ⁴⁷⁹

In my first lecture I repeated the statement which I usually make to those who are just entering the field, and wonder what tasks remain to be done, and which task they personally should select as their contribution to Buddhist Studies. In this last lecture I will repeat my statement yet again: “Everything that has not been done needs to be done. Everything that has been done needs to be done again”. So everything that I have said about the achievements of philology, every suggestion I have made, and all the suggestions of other scholars, all need to be kept permanently under review to see if new evidence can help to prove or disprove what has been proposed.

So much for my hopes. What about my fears? You will have noticed that all the tasks I have just mentioned demand considerable philological skills. They all require a wide knowledge of Indian languages, and in some cases Tibetan and Chinese too. And this is the problem. Where are the philologists who can do this now, and who will train the philologists of the future?

What I should therefore like to do in this final lecture is to make a plea for a bigger and better use of philology in the field of Buddhist studies. Unfortunately, I know that for the most part my plea is destined to fall on stony ground. For a number of reasons the number of philologists capable of working in the field of Indian languages and in the field of Buddhist studies based on those languages - I will say nothing about the comparable situation in Tibetan, Chinese, Khotanese, and a host of other languages in which Buddhist texts are written is decreasing and is likely to decrease even more in the future. There are various reasons for this. One of them, perhaps the most important, is that because the number of teachers of Indian philology is small, and the number of students of philology is also small, compared with other subjeets, in any set of circumstances where the financing of teaching is in direct proportion to the number of students taught, then the amount of money available for teaching philology will decrease as the number of students decreases proportionally. Less money for teaching means less teachers, which means fewer classes, and as the numbers of classes and students continue to decrease, in comparison with other subjects, the amount of money available for teaching philology will decrease even further. This is a circle of the most vicious nature.

And this is very unfortunate, because by and large the number of those interested in Buddhism is growing and I think will grow even more now that Buddhism is one of the religions to be taught in British schools as part of Religious Education. And so I would hope that as the numbers of those studying Buddhism grow, the amount of money available for the teaching of Buddhism will increase and the number of teachers of Buddhism will increase. But if the possibility of learning the languages of Buddhism has diminished, and if the teachers of Buddhism have not been able to gain an adequate philological training in Buddhism, then I suspect that the situation that pertains at the moment, where much of the work in the field of Buddhist studies is done by those who have not had an adequate philological training in the languages required, will continue and will in fact get worse.

I suspect that what I have said in these lectures will have little effect upon some scholars, who would probably say that I am overstating the need for philological training. They would perhaps not describe themselves as philologists, but would nevertheless regard themselves as sufficiently competent in philology to be able to handle all the textual material they need, because in their books and articles about Buddhism they do include Pāli and Sanskrit terms. This, however, is not what I regard as philology. Any work on Buddhism, which makes a claim to be based on original language sources, will be worthless unless it is based upon a full understanding of those sources. This may appear to be stating the obvious, but it seems to me that much of what is alleged to be based on such sources is, in fact, not so based. Some of it seems to be based on translations, with Pāli and Sanskrit equivalents inserted in brackets, to give a veneer of scholarship.

Anyone who writes about Buddhism can sprinkle his article with Pāli and Sanskrit equivalents, but this is not the same as knowing what the words mean, or why they mean it. I say “anyone can do it”, but this is not entirely true. I have read books, published recently, by teachers in recognised academic institutions, where the Pāli and Sanskrit words in brackets do not actually coincide with the English words which are supposed to be their translation. Even if the authors have succeeded in tracking down the correct place on the page in the text, the way in which they quote the reference, with compounds wrongly divided, or case forms misunderstood, makes it clear that tsome of them do not fully understand Pāli and Sanskrit. ⁴⁸⁰

Diacritical marks are often a complete mystery to those who write about Buddhism. One gets the impression that some authors assume that they are just dots and dashes which can be added or removed at will. It sometimes appears that they are added merely to placate critics like myself, who in reviews regularly complain about their omission. ⁴⁸¹ They are inserted in such an inconsistent and incorrect way that it is obvious that the authors have little idea of what they are doing, and do not understand that diacriticals are a matter of spelling, not of aesthetics, or of whim, and result from an attempt to make the Roman alphabet cope with a sound pattern that needs an alphabet with more than 26 letters. Omitting diacritics is therefore a matter of misspelling and inserting them is a matter of correct spelling. It is not just a way of placating some pedantic academic who is a stickler for accuracy.

What is perhaps the most discouraging feature of all is that some of these publications are in fact based upon doctoral dissertations, which means either that the examiners concerned did not detect such blatant errors, or that, hardly less culpable, they did not inform the doctoral candidate of the errors, if they did spot them.

When I began to read Indian Studies more than 40 years ago, the situation in Buddhist studies seemed very simple. If you look at some of the older books on Buddhism available in English about that time you will find that everything seemed cut and dried. We knew exactly when the Buddha lived and died; we knew what he had preached and how he had preached, and we knew how his followers had recited his words and handed them on to their followers; we knew how in three joint recitations held during the two centuries or so after his death this tradition had been continued; and so on. I remember that I was advised to read E.J. Thomas’s The Life of Buddha and The History of Buddhist Thought, because those two books would tell me all I needed to know about Buddhism you will understand that I was training as a philologist, not as a Buddhologist.

In general, books about Buddhism written in English at that time tended to be based on Pāli source materials, partly because of the long history of Britain’s involvement in Burma and Sri Lanka, which had brought British missionaries and civil servants into contact with Theravāda Buddhism. It was thought that with all the Pāli canonical texts and commentaries published and most of them translated, and with the Pali-English Dictionary and other ancillary works available, all problems in Buddhism had been solved, or nearly so. There was a tendency to think that Pāli sources gave the most reliable picture of early Buddhism, being the oldest and the most complete, while Sanskrit sources, being mainly Mahāyāna, were late and suspect, with their grain of truth overlaid with a thick cover of mythology.

This view had, of course, been quite untenable since the discovery of Sanskrit Hīnayāna material in Chinese Turkestan and Gilgit, etc., and to do Thomas justice, he did take account of Sanskrit materials, and he pointed out the danger of neglecting the works of schools preserved in Sanskrit, and in Tibetan and Chinese translations from Sanskrit. As he said, such Sanskrit works needed to be analysed, no less closely than the Pāli. But Thomas wrote his books in 1927 and 1931 respectively, when comparatively little of the Hinayāna material from Chinese Turkestan had been published, and the Gilgit manuscripts had not yet been discovered. It was in the third edition of The Life of Buddha, published in 1949, that he drew attention to the fact that our knowledge of Buddhism during the previous 20 years had consisted chiefly in establishing the fact of a Sanskrit canon parallel to the Pāli, and he was able to include a reference to Waldschmidt’s analysis of the Mahāparinirvānasūtra in his Bibliography. In the preface to the second edition of The History of Buddhist Thought (published in 1951), he again referred to Waldschmidt’s analysis of the Mahāparinirvānasūtra, and its value for determining more of the actual relations between Pāli scriptures and the recensions of the Indian schools, although he did not actually include any material from Waldschmidt in that revised edition of his book. Thomas’s books have been reprinted and are still available, and are still referred to, despite the absence in them of any detailed consideration of all the texts published since they first appeared more than 60 years ago.

Outside Britain, the situation was rather different, and there was less emphasis on Pāli sources. Étienne Lamotte commanded a wide range of languages, and his Histoire du Bouddhisme Indien ⁴⁸² was based upon Sanskrit, Tibetan and Chinese as well as Pāli materials. In Germany a strong school of Sanskrit Buddhist Studies was emerging, based mainly upon the German Turfan material, which was providing material for doctorate after doctorate. Although much more of the Turfan material has been published since 1951, it has not all been translated, or if translated, it is usually into German and only rarely into English. The differences between this Sanskrit material and the Pāli tradition have still not been adequately studied, and hence it has not been adequately taken into account in any English books which deal with Buddhism in a general way. So a recent publication, intended, according to the publishers, as a textbook specifically for students of Religious or Asian Studies, deals with the four noble truths and other Buddhist doctrinal matters in their Pāli versions, without making it clear that these are not necessarily the original forms of these terms, and that in some cases the Sanskrit equivalents are perhaps nearer in form and meaning to the Buddha’s own words. ⁴⁸³

I had myself studied with Professor Sir Harold Bailey, who was well acquainted with the material from Chinese Turkestan and referred to it constantly in his lectures, but since his main orientation was towards Iranian, it was the Khotanese material which he tended to refer to most often, and the impact of this new material upon Indo-Aryan studies, and particularly the Middle Indo-Aryan studies on which I had decided I would concentrate after graduation, was somewhat less emphasised.

The biggest influence upon Middle Indo-Aryan studies came from the posthumous publication of Heinrich Lūders’ Beobachtungen ūber die Sprache des buddhistischen Urkanons by Ernst Waldschmidt in 1954, with its collections of eastern dialect forms in Pāli and Buddhist Sanskrit, including those which had been misunderstood by the later tradition, its observations on the phonological and morphological features which Lūders had identified as belonging to the Urkanon, and its comments upon such details of nominal inflection as the ablative singular and accusative plural endings in -am. As I have sometimes noted in studies I have made of various phenomena in Middle Indo-Aryan, a few examples of something can frequently be explained away piecemeal, but when dozens of examples are put forward together, it is more difficult to do this. All the constituent examples of a phenomenon have, together, a holistic effect, where the evidence as a whole is greater than all the individual parts taken singly.

Sir Harold had already referred to many of these individual features in his lectures, and I find that my student notes from his lectures on the Pāli Dhammapada are full of references to the published portions of the Udānavarga and to the Gāndhārī Dharmapada of which he had made an edition, at least of the portions which were available to him at that time. It was not, however, the individual forms which were so impressive, but the fact that all of them put together really made a coherent whole. Whether they proved that there had been an Ur-kanon or not was beside the point. What was clear was that there was evidence which could be extracted from the texts to show that there had been dialects, not attested in any literary works now extant, with characteristics which could help to explain some of the problematic forms which existed in Buddhist texts in Sanskrit, Prakrit and Pāli.

At that time, there was about half the Udānavarga available in Chakravarti’s edition of the French Pelliot material, ⁴⁸⁴ and a few other vargas that had been published separately. Some years later our knowledge of this text was immeasurably increased by the publication of Bernhard’s edition of the German manuscripts of the Udānavarga, ⁴⁸⁵ at almost the same time as John Brough published all the Gāndhārī Dharmapada material which was available, ⁴⁸⁶ thus revising and augmenting the portions of this in the French and Russian collections which had been made known more than 60 years before. Then came the discovery that there was another version of the Dharmapada available in Buddhist Sanskrit, the so-called Patna Dharmapada, photographed by Rāhula Sankrtyāyana in Tibet in the 1930’s. This disclosure was closely followed by no less than three editions of this text, ⁴⁸⁷ followed finally by an edition of all the French Udānavarga material. ⁴⁸⁸ This means that there is now enough material available to make a full-scale comparative investigation of the Indian source material of the Dharmapada genre, but despite the fact that new translations of the Pāli Dhammapada appear almost every year, few translators seem to realise the help which such a comparative study could give, or if they do realise it, they do little about it.

Other branches of Indian Buddhist textual studies are not so well provided for, but the continuing publication of material from the German Turfan collection and the on-going appearance of the Dictionary of the Turfan Sanskrit material ⁴⁸⁹ means that more and more comparative studies of suttas in the prose nikāyas can be made. Once again, few translators of Pāli suttas make such studies before producing their translations.

As a result of the publications of all these texts in Sanskrit and Prakrit, much of what has been written about the early days of Buddhism, the assembling of texts, the way in which they were transmitted, and the languages in which they existed is now, to a greater or lesser extent, incorrect or misleading. Consequently a meeting was held in Brussels in 1989 to see what should be done to bring Lamotte’s work up to date. Bizarrely, the meeting was timed to coincide with the release of an English translation of his Histoire. ⁴⁹⁰ This was a straight translation from the French, with almost no attempt, except for the Bibliography, to bring it up to date, despite the protests of some of us that the book needed revising, not translating. It seemed nonsense to me, when I was asked to act as a consultant on the section which Lamotte had devoted to the formation of the Buddhist languages, ⁴⁹¹ to be told that I must make no changes whatsoever to the portion sent to me, even though I knew well, in the light of knowledge gained since 1954, that some of it was misleading, or even incorrect.

Consequently, those who now rely on the English translation of Lamotte are making use of something which is not quite as old as Thomas’s books, but still more than 35 years out of date. Reading the books listed in the Bibliographical Supplement will, of course, help to overcome this defect, but a list of publications, most of them prefixed by “see … ”, “see now” or “new study” is not very helpful, since few of the entries give a summary of the conclusions arrived at in the new publications. The very aim at completeness in some of the entries is self-defeating. For example, there is a reference to a dozen publications about Aśoka’s Greek-Aramaic bilingual inscription from Kandahar, ⁴⁹² which have appeared since Lamotte’s Histoire was first published. This leaves the reader uncertain about where to start in his search for extra information about this inscription. In the fifteen lines taken up in listing these works, useful information could have been given about the relative value of some of these publications.

I should strike a warning note. Although comparative studies, e.g. those on the Dharmapada genre which I have just mentioned, can bring positive results, the field to be covered is so large, especially if it is extended to include Chinese and Tibetan translations, that one person can scarcely be expected to cover all of it. Such projects must therefore be joint ones, and although in theory such a solution to the problem should bring good results, this will only happen if there is complete and harmonious co-operation between all participants, and if there is give and take between the proponents of each tradition. If I choose to think that the Middle Indo-Aryan component is nearer the Buddha’s own words, and I therefore disparage everything outside that field, as being merely translations and sometimes faulty translations - of Middle Indo-Aryan originals, then I shall be the loser. If scholars working in the Buddhist Sanskrit field are content to edit their texts and make statements such as “metre incorrect”, ⁴⁹³ whenever the metre of their verses does not fit the classical Sanskrit pattern, then they will be the losers. The task of reading fragments written in a difficult script, and of finding parallels in Pāli, Tibetan or Chinese texts which will help with the reading of illegible akşaras and the understanding of problematic readings, is in itself very considerable. Nevertheless, a study of related fields might well enable those scholars to understand the problems better, to realise how the imposition of Sanskritisation upon a Middle Indo-Aryan original has produced these incorrect forms, to deduce the forms which lie behind their texts, and thus to interpret them better. If those dealing with Tibetan or Chinese believe that their traditions are more faithful transmitters of the Buddha’s teaching than any of the Indian language versions, which have in some way become corrupted, then they are the losers. There is, however, no reason why any of us should be losers. This is a situation where we can all be winners if we are willing to work together, and to stop disparaging those whose interests do not coincide exactly with our own.

I am afraid that some of you may have found this course of lectures hard going. I am sorry about that, but the fact remains that philology is hard going, which is why philologists are rather rare beasts, perhaps even an endangered species. The purpose of this course of lectures has been to show that the study of Buddhism based upon texts is not a simple matter of taking an edition and a translation and assuming that all problems have already been solved.

I will repeat myself, and say again that, in my opinion, no worthwhile original work on early Buddhism, perhaps any sort of Buddhism, can be done by anyone who does not have a good grasp of the relevant languages. And by a good grasp I mean the ability not only to say what the words mean, but also why they mean it, so that a researcher is not restricted to working with translations, but can actually refer to the original text and see if the translator has gone astray. As part of his work he should be able not only to look words up in a dictionary, but also to judge whether the meanings given there are likely to be correct. Dictionary writers are only human, and they too, like other people, are liable to make mistakes, including mistakes of judgement.

For example:

(1) There is in Pāli a word kantaka meaning “thorn”, and then, metaphorically, “obstacle”. In several Theravādin texts there is a statement that “sound (sadda) is an obstacle to the first meditation” (pathamassa jhānassa saddo kantako), where the word for “obstacle” is kantaka. There is a compound “sound or noise is an obstacle” or “having sound as an obstacle” (sadda-kantaka), and we find statements such as “meditations have sound as an obstacle” (saddakantakā jhānā). A Critical Pāli Dictionary, however, although listing the meanings “thorn” and “obstacle” for kantaka, has a separate meaning for the examples I have just given: “a sharp and annoying sound” and it compares the Sanskrit usage. ⁴⁹⁴ It does not specify what Sanskrit usage, nor is the reader given any guidance about the way in which the sentences are to be translated if this sense is adopted.

(2) Again, reference to A Critical Pāli Dictionary will show that it gives two meanings for the compound kañcana-de-piccha. On the face of it, this word should mean “having two golden tail feathers”, but the commentary explains it as “having two wings like gold”. The Dictionary entry states that the literal meaning does not make sense, but no reason is given for this statement. The editors have clearly been unable to agree about the interpretation of this word, and have therefore agreed to differ and, more surprisingly, they have agreed to let the world see their difference, even to the point of writing two separate entries, each signed by its author. ⁴⁹⁵

The two separate entries are mutually incompatible, with different word divisions proposed for the verse in which the word occurs. One editor says “pāda a must contain a finite verb”, and refers to a parallel passage in the Jātakamālā as an authority, but gives no guidance as to how the problematic compound is to be translated. The other editor gives no hint of believing that there must be a finite verb in the first pāda, and says that the verse in the Jātaka-mālā has been reformulated, and is therefore, we assume, no guide as to the original form of the verse. What is a user of A Critical Pāli Dictionary to do? Few will have the lexicographical expertise to solve the matter themselves, and, not unreasonably, will expect the editors to have solved the problem for them. They will be disappointed, because the editors would seem to have abandoned their editorial responsibility. It is unique, in my experience, for two editors of a dictionary to be unable to agree upon a meaning, or, if there is genuine doubt about the meaning, to be unable to agree upon a form of words which would make the difficulty of finding a meaning clear.

So those working in the field of Buddhist Studies must have genuine expertise, the ability to make real judgements. They must be able to tackle textual problems, and not propose emendations to texts simply in order to produce an easier reading. Still less must they give arbitrary meanings to words because the attested meanings do not fit in with their theories of what the Buddha said. Theories must be based on facts and must fit the facts. Facts must not be twisted to fit a theory.

My optimism about an increased interest in the study of Buddhism, because it is one of the religions to be taught in British schools, is perhaps outweighed by my pessimism about the way in which Buddhism will be taught. Despite all the efforts put into this project by advisory committees and bodies whose aim is the promotion of Buddhism, there will inevitably be shortcomings in the teaching. I suspect that there will be a market for school-level text-books on various religions, where opinion will be represented as fact, and everything will be set out clearly as unambiguous and undeniable. I am well aware of the fact that at school level the teaching must of necessity be somewhat cursory, but I hope that the teaching will not be so cursory as to obscure the fact that there are many problems remaining in the interpretation of quite vital aspects of Buddhism. The problematic nature of these concepts is rarely stressed in popular literature, and the views of different schools of Buddhism are rarely given, even, as I said earlier, in a book intended as a handbook for university students. ⁴⁹⁶

If you investigate the translations and interpretations given of many of the basic terms of Buddhism in such text-books, you find that what had been presented as a fact is in many cases only the writer’s opinion, said three times, of course, so that it is true. Very few things in life are clearly black or white. Most of them are some shade of grey, and this is true for Buddhism also. Even for such fundamental concepts as Buddha, nirvāna, mokṣa, āryasatya, etc., the translations which are usually given, and the interpretations which are made on the basis of those translations, are frequently misleading or, if not misleading, tell only half of the story.

It is alleged, and I have no reason to doubt the accuracy of my informants, that when teaching posts in Indian Religion are being filled in Departments of Comparative Religion in some American universities, preference is given to those who have no language ability, so that applicants who do have some philological expertise are forced to hide the fact, in order to have any hope of being appointed. What hope is there for the subject in such circumstances? Even in British universities, if there is no language component to be taught in courses on comparative religion, then the ability to teach languages will certainly not be a qualification when applying for jobs, and might even be taken to be a disqualification by an appointments committee which can see no relevance in the ability to read source materials in the original language. Such an appointments committee might perhaps fear that anyone with that ability would, if appointed, waste his time teaching this irrelevant subject.

I can only view with horror the prospect of more and more courses on Buddhism being taught to those who have no knowledge of language, by those who have no knowledge of language, so that if those being taught want to learn the language they cannot do so, because there is no one to teach them.

Some years ago I read a paper to the Buddhist Forum ⁴⁹⁷ in the School of Oriental and African Studies, in which I drew attention to the great difference to be seen between the state of affairs in the editing of Pāli texts as opposed to editions of Christian gospels in New Testament Greek. I pointed out that whereas we might have an edition of a Pāli text based upon two or three manuscripts or printed editions, or even less, New Testament scholars might be able to make use of 200 or even 2000 manuscripts and other sources. I contrasted the number of New Testament scholars who could handle the Greek of the New Testament, as well as the Greek of the Old Testament and the Hebrew of which that is a translation, with the comparatively small number of scholars capable of handling Pāli, Sanskrit and Prakrit. I was therefore surprised to find, in a recent publication by a Professor of Theology at a British university, a reference to the state of affairs in one branch of the Christian church - and by no means a moribund branch. It was stated that one of the aims of the book was “to stimulate … renewed interest in serious theological study, particularly among ordinands and clergy”, ⁴⁹⁸ where it would appear that interest was waning. There was a clear statement that academic theology at its best is concerned with the central questions, not the proliferation of marginal theories, ⁴⁹⁹ which reminded me of the marginalisation of Buddhist scholars to which I referred in my first lecture. The book also included the revelation that many ordinands were informing the author that they were giving up Greek, i.e. New Testament Greek, because they could not see its relevance to their future ministry. ⁵⁰⁰ They apparently felt no need to read the primary source for their religion in the original language. What was needed, the writer stated, was a revival of serious interest in learning the original languages in which the great treasures of the tradition had been written. It seems, then, that even in a subject so well established in Great Britain as the study of Christianity, there are problems when it comes to training people adequately in philological method. We are then not alone in having problems, but if that is the situation in a field where the facilities are infinitely greater than in the field of Indian and Buddhist Studies, then the outlook for us is bleak indeed.

The difference is that, in the field to which I have just referred, the study of the tradition has always in the past been based upon philology, and the mechanism for teaching philology still exists, and will continue to exist, for a time at least, despite the flight from philology by those newcomers to the field who do not realise its relevance. In the field of Buddhist Studies, however, in many universities the subject has never been taught on a philological basis, even if the teachers were capable of doing so, because the syllabuses and the examination system have not demanded it. The situation in the future can only get worse, because those graduating from such universities - those who will be the teachers of the future - will not have received the necessary training to remedy the situation. I am glad to be able to say that my own university still thinks that competence in language is essential for religious studies, and those reading theology and religious studies have compulsory papers in Hebrew, Greek, or Sanskrit. I hope that this will continue to be the situation, although I know that this deters some applicants for admission, who consequently choose to go to other universities, where life looks a little easier.

Nevertheless, even if knowledge of the relevant languages is not compulsory, I hope that there are those who realise that it is not satisfactory to study Buddhist texts exclusively in English translation, without being able to read anything of the teachings of the Buddha in an oriental language. If anything I have said in any of these lectures has inspired anyone to learn a little more about philology, then my time has not been wasted.

Word Index

akata 38
akusala 15,17,53
akkodhano 89
agada 162
agge 71
anga 132
Anguttara 44
acala 37
ajāta 13,38
ajje 68,72
atthakathā 77,133,134,152,153, 154,165
attaja 74,98
attā 26,27,37
atthe 74
atraja 74,98
advaitavāda 26
adhisessati 156
adhunā 165
anattā 26,27,119
anicca 26,27
anutikā 165
anudaka 108
anupadhiśeṣa 169
anupādisesa 169
anusvāra 85,86,87
anūdaka 108
anodaka 109
apagata-viññāna 156
apadāna 92,169
apāra 164
apeta-viññāna 156
abbhutadhamma 132
abhidhamma 48,137
abhidhamma-dhara 137
abhibhāsana 153
abhihāsana 153
abhūta 38
amata 13,37
amatapada 13
amrta 13
ayya 68
ara 161
arana 150
arahat 30,35,161,163
arahant 168
ari 161,162
ariya 16,52
ariya-sacca 16
ari-hant 162,168
arūpa-jhāna 31,32
arhat 35
avakada-viñaṇa 156
avadāna 169
avijjā 32, 33
asaṅkhata 38
ahimsā 117, 118
ahunā 165
ākiñcaññāyatana 30
ājāniya 160
ātmaja 74,98
ātman 26,27,28,37,39
Ādiccabandhu 162
ābhidhammika 47
āyudha 73
āraka 161
āra-kilesa 161
ārammana 67
ārāma 123,125,126
ārya 68
āryasatya 181
āryā 107
ālambana 67
āva 88
āvāsa 126
āvudha 73
āvuso 63, 73, 97
āsava 11,30,31,34,35
āsavakkhaya 31
itivuttaka 132
Indara 61
ihita 165
utpatya 84
utpāda 66
utrasta 98
udaka-baka 108
udaya-vyaya 108
udāna 132
uddāna 51,145
upaca 84
upapajja 166
upapajjam 166
upapajje 166
upari 156
upādāna 33
upāśake 115
upāsaka 114,115,119
upeca 84
upecca 83,84
upetya 83,84
uposatha 123,124,125,126,128
uppacca 84
uppatitvā 84
uppāta 65,66
uppāda 65
ṛsi 35
eka-uttiraka 47
ekabījin 169
ekavicika 169
evaṃ 67
esin 162
ogha 34
orapāra 164
kañcana-de-piccha 179
kaṇtaka 179
kaṇha 15
kattā 100
katvā 100
kappa 150
kamaradu 86
kamaramu 86
kamma 11,34,89,166
kammarata 86
kammavācā 56
kammic 33
karma 24,120,166
kāma 102
kāmarata 86
kiriyā 105
kusala 15,17,53,159,160
krtvā 100
kriyā 105
kṣana 73
kṣatriya 23,33,81,140,158,168
kṣetra-jña 161
khaṇa 73
khanti 118
khīnāsava 30, 35
khettajina 161
khelāsika 158
gata 162
gada 162
gadana 162
gātha 132
giti 107
guru 161
geyya 132
cūla 16
caitya 121
cora 153
coḷa 153
chaṇa 73
Chanda 67
chandaso 60,109,157
Channa 67
chuddha 156
chettā 100, 101
chetvā 101
janapadanirutti 149
jantāghara 63,68
Jamadaggi 69*
Jamadagni 68, 69
jarāmaraṇa 33
jātaka 132
jāti 33
jāna 108
jāva 88
Jina 35, 36
-jina 161
jināti 158
je 68,70,72,73
-jña 160, 161
jñāna 108
jhāna 7,29,31,32,179
jhānic 33
ñāya 52
țikā 134,165,166
taṇhā 33, 104
taṇhā-kappa 150
tathāgata 35,37,162
tapas 35
tasa-thāvara 163
tasā 163
tipitaka 48,78,132,133,138,146, 154
tipeṭaka 133
tipeṭakin 133
tuccha 156
tuchu 156
tepiṭaka 48,133
tosana 153
trasa 163
thambha 116
thāvarā 163
thuba 67
thūpa 67
Damila-nāḷī 155
dāna 23,34
diṭthi-kappa 150
Dīgha 44
dīpa 101
dīpa-bhāsā 151
du 165
dukkha 26,27,37,83,84
duhita 166
duhitika 166
duhitikā 165
dūta 117,118
deva 25
dvīpa 101
dvīhitikā 165,166
dhamma 11,16,27,54,55,101, 116,117,118,119,120,122, 127,129,132,152,170
dhamma- 127
dhammam 15
dhammathambha 116,117
dhammadāna 116
dhammamamgala 116
dhammamahāmātra 116,117,129
dhammayātrā 116,120,127
dhammalipi 116,117
dhammavijaya 116,118
dhammas 15,17
dhammasavana 116
dhammānuggaha 116
dhamme 74
dharma-bhāṇaka 47
dharma-vijaya 113
nāga 162
nāmarūpa 32
nikāya 44,45,46,47,49,50,78, 79,91,92,93,133,136,142, 143,145,177
nicca 26
nija 69
nijhati 117
nitya 27
nibbāna 11,13,17,28,30,31,32, 34,35,37,39,63,73,119, 120,164
nibbuta 35
nimitta 45
nimināti 84
nimisam 84
nimissam 84
nimmināti 84
nimmissaṃ 84
niya 69
niyāma 105
nirukti 158,159
nirutti 60
nirupadhi 108
nirūpadhi 108
niropadhi 109
niraupadhi 109
nirvacana 158,159,160
nirvāna 13,22,24,41,119,181
nivvāna 35
nivvua 35
nekkhamma 102
nevasaññā-nāsaññāyatana 30
naiṣkāmya 102
naiṣkramya 102
nyāma 105
pakkodano 89
pacceka 13,103,104
pacceka-buddha 36
pacceya 104
pacceya-buddha 104
paññā 31
paṭisambhidā 169
paṇḍita 159
patteya-buddha 14,36,103
pada-bhājaniya 150
pabbajjā 124
paramparā 82
para-(v)vāya 100
pari- 17,63
paritta 48
parinamati 89
parinibbāna 17
parinibbuta 17
parinirvāna 35
parinirvrta 35
pariṣat 105
parisā 105
paryāya 87
parṣat 105
pali- 63
pavāranā 123,125,126
Paseṇai 158
Pasenadi 158,160
pāda 160
pāra 164
pārājika 149
pāli 63,133,134
pāli 134,141
pāli-bhāsā 63
pāli-muttaka 140
pāsamḍā 122
pitaka 45,48,77,133,136,137, 141,142,145
pitus 61
pīti 101
pupphaṃ 15
puraḥ 71
pure 71
puro 71
purohita 23
puṣpā 15
puṣpād 15
peyyāla 87,88
peyyālams 87
prakīṇaka 132
pratisaṃvid 169
pratitya-samutpāda 32, 33
pratyaya 104
pratyeka 13,103,104
pratyeka-buddha 13,103,162
pradhāna 102,103
pravrajita 35
pravrajyā 35
Prasenajit 158
prahāna 102
prīti 101
pretya 24
phala 162
phalesin 162
phassa 32
banaka 47
bahuśruta 42
bahussuta 42
bābhana-samana 122
bāhana 103,160
bāhita 160
buddha 13,29,32,35,36,103,115, 181
Buddha-jñāna 108
Buddha-dhamma 116
Buddha-bhāsita 135
Buddha-yāna 108
Buddhavacana 44,60,63,64,71, 81,92,93,95,104,110,133, 134,135,136,141,142,145, 146,147,148,156,165,168, 169
Budhaśake 115
bodha 32
bodhi 29,30,31,32,33,105,108, 115,120,121
bodhiśakta 105
bodhiṣakta 105
bodhisatta 104,105
bodhisattva 104
brahma-carya 28
brahman 26,27,28,39
brahma-patha 28
brahma-vihāra 28
brahma-sahavyatā 28
brāhmaṇa 23,28,36,41,98,103, 122,159,160,168
brāhmaṇa-śramaṇa 122
brāhmaṇa-samaṇa 36
brūheti 98
bhagavat 158,161
bhagī 161
bhaggaṃ 161
bhajī 161
bhaya 52
bhava 33
bhavati 108
bhav-anta-go 161
bhasta 98
bhāgī 161
bhāgyavā 161
bhāṇaka 44,46,47,48,49,55,56, 59,77,78,81,82,83,90,91, 92,93,136,137,158,169
bhikkhave 63,72
bhikkhavo 63
bhikkhu 34,62,73,124,125,126, 154,157,158
bhikkhunī 63,73
bhikṣu 34
bhūnahū 73
bheda 124,125
bhoti 108
Magadha-nāḷī 155
magga-jina 161
mañgala 116
majura 70
majūla 70
Majjhima 44
majhima 47
mayūra 70
mahā 16
mahāpadesa 91,134,137
mahāmātra 115,116,128,129
Mahāvīra 35
mātikā 48,51,137
mātikā-dhara 137
Mātā 66
Māyā 66
mukta 35
muktā 105
mucati 85
muccati 85,89
mucceya 85
mucchati 89
mucyate 85
muñcati 85
muṇcati 85
muttā 105
muttu 105
muni 159
muhur 61
mūrchati 89
mūrta 105
mokkha 100,119
mokṣa 29,41,181
mora 70
yakṣa 162
Yamataggi 68,69
yātrā 116
yādeti 63
yāna 108
yāvat 88
ye 68,73
yo 149
rañj- 159
ramati 101
rahābhāva 161
rāga 161
rāj- 159
rāja 23
rājā 159
rūpa 26
rūpa-jhāna 29,31,32
lahudaṇdatā 118
lipati 85,86
lipi 67
lipī 42
lippati 85,87
libi 67
limpati 85,87
vakka 99
vagga 142
vagguposatha 124
vacana-vidu 160
vatika 84
vattika 84
vada-jña 160
vadaññu 160
vadānya 160
vayadi 81
vari 156
vā 152
vāk 99
vākya 99
vāceti 42
vikaṭa 61
vikrta 61
vijjā 33
vijñāna 92
viññāna 32,92
vinaya 92,134,147
vinaya-dhara 137
vibhattavā 161
viśva 152
viṣama 152
visama 152,156
vissa 152,156
vihāra 56,78,110,111,114,121, 125,126
vṛttika 84
vedha 134
vedanā 32
vedalla 132
veyyākaraṇa 132
veśma 152,156
veśma-dharma 152
vêssa 152
vaiśya 23,33,81
vrajati 82
vratika 84
satta 61
śa’iṣadi 156
sata 38
śayutaka 47
śānta 101
śramaṇa 29,34,35,36,37,39,41, 103,122,168
śramaṇa-brāhmaṇa 122
śrāvaka 13,103
śruti 42
śleṣa 161
Saṃyutta 44
saṃsāra 24,25,28,30,31,32,34, 36,37,38,39,41,92,104, 119,164
saṃskrta 157
sakkata-bhāsāya 157
sañkhārā 32
sañgāyana 78
sañgiti 43,47,48,52,67,78,81, 90,91,114,123,127,128, 132,135,136,139,140,141, 143,145,146
sañgītikāra 135,145
sañgitikāraka 135
sañgha 56,77,92,115,120,122, 124,125,126,128,141,145
sañghabheda 122,125,128
sañghādisesa 169
sañghāvaśeṣa 169
sacca 16
sacce 74
sañcicca 84
sañj- 105
saññāvedayitanirodha 31,32
sata 38
satī 101
satkṛta 157
sattva 105
satyam 74
sadda 179
sadda-kaṇtaka 179
saddhamma 120
santa 101
sapta 61
samagga 123,125,126
samaṇa 103,159
samaṇa-brāhmaṇa 122
samāpatti 31
Sambuddha 35
sambodhi 120,121
sammappadhāna 102,160,169
sammappahāna 102,160
samyakprahāna 102,160,169
saya- 100
sayati 82
sayissati 156
sarvajña 35
salāyataṇa 32
sallle 121
sālā 122
sāsana 113
Siddha 35, 37
siddhattha 37
siddhi 37
sibbanī 104
siva 37
sukka 15
sukha 26,27,37
Sugata 35
sutta 16,43,46,47,50,52,54,55, 88,90,92,93,104,127,132, 134, 141, 142, 143, 144, 146, 147,150,151,156,160,164, 177
sutta-dhara 137
subhāvitattano 161
subhāṣita 104
suhita 166
sūkta 104
sūtra 104
sota 104
sotaviyā 42
sotāpattiyañga 52
sōtta 104
stūpa 114,121
sthāvara 163
srotaḥ 104
svarabhakti 61,105,109,152,161, 162
hamje 68
hatthi-vatika 84
hatthi-vattika 84
hanta 67
handa 67
huveyu 105
hevaṃ 67
hveyu 105

“Buddhist Studies: The State of the Art”, University of California at Berkeley, October 28-31, 1993. ↩
Subsequently published in a slightly revised form in Religion (1994) 24, 165-72. ↩
La Loubère, 1693. ↩
See Pruitt, 1987, 121-31. ↩
Clough, 1824. ↩
Almond, 1988. ↩
Bendall, 1883. ↩
Preliminary remarks 7. ↩
O’Flaherty, 1975, 173-75 (173). ↩
api ssu nam mārisa amanussā rittam pi pattam sise nikkuijeyyum, D III 203,21-22. ↩
T.W. & C.A.F. Rhys Davids, 1921, 195. ↩
rittam pi pattan ti, bhikkhūnam patta-sadisam eva loha-pattam hoti. tam sise nikkujittam yāva gala-vātakā bhassati. atha nam majjhe ayo-khileana ākotenti, Sv 968,37-39. ↩
Walshe, 1987, 477. ↩
Brough, 1962, 195 (ad GDhp 75). ↩
e.g. in Th 1279 (sacce atthe ca dhamme ca āhu santo patitthitā), sacce is probably the locative case, while atthe and dhamme are in the nominative case. See Norman, EV I, 292. ↩
The latest views on this subject can be seen in Bechert, 1980. ↩
Th 492. ↩
Kloppenborg, 1973, 4. ↩
See Norman, 1982, 377-91 (=CP II, 210-23). ↩
See Norman, 1991C, 3-9 (=CP IV, 218-25). ↩
See Norman, 1994, 209-24. ↩
Kalupahana, 1986, 161 (ad Dhp 21). ↩
See Norman, 1991D, 1-11 (7) (=CP IV, 251-63 [261]). ↩
See Norman, 1983A, 92-106 (= CP II, 233-49). ↩
See Norman, 1989B, 29-53 (52, note 139) (= CP IV, 92-123 [ 122, note 1]). ↩
See Lūders, 1954, §§ 188-219. ↩
kanhaṃ dhammam vippahāya, sukkaṃ bhāvetha pand̄ito, Dhp 87ab. ↩
Carter & Palihawadana, 1987, 170. ↩
Kalupphana, 1986, 121. ↩
Rādhakrishnan, 1950, 87. ↩
kanhan ti, akusala-dhammaṃ, sukkan ti, kusala-dhammaṃ, Spk III 132,20-21 (ad S V 24,21). ↩
kinhe dhamme viprahāya śukre bhāvetha panditā, Patna Dharmapada 263ba; krṣnām dharmām viprahāya śuklām bhāvayata bhikgavaḅ, Udāna-v 16.14ab. ↩
yathāpi bhamaro pupphaṃ vanṇagandham aheṭhayaṃ paleti rasam ādāya, Dhp 49abc. ↩
Carter & Palihawadana, 1987, 443, note 14. ↩
See Norman, 1990B, 1-13 (=CP IV, 170-74). ↩
satañ ca dhammo na jaram upeti, santo have sabbhi pavedayanti, Dhp 151 cd . ↩
See Norman, 1983C, 36. ↩
In the Theragāthā we find that Dabba refers to himself as parinibbuto (Th 5) as does Ānanda (Th 1047). In the Therīgāthā we find the same of Ubbirī (Thī 53), and the Pañcasatā Paṭācārā (Thī 132). ↩
It is being prepared by Dr Margaret Cone, Assistant Director of Research in Pāli Lexicography, in the Faculty of Oriental Studies, University of Cambridge. ↩
PED, 738. ↩
See Gombrich’s review of Cousins (et al.), 1974, in JRAS 1977, 132-33. ↩
Note the story of Anāthapindika and the purchase of the Jetavana (Vin II 158 foll.). ↩
For an assessment of this, see Tsuchida Ryūtarō, 1991, 51-95. ↩
See MW, s.v. rājan. ↩
See Brough, 1953, 5, note 3. ↩
e.g. the Rajputs into the ksatriya caste. ↩
See Trautman, 1981, 316-30, referred to by Gombrich, 1992, 161. ↩
See Dhp 17-18. ↩
Sn 117. ↩
For the cty see Brough, 1962, 202 and Pj II 13,1 foll. ↩
See Collins, 1982, 46-47. ↩
Norman, 1977, 329. ↩
Sakka’s lifespan is said to be saṭthiñ ca vassasatasahassāni tisso ca vassakoṭiyo (Ja II 312,19-20 ). ↩
Sakko hi indo devānam aparimutto jātiyā jarāya maraṇena sokehi … aparimutto dukkhasmā ti vadāmi (A I 144,24-26). ↩
Sakko … cavi, añño Sakko nibbatti, so pi devarajjam kāretvā āyukkhayena cavi. eten’ upāyena chattimssa Sakkā cavimsu (Ja II 312,19-22). ↩
chattimsakkhattum devindo devarajjam akārayim (A IV 90,5). ↩
Sakko pana sotāpanno jāto sotāpanno va hutvā, Bhagavato purato yeva cavitvā taruna-Sakko hutvā nibbatti (Sv 732,29-31). The same episode is referred to at Dhpa III 270,15-16. ↩
devatānam hi cavamānānam attabhāvassa gatāgatatthānam nāma na paññāyati, dīpa-sikhā-gamanam viya hoti. tasmā sesa-devatā na jānimsu (Sv 732,33-34). ↩
See Norman, 1981, 19-29 (22) (= CP II, 200-209). ↩
See Thomas, 1949, 35. ↩
See Potter, 1981, 6. ↩
M I 136,6. ↩
See Norman, 1981, 22, and Gombrich, 1990, 14. ↩
Dhp 279. ↩
Mūller, 1881, 69. ↩
Rādhakrishnan, 1950, 147. ↩
Buddharakkhita, 1985, 52. ↩
Kalupahana, 1986, 139. ↩
Jayasekera, 1992, 92. ↩
Carter & Palihawadana, 1987, 312. Strangely enough they translate anattā in the gloss as “not self”. ↩
M I 141,11 . ↩
For the occurrences of brahman in the Pāli canon, see Bhattacharya, 1989. ↩
triyamānam Brahma-pathe ti catubbidhe pi brahma-vihāra-pathe, brahme va setthe phala-samāpatti-pathe samāpajjana-vasena pavattamānam. Th-a III 9,9-11 (ad Th 689). ↩
Thomas, 1949, 126. ↩
The Tevijja-sutta (D I 235-53). ↩
mettā, karunā, muditā, and upekkhā. ↩
so cattāro brahma-vihāre bhāvetvā kāyassa bhedā param maranā Brahmalokūpago ahosi, D II 196,7-8. ↩
See D I 52-59. ↩
See MacQueen, 1984, 291-307. ↩
In the Mahāsaccakasutta (M I 237-51). ↩
See C.A.F. Rhys Davids, 1909, v. ↩
M I 160-75. ↩
See Norman, 1990A, 24-35 (33, note 17) (= CP IV, 124-38 [127, note 1]). ↩
nāyam dhammo nibbidāya na virāgāya na nirodhāya na upasamāya na abhiññāya na sambodhāya na nibbānāya samvattati, M I 165,10-12 = 166,29-31. ↩
Norman, 1990A, 24-35 (26) (= CP IV, 124-38 [128]). ↩
M I 164,15 . ↩
M I 165,35 . ↩
ākiñcaññāyatanapariyosānā satta samāpattiyo mam jānāpesi, Ps III 171,22-23. ↩
Although saññāvedayita is usually translated as a dvandva compound, this is not necessarily correct. Grammatically, it could as well be taken as a tatpurusa compound, with the past participle vedayita being used as an action noun. This interpretation would depend upon the occurrence of saññā with the verb vedayati. This combination seems not to occur in the Pāli canon as we have it now, but it is possible that it existed at an earlier date, when the precise signification of technical terms had not yet been fixed. ↩
M I 165,35 . ↩
M I 175,3-4. ↩
See Schmithausen, 1981, 249 (addendum to ch. H), quoting Nagasaki. ↩
D II 72-168. ↩
D II 156,17 . ↩
D II 156,18-19. ↩
Alsdorf, 1965, 3-6, and Norman, 1989A, 393-97 (= CP IV, 264-70). ↩
Alsdorf, 1965, 4-5. ↩
Schubring, 1962, §§ 84, 97. ↩
The āsavas and the oghas are kāma “sensual pleasure”, bhava “existence”, ditthi “speculative view”, and avijjā “ignorance”. ↩
See Norman, EV I, 133-34. ↩
See Jacobi, 1884, xix. ↩
For the number of Buddhas see Gombrich, 1980, 62-72 (68), where he suggests that the number 24 was taken over by the Buddhists from the Jains. ↩
See Norman, 1983A, 92-106. ↩
Vetter (1985,74) may not be correct in translating amata as “immortality”. This translation perhaps gives the wrong impression, since the Buddha was presumably trying to gain release from samsāra, i.e. he was trying to find a state where there was no rebirth, and therefore no dying leading to rebirth. For this reason nibbāna is described as being without birth, without death, without gati, etc. ↩
See Erdosy, 1993, 40-56 (51). See also Cribb, 1985, quoted by von Hinūber (1993, 111). ↩
See Bechert, 1991A, 3-19 (10). ↩
89-77 B.C.E. according to Bechert, 1991A, 3-19 [9]. ↩
See Collins, 1992, 121-35. ↩
Sv 13,23-24; 15,2-13. ↩
See Norman, 1989B, 29-53 (33) (= CP IV, 92-123 [98]). ↩
Miln 342,1. ↩
See Norman, 1983C, 9. ↩
In particular the Sarvāstivādins and the Mūla-Sarvāstivādins; see Schmithausen, 1987. ↩
See Norman, 1989B, 33, note 26 (= CP IV, 97, note 3). ↩
See BHSD, s.vv. bhānaka and dharma-bhānaka. ↩
Tambiah, 1968, 100. ↩
Tambiah, 1968, 99. ↩
Tambiah, 1968, 100. ↩
T.W. Rhys Davids, 1881, xxiii (quoted by Tambiah, 1968, 100-101). ↩
See Norman, 1983C, 8. ↩
Allon, 1994. ↩
Cousins, 1983, 1-11. ↩
See Gethin, 1992A, 149-72. ↩
See Norman, 1983C, 126. ↩
See I.B. Horner, in the Foreword to Nārada, 1969, viii. ↩
See Oldenberg & Pischel, 1883, xiv. ↩
Sp 695-96. ↩
See Norman, 1983C, 87; and 1988, 1-27 (15) (= CP III, 225-43 [236]). ↩
na bhikkhave buddhavacanam chandaso āropetabham, Vin II 139,13-14. ↩
Burrow, 1973, 123-40. ↩ ↩²
Burrow, 1973, 125. ↩
See Elizaremkova, 1989, 1-17; and Pinault, 1989, 35-96. ↩
For other evidence see Bloomfield & Edgerton, 1932, who give evidence for voicing and unvoicing ( §§ 44-79 ), aspiration and de-aspiration ( §§ 80-124 ), ks/ts/ps > cch (§§ 183-86), y>j, and j>y(§§ 192-93), the development of -r→-a-/-i-i-u- (§§ 631-44). ↩
e.g. ānapayati, vattati, vaḍ̣hati (Mahābhāsya [ed. F. Kielhorn], I, 259). ↩
See Norman, 1989C, 369-92 (=CP IV, 46-71). ↩
e.g. Lūders, 1954. ↩
e.g. Lévi, 1912, 495-514. ↩
von Hinūber, 1991A, 183-93 (187). ↩
See Norman, 1970, 132-43 (136-37) (= CP I, 93-107 [98-100]). ↩
See Lūders, 1954, §§ 89-90. ↩
See Lūders, 1954, §§ 88. ↩
See Norman, 1980B, 173-84 (177) (= CP II, 148-61 [154]). ↩
See Norman, 1976C, 119-21 (= CP I, 240-42). ↩
See Norman, 1979A, 279-87 (282) (= CP II, 59-70 [64]). ↩
By von Hinūber, 1993, 101-13. ↩
See Norman, 1967A, 160-70 (= CP I, 47-58 [50-51)]. ↩
See Schwarzschild, 1961, 211-17 (= Collected Articles, 104-16). ↩
Brough, 1962, 249. ↩
The antiquity of the change of y- > j - is shown not only by the examples given in Vedic Variants (see Bloomfield & Edgerton, 1932), but also by the fact that at Rāmāyana 7.4.12 the etymology of yakṣa implies a form with j - (cf. Pkt jakkha). See T. Burrow, review of U.P. Shah: The Vālmīki-Rāmāyana, Vol. I, Baroda 1972, in JRAS 1974, 74. ↩
See Brough, 1980, 42. ↩
See Norman, 1980A, 61-77 (74, note 43) (=CP II, 128-47 [ 144, note 1]). ↩
See Norman, 1980A, 61-77 (65) (=CP II, 128-47 [ 133]). ↩
See Bechert, 1991A, 3-19. ↩
See Bechert, 1991A, 13. ↩
See Bechert, 1991A, 12. ↩
von Hinūber, 1993, 101-13. ↩
von Hinūber, 1993, 102. ↩
See Saksena, 1936, 713-14. ↩
von Hinūber is misled by the āgamaśabdakośa (see ūberbl § 49) into believing that this usage is only attested once. As Schwarzschild (1961) makes clear, it is quite common. For additional examples see Oberlies, 1993, 78, s.v. je. For Pāli ye after infinitives see Norman, EV II 418. ↩
D II 154,9-15. ↩
Bechert, 1991A, 13. ↩
sacce atthe ca dhamme ca āhu santo patitthitā, Th 1229 cd . ↩
In EV I 1229, I took the former interpretation; Udāna-v 8.14 takes the latter, writing satyam. ↩
See von Hinūber, 1982, 133-40 ( 140 ). ↩
89-77 B.C.E., according to Bechert, 199A1, 9. ↩
pitakattayapāliñ ca tassa atthakatham pi ca mukhapāthena ānesum pubbe bhikkhu mahāmati|| hānim disvāna sattānam tadā bhikkhu samāgatā| ciratthitattham dhammassa potthakesu likhāpayum|| Dī 20.20-21. ↩
Sp 695-96. ↩
See Norman, 1983C, 11. ↩
Bechert, 1992, 45-53 (52). ↩
von Hinūber, 1989A. ↩
Deraniyagala, 1990A, 149-68; and 1990B, 251-92. ↩
See Norman, 1976A, 35-42 (= CP II, 52-58). ↩
Initial a - and aleph, ga and gimel, and tha and theth (with the cross in the middle of theth reduced to a dot in tha). See MW, xxvii. ↩
As Tambiah (1968, 85-131) reported of a modern Thai monastery he visited. ↩
Lūders, 1954, § 105. ↩
See Brough, 1962, 218. ↩
avithidu so vayadi, GDhp 144. ↩
avisjhitah sa vrajati, Udāna-v 1.6 ↩
abbhutthito va sayati, Ja IV 494,2*. ↩
Utt 25.7 ↩
L. Alsdorf, 1962, 110-36 (134). ↩
See Bechert, 1991A, 9. ↩
See Norman, EV II, 109 (on Th1 248). ↩
Showing the palatalisation of -a→-e- before -c-. See Norman, 1976B, 328-42 (= CP I, 220-37). ↩
Nidd I 89,17-29. ↩
cf. Skt go-vratika, bidāla-vratika, cāndra-vratika. ↩
Th 32. ↩
The same confusion is found in BHS (see BHSD, s.vv. nimināti and nirminoti), e.g. at Mvu II 176,12 where Jones emends (1952, 170, note 3) nimmin- to nimin-, because Ja III 63,9,10 has nimini and nimineyya. ↩
PDhp 46. ↩
See PED, s.v. muñcati. ↩
GDhp 63. ↩
kāmarata at A IV 438,19 and kammarata at It 71,16. ↩
Brough, 1962, 79. ↩
Brough, 1962, 266-67 (ad GDhp 292). ↩
See Brough, 1954, 361. ↩
See Norman, 1989B, 29-53 (48, note 120) (= CP IV, 92-123 [ 117, note 2]). ↩
Sadd 374,29 quotes “dānam datvā ti tamcetanam pariyodāpetvā” from Tikap-a 269,20, where, however, tam is missing in Ee. We can postulate that it had been written/read as ti because of this Burmese phenomenon, and the second ti fell out by haplology. ↩
See Report for 1883, JPTS 1883, xii. ↩
e.g. Vin II 103,21 ↩
Vin II 106,4. See von Hinūber, 1991B. ↩
Dhp 71. ↩
See Morris, 1884, 92. ↩
PDhp 107. ↩
Udāna-v 9.17. ↩
See Norman, 1978, 28-47 (32) (= CP II, 30-51 [34]). ↩
See Norman, 1984-85, 1-14 (13) (= CP III, 126-36 [135]). ↩
The break between North India and Ceylon was clearly not an abrupt one, or even a complete one, for Theravādins continued to reside near Bodhgayā for some centuries. The fame of the Sinhalese commentaries was sufficiently widespread in North India to attract Buddhaghosa to Ceylon. ↩
See Norman, 1983C, 108, 110, 111. ↩
For the question of the canonicity of these texts in Burma, see the discussion referred to in the eighth lecture. ↩
Vin V 48,29. ↩
Parivāro niṭthito,Vin V 226,6-7. ↩
D II 123-26. ↩
Alagaddūpamasutta, MI 130-42. ↩
M I 256-71. ↩
tad ev’ idam viñ̃̃anam sandhāvati samsārati, anaññam, M I 256,19-20. ↩
idam mahad bhūtam anantam apāram vijñāna-ghana eva, BUp II.4.12. ↩
aññatra paccayā n’ atthi viññannassa sambhavo, M I 258,20. ↩
na tatth’ Ānanda Tathāgatassa dhammesu ācariya-mutthi, D II 100,4=S V 153,19. 55 Gombrich, 1988, 29-46. ↩
For a survey see von Hinūber, 1989B, 341-67. ↩
von Hinūber (ūberbl § 15) writes of an archaising dialect. I would rather think of an archaising scribe. The apparent resemblances to the Girnār (=G) version of the Aśokan inscriptions are not conclusive, since we face there the same problems as in Pāli. The G version represents a “translation” of an Eastern version which we may assume was sent from Aśoka’s capital Pāṭsliputra. We have no certain way of telling whether the G version accurately represents the vernacular spoke in Western India at that time, or whether it merely represents the scribe’s attempt to produce what he thought was appropriate in the circumstances. It is arguable that the scribe at G was trying to Sanskritise, to the best of his ability, what he had received. ↩
See PED, s.v. attaja. ↩
Skt brmhayati should have developed > *bumhayati > *būhayati/būheti (with the -um-/-ū- alternation), as in AMg. ↩
See von Hinūber, 1982, 133-40 for absolutives in -ttā, and Norman, 1980B, 183, note 21 , and 1985,32-35, for absolutives in -tā. I assume that the latter are m.c., but von Hinūber (ūberbl § 498) lists an absolutive in -tā. ↩
e.g. bāhita-papo ti brāhmaṇo, Dhp 388. ↩
See Norman, 1989C, 369-92 (377-79) (= CP IV, 46-71 [56-58]). ↩
There are also examples of the retention or restoration of a long vowel before a doubled consonant, which goes against the general rule of two morae. Probably a word such as dātta is a borrowing from a North-Western dialect (see Turner, 1973, 424-28 [= Collected Papers, 450-35]); Geiger, 1994, § 7 (despite Geiger, dātta is quotable from Mil 33,3 foll.). We also find a long -ā - occurring before a doubled consonant as a result of the crasis of two vowels in a compound, and also in such words as yvāssa ( = yo assa, M I 137,11) and tyāssa ( = te assa, Dhp-a I 116,20) as products of a nonhistorical sandhi process, as well as gavāssa ca (= gavā assā ca, Ja III 408,21). All these examples seem to be based upon a knowledge of Sanskrit, since they go against the general rules of MIA dialects. ↩
katvā ti … karitvā, Ja II 317,21’ (ad katvā, 317,14*); kattā kārako, Ja IV 274,9’ (ad kattā, 274,2*). ↩
chetvā, Th 1263; chettā chedako, Th-a III 199,11. ↩
dīpam kayirātha meddhāvī yam ogho nābhiktrati, Dhp 25. ↩
dvīpam karoti medhāvī, Udāna-v 4.5. ↩
Quoted by Brough, 1962, 209 (ad GDhp 111). ↩
Dhp 79. ↩
dhamme sadā ramati pandito, Dhp 205. ↩
dhammapīti ti dhammapāyako dhammam pivanto ti attho, Dhp-a II 126,15. ↩
The Gāndhārī Dharmapada is written in a dialect which retains many of the consonant groups of Skt, and there we find dhama-pridi, GDhp 224. See Brough, 1962, 244. ↩
See Sasaki, 1986, 3. ↩
Mvy 6444 “renunciation as regards desires”, quoted in BHSD. ↩
See Gethin, 1992B, 70. ↩
Gethin, 1992B, 72. ↩
As Gethin notes (1992B, 71). ↩
bāhitapāpo tti brāhmano, Dhp 388. ↩
samitatta hi pāpānam samano tti pavuccati, Dhp 265. ↩
śamitatvāt tiu pāpānām śramano hi nirucyate, Udāna-v 11.14. ↩
See Walleser, 1914, 4, note 1; von Hinūber (1994, 132, note 28) follows Mayrhofer (1976, 492, s.v. sūtram) in thinking this proposed etymology is unnecessary. ↩
See EV 1663. ↩
MW, s.v. bodhisattva. ↩
See Bollée, 1974, 27-39 (36, note 27). ↩
See von Hinūber, 1993, 113, referring to Lūders, 1940, 179-90. ↩
See DEDR 4959. ↩
See Sander, 1985, 144-60. ↩
Nakatani, 1987. ↩
Bernhard, 1965. ↩
See Bechert, 1991A and de Jong, 1974, 49-82 (53). ↩
See Norman, 1993, 113-23, where the relevant information is given. ↩
cf. udaka-vaya, GDhp 317. ↩
Dlpp 113. ↩
Karashima, 1992. ↩
Karashima, 1992, 266 and 1991, 607-43. ↩
See Matsumura, 1989, 80 (ad [26]). ↩
See Matsumura, 1989, 81 (ad [32]); 82 (ad [35]). ↩
Udāna-v 6.10. ↩
See von Hinūber, 1993, 101-13 (109). ↩
See Schmithausen, 1987, 304-81. ↩
Rahula, 1956, 54-55. ↩
Warder, 1970, chapter 8. ↩
Lamotte, 1988, 223. ↩
Mhv 5.20 . ↩
cf. Dip 6.55; Mhv 5.72. ↩
Mhv 5.79-80. ↩
Mhv 5.175 . ↩
sāsanassa dāyādo, Mhv 5.197. ↩
Mhv 5.209. ↩
Mhv 5.280 ↩
Mhv 5.189 ↩
See Norman, 1973, 68-69 (= CP I, 166-67). ↩
See Norman, 1967B, 26 (= CP I, 68-69). ↩
See Lamotte, 1988, 228. ↩
Hultzsch, 1925, xlix. ↩
Hultzsch, 1925, liii. ↩
Gombrich, 1971, 16-17 (quoted by Southwold, 1985, 30). ↩
M I 142. ↩
See Norman, GD II, xxvii. ↩
Beal, 1884, Part II, 164. ↩
See Norman, 1983B, 277-92 (291, note 82) (= CP II, 250-68 [267, note 1]). ↩
See Norman, 1983B, 288 (= CP II, 268). ↩
Mhv 5.274 ↩
Dip 7.35-41. ↩
At Dip 7.36 vagga ( < Skt vyagra) is opposed to samagga, according to PED (s.v. vagga2). Oldenberg (1879,157) translated vagguposatha correctly, and it is not clear why Law (1957-58, 183) differed from him and, by dividing the compound vaggu ( < Skt valgu) + posatha (instead of vagga + uposatha), translated “pleasant uposatha”, although this is highly inappropriate in the context. ↩
Dip 7.44-54. ↩
Dip 7.53 . ↩
Dip 7.53. It would appear that bhikkhuno is a genitive plural form (= bhikkhūnam). For genitive plural forms in -o, see Norman, 1976C, 124 (= CP I, 244). ↩
Causing schism is dealt with in the tenth sanghādīzesa rule. Anyone attempting to cause schism should be told to desist. If after three admonitions he still persists, then it is a breach of the rule. The penalty for this is laid down at Vin III 185,37-38: samgho va tassā āpattiyā parivāsam deti mūlāya patikassati mānattaṃ deti abhbett; “placing on probation, sending back to the beginning, inflicting the mānatta discipline, rehabilitation”. It appears that the schismatic bhikkhus at Kosambi needed to be re-ordained (bhedānuvattakā bhikkhū puna upasampajjeyyum, Vin II 201,1-2). ↩
See Lamotte, 1988, 253. ↩
See DPPN, s.v. Asoka. ↩
Divy 427. ↩
See Takakusu, 1896, 73. ↩
Cutler, 1994, 33. ↩
Lamotte, 1988, 172-73. ↩
Lamotte, 1988, 301. ↩
Lamotte, 1988, 562. ↩
Brough, 1962, 33. ↩
See Thomas, 1951, 276 foll. ↩
e.g. M I 133,24-25. ↩
Dip 4.18-20. ↩
Ps II 106,8 foll. See von Hinūber, 1994. ↩
Vin V 3,14 (Khemanāmo tipetaki [m.c.]). ↩
Mil 18,15 (tepitakam buddhavacanaṃ); 19,14 (thero tepitako); 90,4 ([Nāgaseno] … so pi tipetako [m.c.]) ↩
See Norman, 1983C, 26. ↩
Sp 230,28 quoting Mil 148,6 foll. (but in a slightly different form from the text as we have it). ↩
Beal, 1884, Part II, 164-65. ↩
M I 10 and II 169,13. See Collins, 1990, 89-126 (108, note 15). ↩
Quoted by Hardy, 1902, x, note 6 (on xi) and xi, note 1. ↩
Collins, 1990, 108, note 11. ↩
The cty is also found at Ja 17. ↩
von Hinūber (1977, 237-46 [244]), draws attention to: catunnam pi catasso gāthā bandhitvā pālim eva ugganhāpetvā uyyoesi, Ja VI 353,11-12: “Composing four verses, and making the four learn them, he sent them off”. ↩
D II 123-26. ↩
D II 124,10 foll. = A II 168,11 foll. See Collins, 1990, 109, note 18. ↩
āgacchantu bhaddantā imam suttantam pariyāpunissanti pur’ āyam suttanto palujjati, Vin I 140,37 foll. See Norman, 1988, 1-27 (14) (= CP III, 225-43 [236]). ↩
See Collins’ Preface to Horner and Jaini, 1985, ix. ↩
Jayawickrama, 1962, 100 (§ 20, note 1). pathamasanigitīyaṃ sangittañ ca asangittañ ca (Sp 18,3-4). ↩
The justification for the Madhupinḍakasutta of the Majjhima-nikāya (18) and other suttas is that (in that particular case) Mahākaccāna had answered in exactly the same way as the Buddha would have done, had he been asked. ↩
See Norman, GD II, xxxv. ↩
Vitandavādī pan’ āha: Kathāvathuṃ kasmā gahitam? As 3,25-26. ↩
As 5,32 . ↩
Sp 16,14-15. ↩
Nidd-a 11,14-15. ↩
See Norman, 1983C, 84-85. ↩
Kathāvatthuvisuddhiya, Ap 37,1; Kathāvatthuvisārada 550,21, both times in conjunction with Abhidhammanayañña. ↩
As 26,3. ↩
See Norman, 1983C, 9. The Digha-bhānakas put the Khuddaka-gantha into the Abhidhammapitaka, classifying the texts in the following way: Jāṭaka Mahāniddesa Cūlaniddesa Paṭisambhidāmagga Suttanipāta Dhammapada Udāna Itivuttaka Vimānavatthu Petavatthu Theragāthā Therīgāthā. The Majjhima-bhānakas, however, put these texts in the Sutta-piṭaka, together with Cariyāpiṭaka Apadāna and Buddhavamsa. Both groups of bhānakas thus omitted the Khuddakapātha from the canon. We must assume that the Digha-bhānakas closed their list of the Khuddakanikāya before Cariyāpiṭaka Apadāna and Buddhavamsa were added to it, while the Majjhima-bhānakas closed their list before the Khuddakapātha was reckoned as being canonical. ↩
See Norman, 1983C, 8-9. ↩
See Paranavitana, 1970, civ, cvi. ↩
Lévi, 1912, 495-514. ↩
Lūders, 1954. ↩
Lamotte, 1988, 558. ↩
Lamotte, 1988, 562. ↩
Dip 5.37. ↩
The break between North India and Ceylon was clearly not an abrupt one, nor even a complete one, for Theravādins continued to reside near Bodhgayā for some centuries. The fame of the Sinhalese commentaries was sufficiently widespread in North India to attract Buddhaghosa to Ceylon. ↩
See Norman, 1983C, 108, 110, 111. ↩
See Collins, 1990, 92. ↩
As Collins points out (1990, 117, note 55); cf. Mvu II 157, where the lady is called Mrgl. ↩
Bode (1909, 4-5) states that the Burmese tradition adds to the fifteen ancient texts of the Khuddaka-nikāya four other works - the Milindapañha, the Suttasangaha, the Petakopadesa and the Nettipakarana. ↩
Duroiselle, 1911, 119-21. ↩
imāni bhamte dhammapaliyāyāni: vinayasamukase aliyavasāni anāgatabhayāni munigāthā moneyazute upatizapazine e ca lāghulovāde musāvādam adhigicyu bhagavatā budhena bhāsite. There are many discussions of these titles. See Winternitz, 1933, 606 foll. ↩
For such alternative titles, see Norman, GD II, xxvi-xxvii. ↩
See also Balbir, 1993, 47-59 (53). ↩
Winternitz, 1933, 607. ↩
M 62. ↩
Balbir, 1993, 50. ↩
See Hallisey, 1990, 155-95 (162). ↩
See Luce, 1974, 125-27. ↩
Dip 5.37 ↩
“Council” is misleading as a translation of sangiti. It was a recitation carried out by the Theravādins alone. The Dīpavamsa tells us that most of the Northern Buddhists sects separated from the Theravādins in the second century after the Buddha’s death. Few of these sects are likely to have known of the sangiti, much less have regarded it as authoritative. ↩
“Closed” implies that no new texts were admitted to the canon after that time, not that the texts themselves underwent no change. The Abhidhamma texts, excluding the Dhammasangani and Kathāvatthu, probably consisted of nothing more than the mātikā lists, with a few examples to show how they should be elaborated. I assume that the Dhammasangani, since it incorporates a portion of ancient cty, is earlier than the Third Council in its present form. I accept the tradition that the Kathāvatthu was compiled by Tissa and was recited at the Third Council. ↩
e.g. Anangana-jātaka (Mvu II 271-76), Punyavanta-jātaka (III 33-41). ↩
Mvu I 2. ↩
See Norman, 1983C, 84. ↩
Mahāsangitika bhikkhū vilomam akaṃsu sāsanami, bhinditvā mūlasanigaham aññam akaṃsu saṃgaham, Dip 5.32. ↩
Dīp 5.32-35. ↩
It seems likely that at the time of the Second Council individual suttas and Jātakas still had an independent existence, and after the schism the sects were able to arrange them as they wished. It is clear that at the time of the composition of the Niddesa the Sn collection was still regarded as separate vaggas, not as a whole. This would explain why cities were made on two vaggas only. The existence of the cty on the Khaggavisāna-sutta presumably shows that this was still regarded as an individual sutta. ↩
Lamotte, 1988, 573. ↩
Brough, 1962, 43. ↩
Lamotte, 1988, 562, 140 foll. (especially 163 foll.). ↩
The author of Pj II (477,13-14) notes that there was no old commentary upon Sn 67778: avasāne gāthādvayam eva pana Mahā-athhakathāyam vinicchitapāthe n’ atthi. See Norman, 1978, 28-47 (34) (= CP II, 30-51 [36-37]). ↩
See Collins, 1990, 89: “The Pāli canon … should be seen as a product of [the Theravāda] school, as part of a strategy of legitimisation by the monks of the Mahāvihāra lineage in Sri Lanka in the early centuries of the first millennium A.D.” ↩
As Collins (1990, 116, note 55) reminds us. ↩
Collins, 1990, 111. ↩
Collins, in the Preface to Horner and Jaini, 1985, x. ↩
See Horner and Jaini, 1985, and Jaini, 1986. ↩
See Hallisey, 1993, 97-130. ↩
M III 234,34 foll. ↩
Grierson, 1926. ↩
See Vin I 101,18 foll. ↩
Vin III 23,37 foll. ↩
M III 230-37. ↩
Dip 12.9; Mhv 13.21. ↩
= M I 175-84. ↩
Mhv 14.65 . ↩
Sīhaladīpam pana ābhatā ‘tha vasinā Mahā-Mahindena thapitā Sīhalabhāsāya dīpavāsīnam atthāya (Ps I 1,23*-24* and Lamotte, 1988, 557). ↩
Ps I 1, 21* foll. ↩
See Brough, 1962, 226. ↩
Dbp 266. ↩
It is so taken by PED, s.v. vissa 1. ↩
vissan ti, visamaṃ vissa-gandham vā kāya-kammādikam dhammaṃ samādāya, Dbp-a III 393,2-4. ↩
Udāna-v 32.18 ↩
GDhp 67. ↩
Brough, 1962, 266. ↩
Mvu III 422,13*. ↩
Th 170 . ↩
Th 613. ↩
abhibhāsanan ti tosanaṃ appatisāra-hetutāya cittassādhippamodanato, Th-a II 260,2-3. ↩
abhihāsana < *abhihassana < abhiharṣana. ↩
CPD, s.v. avitakka. ↩
ese eke ekatthe same samabhāge tajjāte tañ ñevā ti, Mp I 71,13; ese eke ekatthe same sabhāge tajjāte tañ ñevā ti, II 273,16-17. ↩
ese se ekatthe same samabhāge tajjāte ti, Kv 26,20-21 et passim. ↩
See Norman, 1979A, 279-87 (281) (= CP II 59-70 [61]). ↩
Childers, 1875, x. ↩
Mhv 37.230 ↩
C.A.F. Rhys Davids, 1974, xxviii. ↩
Adikaram, 1953, 12. ↩
Adikaram, 1953, 14. ↩
Adikaram, 1953, 16. ↩
Adikaram, 1953, 9. ↩
Sp 747,23. ↩
Sp 702,23-27. ↩
e.g. Sp 697,1. ↩
Mhv 37.227 ↩
Sp 646,11. ↩
e.g. by Adikaram, 1953. ↩
Mori, 1988, 119-67 (143). ↩
Sp 519,26. ↩
Sn 677-78. ↩
avasāne gāthādvayam eva pana Mahā-atṭhakathāyam vinicchitapāthe n’ atthi, tenāvocumha “visatigāthāsū” ti, Pj II 477,13-14. This seems to be the only statement of this kind in the cty. ↩
Adikaram, 1953, 12. ↩
Brough, 1962, 192. ↩
Dhp 41. ↩
GDhp 153. ↩
chandaso āropemā ti, vedam viya sakkata-bhāsāya vācanāmaggam āropema (Sp 1214,16-17) ↩
See Brough, 1980, 35-42 (39). ↩
D I 52 foll. ↩
Vogel, 1970. ↩
pacato ti, pare dandena pilentassa (Sv 159,16-17). ↩
Vogel, 1970, 26. ↩
bālā denti, panditā ganhantī ti dasseti (Sv 166,16). ↩
Vogel, 1970, 22. ↩
See Collins, 1990, 89-126 (117, note 55). ↩
At Vin II 188,37 and Dhp-a I 140,1. See Lamotte, 1970, 107-15. ↩
See Norman, 1979B, 325-31 (329, note 23) (= CP II, 187-93 [191, note 2]). ↩
Sp 1275,17. ↩
Aggavanna has a rule to deal with this: Sadd sūtra 88 (622,24): do jassa: Pasenadi. ↩
For similar separate transmission of text and commentary, see von Hinūber, 1981, 7486 (78). ↩
bāhita-pāpo ti brāhmano, Dhp 388. ↩
samitattā hi pāpānam samano ti pavuccati, Dhp 265. ↩
dhammena pare rañjeti, D III 93,14. See Gombrich, 1992, 159-78 (174). ↩
See Hara, 1969, 489-99. ↩
See Gombrich, 1992, 174. ↩
See Norman, 1980B, 173-84 (177-78) (= CP II, 148-64 [154-55]). ↩
See Norman, 1980B, 173-84 (178) = (CP II, 148-64 [155-56]). ↩
See Gethin, 1992, 70. ↩
Gethin, 1992, 71. ↩
See MW, s.v. vadānya. ↩
P II 415 (ad Sn 487). ↩
Sn 84. ↩
Pj II 162: maggajino ti maggena sabbakilese vijitāvī. ↩
See Dhadphale, 1980, 205, and Norman, 1980B, 173-84 (183, note 6) (= CP II, 14861[149, note 3]). ↩
Pj I 107,19 foll.; translated at MRI, 116. ↩
Ps I 42,19, quoting M I 280. ↩
ārakattā arīnam arānañ ca hatattā paccayādīnam arahattā pāpa-karane rahā-bhāvā ti imehi tāva karanehi so bhagavam arahan ti veditabbo, Sv 146,10-12; cf. Vism 198,11 foll. cf. ārakattā arahā hoti, A IV 145,2; kilesārīnam hatattā arahatā (instr.sg.), Mp V 84,22. ↩
MRI 217 foll.; cf. Ud-a 131,15. ↩
See also Ud-a 132,9 for agada/āgada. ↩
Pj II 105,27. ↩
Sn 54. ↩
See Geiger, 1994, § 193A. ↩
añgārino dāni dumā bhadante phalesino chadanam vippahāya, Th 527 ab . ↩
yathā phalesī mūle tarum chettu tam eva icchasi, Th 1121 ab. ↩
See Burford, 1991, 9. ↩
Sn 146. ↩
See MW, s.v. trasa. ↩
See MW, s.v. sthāvara. ↩
tasā, sa-tanhānam sa-bhayānam c’ etam adhivacanaṃ, tittthantī ti thāvarā, pahīnatanhāgamanānam arahantānam etam adhivacanaṃ, Pj I 245 (ad Khp IX. 4= Sn 146). ↩
Horner, 1957, 290 (= M II 105). ↩
so bhikkhu jahāti orapāram urago jinnam iva tacam purānam, Sn 1-17. ↩
Pj II 12-14, discussed by Brough, 1962, 201 foll. ↩
oram manussaloko pāram devaloko, Pj II 13,9. ↩
Udāna-v 18.21 , etc. ↩
Bu-up 7. ↩
Malalasekera (1935, cix) dates Mhv-t to the eighth or ninth century. ↩
Mhv-t 193, 305. ↩
M I 255,16 (which is quoted at Spk I 13,29). ↩
Both PTS editions read ahu tañ ñeva aññatarassa. Spk-pt (Be 1961) 41,25 reads ahunā idān’ eva. (Information received from L.S. Cousins in a letter of 25 December 1973). ↩
C.A.F. Rhys Davids, 1909, xvi-xvii. ↩
von Hinūber, 1981, 74-86. ↩
panditā ti pandiccena samannāgatā, Sp 552,24. ↩
von Hinūber, 1971, 241-49. ↩
The extension of an absolutive by a nasal can be found elsewhere in Pāli. See Geiger, 1994, § 214. ↩
See Norman, 1989C, 375 (=CP IV, 53). ↩
See, for example, Skilling, 1993. ↩
von Hintiber (1993, 113): “By using first of all the invaluable evidence of the Theravāda Tipitaka and linking it to the events of general cultural history, it will not only be possible to trace the history of MIA within this general development, but, at least to a certain degree, to find even approximate dates for certain passages of early Buddhist texts”. ↩
Bechert (1991A, 19): “If we view our question in its broader ramifications, its answer will prove to be an important element in the task of elaborating an accurate understanding of the entire linguistic, literary and religious development in India during the 5th to the 1st century B.C.E.” ↩
See, for example, the review of Wiltshire, 1990 (BSOAS 55, 1, 1992, 144-45). ↩
See, for example, the review of Stargardt, 1990 (JRAS, 2, 1, 1992, 114-17). ↩
Lamotte, 1958. ↩
See the review of Harvey, 1990 (BSOAS 55, 1, 1992, 142-43). ↩
Chakravarti, 1930. ↩
Bernhard, 1965. ↩
Brough, 1962. ↩
(1) Shukla, 1979; (2) Roth, 1980; (3) Cone, 1989. ↩
Nakatani, 1987. ↩
SWTF, 1973-. ↩
Lamotte, 1988. ↩
Lamotte, 1988, 549-93. ↩
Lamotte, 1988, 743. ↩
Matsumura, 1989, 69-100 (e.g. “Pāda b unregelmāăig”, 81, note [32].2). ↩
CPD III, 58. ↩
CPD III, 29-30. ↩
Harvey, 1990. ↩
Norman, 1990C. ↩
Young, 1992, 3. ↩
Young, 1992, 16. ↩
Young, 1992, 2. ↩

🪔🪔🪔 Tidipa

Explorer

Norman - A Philological Approach To Buddhism (1997)

A Philological Approach To Buddhism

Contents

Foreword

Abbreviations

Bibliography

I - Buddhism and Philology

II - Buddhism and Its Origins

III - Buddhism and Oral Tradition

IV - Buddhism and Regional Dialects

V - Buddhism and Writing

VI - Buddhism and Sanskritisation

VII - Buddhism and Aśoka

VIII - Buddhism and Canonicity

IX - Buddhism and the Commentarial Tradition

X - Philology and Buddhism

Word Index

Graph View

Table of Contents

Backlinks

🪔🪔🪔 Tidipa

Explorer

Norman - A Philological Approach To Buddhism (1997)

A Philological Approach To Buddhism

Contents

Foreword

Abbreviations

Bibliography

I - Buddhism and Philology

II - Buddhism and Its Origins

III - Buddhism and Oral Tradition

IV - Buddhism and Regional Dialects

V - Buddhism and Writing

VI - Buddhism and Sanskritisation

VII - Buddhism and Aśoka

VIII - Buddhism and Canonicity

IX - Buddhism and the Commentarial Tradition

X - Philology and Buddhism

Word Index

Footnotes

Graph View

Table of Contents

Backlinks