The Buddhist Publication Society is an approved charity dedicated to making known the Teaching of the Buddha, which has a vital message for people of all creeds.

Founded in 1958, the BPS has published a wide variety of books and booklets covering a great range of topics. Its publications include accurate annotated translations of the Buddha’s discourses, standard reference works, as well as original contemporary expositions of Buddhist thought and practice. These works present Buddhism as it truly is — a dynamic force which has influenced receptive minds for the past 2500 years and is still as relevant today as it was when it first arose.

[25|xxiii]The history of the development of Buddhist literature seems to be marked by periods in which the received teachings and established scriptures are assimilated and consolidated and periods of mature creativity when the essence of that transmission is expressed afresh. Bhadantācariya Buddhaghosa’s Visuddhimagga is a classic text of the latter type. It represents the epitome of Pali Buddhist literature, weaving together its many strands to create this wonderful meditation manual, which even today retains the clarity it revealed when it was written.

There are occasions when people like to make much of the supposed differences in the various traditions of Buddhism that have evolved in different times and places. What I find especially encouraging about a book such as this is that it shows so clearly how much all schools of Buddhism have fundamentally in common. Within a structure based on the traditional three trainings of ethical discipline, concentration and wisdom are detailed instructions on how to take an ethical approach to life, how to meditate and calm the mind, and on the basis of those how to develop a correct understanding of reality. We find practical advice about creating an appropriate environment for meditation, the importance of developing love and compassion, and discussion of dependent origination that underlies the Buddhist view of reality. The very title of the work, the Path of Purification, refers to the essential Buddhist understanding of the basic nature of the mind as clear and aware, unobstructed by disturbing emotions. This quality is possessed by all sentient beings which all may realize if we pursue such a path.

Sometimes I am asked whether Buddhism is suitable for Westerners or not. I believe that the essence of all religions deals with basic human problems and Buddhism is no exception. As long as we continue to experience the basic human sufferings of birth, disease, old age, and death, there is no question of whether it is suitable or not as a remedy. Inner peace is the key. In that state of mind you can face difficulties with calm and reason. The teachings of love, kindness and tolerance, the conduct of non-violence, and especially the Buddhist theory that all things are relative can be a source of that inner peace.

While the essence of Buddhism does not change, superficial cultural aspects will change. But how they will change in a particular place, we cannot say. This evolves over time. When Buddhism first came from India to countries like Sri Lanka or Tibet, it gradually evolved, and in time a unique tradition arose. This is also happening in the West, and gradually Buddhism may evolve with Western culture.

Of course, what distinguishes the contemporary situation from past transmissions of Buddhism is that almost the entire array of traditions that evolved elsewhere is now accessible to anyone who is interested. And it is in such a context that I welcome this new edition of Bhikkhu Ñāṇamoli’s celebrated English translation of the Path of Purification. I offer my prayers that readers, wherever they are, may find in it advice and inspiration to develop that inner peace that will contribute to creating a happier and more peaceful world.

May 2000

[26|xxiv]Bhikkhu Ñāṇamoli’s translation of the Visuddhimagga not only makes available in fluent English this difficult and intricate classical work of Theravāda Buddhism, the high point of the commentarial era, but itself ranks as an outstanding cultural achievement perhaps unmatched by Pali Buddhist scholarship in the twentieth century. This achievement is even more remarkable in that the translator had completed the first draft within his first four years as a bhikkhu, which is also the amount of time he had been a student of Pali.

The Buddhist Publication Society first issued this work beginning in 1975, with the kind consent of the original publisher, Mr. Ānanda Semage of Colombo. This was a reprint produced by photolithographic process from the 1964 edition. The 1979 reprint was also a photolithographic reprint, with some minor corrections..

For this edition the text has been entirely recomposed, this time with the aid of the astonishing electronic typesetting equipment that has proliferated during the past few years. The text itself has not been altered except in a few places where the original translator had evidently made an oversight. However, numerous minor stylistic changes have been introduced, particularly in the lower casing of many technical terms that Ven. Ñāṇamoli had set in initial capitals and, occasionally, in the paragraphing.

Buddhist Publication Society, 1991

This fourth edition had to be retypeset again because the digital files of the previous edition, prepared “with the aid of the astonishing electronic typesetting equipment” (as mentioned in the Foreword to the Third Edition) were lost.

Like in the previous edition, the text itself has not been altered except in a few places where Ven. Ñāṇamoli had evidently made an oversight. A few minor stylistic changes have been introduced again, such as the utilisation of the Critical Pali Dictionary system of abbreviation instead of the PTS system

The BPS would like to thank John Bullitt, Ester Barias-Wolf, Michael Zoll,

Manfred Wierich and all others who helped with this project.

Buddhist Publication Society, 2010

[27|xxv]Originally I made this translation for my own instruction because the only published version was then no longer obtainable. So it was not done with any intention at all of publication; but rather it grew together out of notes made on some of the book’s passages. By the end of 1953 it had been completed, more or less, and put aside. Early in the following year a suggestion to publish it was put to me, and I eventually agreed, though not without a good deal of hesitation. Reasons for agreeing, however, seemed not entirely lacking. The only previous English version of this remarkable work had long been out of print. Justification too could in some degree be founded on the rather different angle from which this version is made.

Over a year was then spent in typing out the manuscript during which time, and since, a good deal of revision has taken place, the intention of the revision being always to propitiate the demon of inaccuracy and at the same time to make the translation perspicuous and the translator inconspicuous. Had publication been delayed, it might well have been more polished. Nevertheless the work of polishing is probably endless. Somewhere a halt must be made.

A guiding principle—the foremost, in fact—has throughout been avoidance of misrepresentation or distortion; for the ideal translation (which has yet to be made) should, like a looking glass, not discolour or blur or warp the original which it reflects. Literalness, however, on the one hand and considerations of clarity and style on the other make irreconcilable claims on a translator, who has to choose and to compromise. Vindication of his choice is sometimes difficult.

I have dealt at the end of the Introduction with some particular problems. Not, however, with all of them or completely; for the space allotted to an introduction is limited.

Much that is circumstantial has now changed since the Buddha discovered and made known his liberating doctrine 2,500 years ago, and likewise since this work was composed some nine centuries later. On the other hand, the Truth he discovered has remained untouched by all that circumstantial change. Old cosmologies give place to new; but the questions of consciousness, of pain and death, of responsibility for acts, and of what should be looked to in the scale it values as the highest of all, remain. Reasons for the perennial freshness of the Buddha’s teaching—of his handling of these questions—are several, but not least among them is its independence of any particular cosmology. Established as it is for its foundation on the self-evident insecurity of the human situation (the truth of suffering), the structure of the Four Noble Truths provides an unfailing standard of value, unique in its simplicity, its completeness and its ethical purity, by means of which any situation can be assessed and a profitable choice made.

Now I should like to make acknowledgements, as follows, to all those without whose help this translation would never have been begun, persisted with or completed. [28|xxvi]To the venerable Ñāṇatiloka Mahāthera (from whom I first learned Pali) for his most kind consent to check the draft manuscript. However, although he had actually read through the first two chapters, a long spell of illness unfortunately prevented him from continuing with this himself.

To the venerable Soma Thera for his unfailing assistance both in helping me to gain familiarity with the often difficult Pali idiom of the Commentaries and to get something of the feel—as it were, “from inside”—of Pali literature against its Indian background. Failing that, no translation would ever have been made: I cannot tell how far I have been able to express any of it in the rendering.

To the venerable Nyanaponika Thera, German pupil of the venerable Ñāṇatiloka Mahāthera, for very kindly undertaking to check the whole manuscript in detail with the venerable Ñāṇatiloka Mahāthera’s German translation (I knowing no German).

To all those with whom I have had discussions on the Dhamma, which have been many and have contributed to the clearing up of not a few unclear points.

Lastly, and what is mentioned last bears its own special emphasis, it has been an act of singular merit on the part of Mr. A. Semage, of Colombo, to undertake to publish this translation.

Island Hermitage Dodanduwa, Sri Lanka

Ñāṇamoli Bhikkhu, Vesākhamāse, 2499: May, 1956

Sinhalese script: Hewavitarne Bequest edition, Colombo.

Burmese script: Hanthawaddy Press edition, Rangoon, 1900.

Siamese script: Royal Siamese edition, Bangkok.

Latin script: Pali Text Society’s edition, London. Harvard University Press edition, Harvard Oriental Series, Vol. 41, Cambridge, Mass., 1950.

English: The Path of Purity by Pe Maung Tin, PTS, London. 3 vols., 1922–31.

German: Visuddhimagga (der Weg zur Reinheit) by Nyanatiloka, Verlag Christiani, Konstanz, 1952. Reprinted by Jhana-Verlag, Uttenbühl, 1997.

Sinhala: Visuddhimārga-mahāsannē, ed. Ratanapala Medhaṅkara et al, 2 vols., Kalutara, 1949. (Also called Parākramabāhu-sannaya. A Pali-Sinhala paraphrase composed by King Paṇḍita Parākramabāhu II in the 13th cent. CE.) Visuddhimārgaya, Sinhala translation by Paṇḍita Mātara Sri Dharmavaṃsa Sthavira, Mātara, 1953. Etc.

French: Le Chemin de la pureté, transl. by Christian Maës, Editions Fayard, Paris 2002.

Italian: Visuddhimagga: Il sentiero della purificazione, transl. of samādhi-bheda by Antonella Serena Comba, Lulu.com, Raleigh, 2008.

Buddhaghosuppatti, edited and translated into English, by J. Gray, Luzac and Co., London, 1892.

Critical Pali Dictionary (Pali-English), Vol. I (letter a), Copenhagen, 1924–48.

Cūḷavaṃsa or Minor Chronicle of Ceylon (or Mahāvaṃsa Part II), English translation by W. Geiger, PTS London.

Dīpavaṃsa (Chronicle of Ceylon), English translation by H. Oldenberg, London, 1879.

The Early History of Buddhism in Ceylon, by E. W. Adikaram, Sri Lanka, 1946.

Guide through Visuddhimagga, U. Dhammaratana, Sarnath, 1964.

History of Indian Literature, by M. Winternitz, English translation by Mrs. S. Ketkar and Miss H. Kohn, Calcutta University, 1933.

History of Pali Literature, by B.C. Law, London, 1933 (2 Vols.).

The Life and Work of Buddhaghosa, by B.C. Law, Thacker, and Spink, Calcutta and Simla, 1923.

Mahāvaṃsa or Great Chronicle of Ceylon, English translation by W. Geiger, PTS, London.

Pali-English Dictionary, Pali Text Society, London.

The Pali Literature of Ceylon, by G.P. Malalasekera, Royal Asiatic Society, London, 1928. Reprinted by BPS, Kandy, 1994.

Pali Literature and Language, by W. Geiger, English translation by Batakrishna Ghosh, Calcutta University, 1943.

Paramatthamañjūsā, Ācariya Dhammapāla, commentary to the Visuddhimagga (Visuddhimaggamahā-ṭīkā). Vidyodaya ed. in Sinhalese script, Colombo (Chapters I to XVII only). P.C. Mundyne Pitaka Press ed. in Burmese script, Rangoon, 1909 (Chapters I to XI), 1910 (Chapters XII to XXIII). Siamese ed. in Siamese script, Bangkok. Latin script edition on Chaṭṭha Saṅgāyana CDROM of Vipassana Research Institute, Igatpuri. No English translation.

Theravada Buddhism in Burma, by Niharranjan Ray, Calcutta University, 1946 (pp. 24 ff.).

Vimuttimagga, Chinese translation: Jiĕ-tu-dào-lùn by Tipiṭaka Saṅghapāla of Funan (6th cent. CE). Taishō edition at T 32, no. 1648, p. 399c–461c (Nanjio no. 1293).

The Path of Freedom (Vimuttimagga), privately circulated English translation from the Chinese by N.R.M. Ehara, V.E.P. Pulle and G.S. Prelis. Printed edition, Colombo 1961; reprinted by BPS, Kandy 1995. (Revised, BPS edition forthcoming in 2010.)

Vimuttimagga and Visuddhimagga—Comparative Study, by P.V. Bapat, Poona, 1937. (Reprinted by BPS, 2010)

All editions Pali Text Society unless otherwise stated.

Numbers in square brackets in the text thus [25] refer to the page numbers of the Pali Text Society's edition of the Pali.

Paragraph numbers on the left correspond to the paragraph numbers of the Harvard edition of the Pali.

Chapter and section headings and other numberings have been inserted for clarity.

The works of Bhadantācariya Buddhaghosa fill more than thirty volumes in the Pali Text Society’s Latin-script edition; but what is known of the writer himself is meager enough for a page or two to contain the bare facts.

Before dealing with those facts, however, and in order that they may appear oriented, it is worth while first to digress a little by noting how Pali literature falls naturally into three main historical periods. The early or classical period, which may be called the First Period, begins with the Tipiṭaka itself in the 6th century BCE and ends with the Milindapañhā about five centuries later. These works, composed in India, were brought to Sri Lanka, where they were maintained in Pali but written about in Sinhalese. By the first century CE, Sanskrit (independently of the rise of Mahayana) or a vernacular had probably quite displaced Pali as the medium of study in all the Buddhist “schools” on the Indian mainland. Literary activity in Sri Lanka declined and, it seems, fell into virtual abeyance between CE 150 and 350, as will appear below. The first Pali renascence was under way in Sri Lanka and South India by about 400 and was made viable by Bhadantācariya Buddhaghosa. This can be called the Middle Period. Many of its principal figures were Indian. It developed in several centres in the South Indian mainland and spread to Burma, and it can be said to have lasted till about the 12th century. Meanwhile the renewed literary activity again declined in Sri Lanka till it was eclipsed by the disastrous invasion of Magha in the 11th century. The second renascence, or the Third Period as it may be termed, begins in the following century with Sri Lanka’s recovery, coinciding more or less with major political changes in Burma. In Sri Lanka it lasted for several centuries and in Burma for much longer, though India about that time or soon after lost all forms of Buddhism. But this period does not concern the present purpose and is only sketched in for the sake of perspective.

The recorded facts relating from the standpoint of Sri Lanka to the rise of the Middle Period are very few, and it is worthwhile tabling them.^[1] [30|xxviii]Why did Bhadantācariya Buddhaghosa come to Sri Lanka? And why did his work become famous beyond the island’s shores? The bare facts without some interpretation will hardly answer these questions. Certainly, any interpretation must be speculative; but if this is borne in mind, some attempt (without claim for originality) may perhaps be made on the following lines.

Up till the reign of King Vaṭṭagāmaṇi Abhaya in the first century BCE the Great Monastery, founded by Asoka’s son, the Arahant Mahinda, and hitherto without a rival for the royal favour, had preserved a reputation for the saintliness of its [32|xxx]bhikkhus. The violent upsets in his reign followed by his founding of the Abhayagiri Monastery, its secession and schism, changed the whole situation at home. Sensing insecurity, the Great Monastery took the precaution to commit the Tipiṭaka for the first time to writing, doing so in the provinces away from the king’s presence. Now by about the end of the first century BCE (dates are very vague), with Sanskrit Buddhist literature just launching out upon its long era of magnificence, Sanskrit was on its way to become a language of international culture. In Sri Lanka the Great Monastery, already committed by tradition to strict orthodoxy based on Pali, had been confirmed in that attitude by the schism of its rival, which now began publicly to study the new ideas from India. In the first century BCE probably the influx of Sanskrit thought was still quite small, so that the Great Monastery could well maintain its name in Anurādhapura as the principal centre of learning by developing its ancient Tipiṭaka commentaries in Sinhalese. This might account for the shift of emphasis from practice to scholarship in King Vaṭṭagāmani’s reign. Evidence shows great activity in this latter field throughout the first century BCE, and all this material was doubtless written down too.

In the first century CE, Sanskrit Buddhism (“Hīnayāna,” and perhaps by then Mahāyāna) was growing rapidly and spreading abroad. The Abhayagiri Monastery would naturally have been busy studying and advocating some of these weighty [33|xxxi]developments while the Great Monastery had nothing new to offer: the rival was thus able, at some risk, to appear go-ahead and up-to-date while the old institution perhaps began to fall behind for want of new material, new inspiration and international connections, because its studies being restricted to the orthodox presentation in the Sinhalese language, it had already done what it could in developing Tipiṭaka learning (on the mainland Theravāda was doubtless deeper in the same predicament). Anyway we find that from the first century onwards its constructive scholarship dries up, and instead, with the reign of King Bhātika Abhaya (BCE 20–CE 9), public wrangles begin to break out between the two monasteries. This scene indeed drags on, gradually worsening through the next three centuries, almost bare as they are of illuminating information. King Vasabha’s reign (CE 66–110) seems to be the last mentioned in the Commentaries as we have them now, from which it may be assumed that soon afterwards they were closed (or no longer kept up), nothing further being added. Perhaps the Great Monastery, now living only on its past, was itself getting infected with heresies. But without speculating on the immediate reasons that induced it to let its chain of teachers lapse and to cease adding to its body of Sinhalese learning, it is enough to note that the situation went on deteriorating, further complicated by intrigues, till in Mahāsena’s reign (CE 277–304) things came to a head.

With the persecution of the Great Monastery given royal assent and the expulsion of its bhikkhus from the capital, the Abhayagiri Monastery enjoyed nine years of triumph. But the ancient institution rallied its supporters in the southern provinces and the king repented. The bhikkhus returned and the king restored the buildings, which had been stripped to adorn the rival. Still, the Great Monastery must have foreseen, after this affair, that unless it could successfully compete with Sanskrit it had small hope of holding its position. With that the only course open was to launch a drive for the rehabilitation of Pali—a drive to bring the study of that language up to a standard fit to compete with the “modern” Sanskrit in the field of international Buddhist culture: by cultivating Pali at home and abroad it could assure its position at home. It was a revolutionary project, involving the displacement of Sinhalese by Pali as the language for the study and discussion of Buddhist teachings, and the founding of a school of Pali literary composition. Earlier it would doubtless have been impracticable; but the atmosphere had changed. Though various Sanskrit non-Mahayana sects are well known to have continued to flourish all over India, there is almost nothing to show the status of the Pali language there by now. Only the Mahāvaṃsa [XXXVII.215f. quoted below] suggests that the Theravāda sect there had not only put aside but lost perhaps all of its old non-Piṭaka material dating from Asoka’s time.^[2] One may guess that the pattern of things in Sri Lanka only echoed a process that had gone much further in India. But in the [34|xxxii]island of Sri Lanka the ancient body of learning, much of it pre-Asokan, had been kept lying by, as it were maturing in its two and a half centuries of neglect, and it had now acquired a new and great potential value due to the purity of its pedigree in contrast with the welter of new original thinking. Theravāda centres of learning on the mainland were also doubtless much interested and themselves anxious for help in a repristinization.^[3]Without such cooperation there was little hope of success.

It is not known what was the first original Pali composition in this period; but the Dīpavaṃsa (dealing with historical evidence) belongs here (for it ends with Mahāsena’s reign and is quoted in the Samantapāsādikā), and quite possibly the Vimuttimagga (dealing with practice—see below) was another early attempt by the Great Monastery in this period (4th cent.) to reassert its supremacy through original Pali literary composition: there will have been others too.^[4]Of course, much of this is very conjectural. Still it is plain enough that by 400 CE a movement had begun, not confined to Sri Lanka, and that the time was ripe for the crucial work, for a Pali recension of the Sinhalese Commentaries with their unique tradition. Only the right personality, able to handle it competently, was yet lacking. That personality appeared in the first quarter of the fifth century.

Sources of information about that person fall into three groups. There are firstly the scraps contained in the prologues and epilogues to the works ascribed to him. Then there is the account given in the second part of the Sri Lankan Chronicle, the Mahāvaṃsa (or Cūḷavaṃsa as the part of it is often called), written in about the 13th century, describing occurrences placed by it in the 5th century, and, lastly, the still later Buddhaghosuppatti (15th cent.?) and other later works.

It seems still uncertain how to evaluate the old Talaing records of Burma, which may not refer to the same person (see below). India herself tells us nothing at all.

It seems worthwhile, therefore, to give a rendering here of the principal passage from the prologues and epilogues of the works ascribed to him by name; for they are few and short, and they have special authentic value as evidence. The Mahāvaṃsa account will be reproduced in full, too, since it is held to have been composed from evidence and records before its author, and to have the ring of truth behind the legends it contains. But the later works (which European scholars hold to be legendary rather than historical in what they add to the accounts already mentioned) can only be dealt with very summarily here. [35|xxxiii]The books actually ascribed to Bhadantācariya Buddhaghosa have each a “postscript” identical in form with that at the end of Chapter XXIII of the present work, mentioning the title and author by name. This can be taken to have been appended, presumably contemporaneously, by the Great Monastery (the Mahāvaṃsa) at Anurādhapura in Sri Lanka as their official seal of approval. Here is a list of the works (also listed in the modern Gandhavaṃsa and Sāsanavaṃsa with one or two discrepancies):^[5]

Commentaries to the Vinaya Piṭaka

Commentaries to the Sutta Piṭaka

Commentary to Suttanipāta

Commentaries to the Abhidhamma Piṭaka

Beyond the bare hint that he came to Sri Lanka from India his actual works tell nothing about his origins or background. He mentions “The Elder Buddhamitta with whom I formerly lived at Mayūra suttapaṭṭana” ([M-a] epil.),^[6] and “The well known Elder Jotipāla, with whom I once lived at Kañcipura and elsewhere” ([A-a] epil.).^[7]Also the “postscript” attached to the Visuddhimagga says, besides mentioning his name, that he “should be called ‘of Moraṇḍacetaka.’” ^[8] And that is all. [36|xxxiv]On coming to Sri Lanka, he went to Anurādhapura, the royal capital, and set himself to study. He seems to have lived and worked there during the whole of his stay in the island, though we do not know how long that stay lasted. To render his own words: “I learned three Sinhalese commentaries—the Mahā-aṭṭha-[kathā], Mahāpaccarī, Kuruṇḍī—from the famed elder known by the name of Buddhamitta, who has expert knowledge of the Vinaya. Set in the grounds of the Mahā Meghavana Park [in Anurādhapura] there is the Great Monastery graced by the [sapling from the] Master’s Enlightenment Tree. A constant supporter of the Community, trusting with unwavering faith in the Three Jewels, belonging to an illustrious family and known by the name of Mahānigamasāmi (Lord of the Great City), had an excellent work-room built there on its southern side accessible to the ever virtuously conducted Community of Bhikkhus. The building was beautifully appointed, agreeably endowed with cool shade and had a lavish water supply. The Vinaya Commentary was begun by me for the sake of the Elder Buddhasiri of pure virtuous behaviour while I was living there in Mahānigamasāmi’s building, and it is now complete. It was begun by me in the twentieth year of the reign of peace of the King Sirinivāsa (Of Glorious Life), the renowned and glorious guardian who has kept the whole of Lanka’s island free from trouble. It was finished in one year without mishap in a world beset by mishaps, so may all beings attain…’’ ([Vin-a] Epilogue).

Mostly it is assumed that he wrote and “published” his works one by one as authors do today. The assumption may not be correct. There is an unerring consistency throughout the system of explanation he adopts, and there are cross-references between works. This suggests that while the Visuddhimagga itself may perhaps have been composed and produced first, the others as they exist now were more likely worked over contemporaneously and all more or less finished before any one of them was given out. They may well have been given out then following the order of the books in the Tipiṭaka which they explain. So in that way it may be taken that the Vinaya Commentary came next to the Visuddhimagga; then the Commentaries on the four Nikāyas (Collections of Suttas), and after them the Abhidhamma Commentaries. Though it is not said that the Vinaya Commentary was given out first of these, still the prologue and epilogue contain the most information. The four Nikāya Commentaries all have the same basic prologue; but the Saṃyutta Nikāya Commentary inserts in its prologue a stanza referring the reader to “the two previous Collections” (i.e. the Dīgha and Majjhima Nikāyas) for explanations of the names of towns and for illustrative stories, while the Aṅguttara [37|xxxv]Nikāya Commentary replaces this stanza with another referring to “the Dīgha and Majjhima” by name for the same purpose. The point may seem laboured and even trivial, but it is not irrelevant; for if it is assumed that these works were written and “published” in some historical order of composition, one expects to find some corresponding development of thought and perhaps discovers what one’s assumption has projected upon them. The more likely assumption, based on consideration of the actual contents, is that their form and content was settled before any one of them was given out.

Sometimes it is argued that the commentaries to the Dhammapada and the Jātaka may not be by the same author because the style is different. But that fact could be accounted for by the difference in the subject matter; for these two commentaries consist mainly of popular stories, which play only a very minor role in the other works. Besides, while this author is quite inexorably consistent throughout his works in his explanations of Dhamma, he by no means always maintains that consistency in different versions of the same story in, say, different Nikāya Commentaries (compare for instance, the version of the story of Elder Tissabhūti given in the commentary to [A] 1:2.6, with that at [M-a] I 66; also the version of the story of the Elder Mahā Tissa in the [A-a], same ref., with that at [M-a] I 185). Perhaps less need for strictness was felt with such story material. And there is also another possibility. It may not unreasonably be supposed that he did not work alone, without help, and that he had competent assistants. If so, he might well have delegated the drafting of the Khuddaka Nikāya commentaries—those of the Khuddakapāṭha and Suttanipāta, Dhammapada, and the Jātaka—or part of them, supervising and completing them himself, after which the official “postscript” was appended. This assumption seems not implausible and involves less difficulties than its alternatives.^[9]These secondary commentaries may well have been composed after the others.

The full early history of the Pali Tipiṭaka and its commentaries in Sinhalese is given in the Sri Lanka Chronicle, the Dīpavaṃsa, and Mahāvaṃsa, and also in the introduction to the Vinaya Commentary. In the prologue to each of the four Nikāya Commentaries it is conveniently summarized by Bhadantācariya Buddhaghosa himself as follows: “[I shall now take] the commentary, whose object is to clarify the meaning of the subtle and most excellent Long Collection (Dīgha Nikāya) … set forth in detail by the Buddha and by his like [i.e. the Elder Sāriputta and other expounders of discourses in the Sutta Piṭaka]—the commentary that in the beginning was chanted [at the First Council] and later re-chanted [at the Second and Third], and was brought to the Sīhala Island (Sri Lanka) by the Arahant Mahinda the Great and rendered into the Sīhala tongue for the benefit of the islanders—and from that commentary I shall remove the Sīhala tongue, replacing it by the graceful language that conforms with Scripture and is purified and free from flaws. Not diverging from the standpoint of the elders residing in the Great Monastery [in Anurādhapura], who illumine the elders’ heritage and are all well [38|xxxvi]versed in exposition, and rejecting subject matter needlessly repeated, I shall make the meaning clear for the purpose of bringing contentment to good people and contributing to the long endurance of the Dhamma.”

There are references in these works to “the Ancients” (porāṇā) or “Former Teachers” (pubbācariyā) as well as to a number of Sinhalese commentaries additional to the three referred to in the quotation given earlier. The fact is plain enough that a complete body of commentary had been built up during the nine centuries or so that separate Bhadantācariya Buddhaghosa from the Buddha. A good proportion of it dated no doubt from the actual time of the Buddha himself, and this core had been added to in India (probably in Pali), and later by learned elders in Sri Lanka (in Sinhalese) as references to their pronouncements show (e.g. XII.105 and 117).

This body of material—one may guess that its volume was enormous—Bhadantācariya Buddhaghosa set himself to edit and render into Pali (the Tipiṭaka itself had been left in the original Pali). For this he had approval and express invitation (see, e.g., the epilogue to the present work, which the Elder Saṅghapāla invited him to compose). Modern critics have reproached him with lack of originality: but if we are to judge by his declared aims, originality, or to use his own phrase “advertising his own standpoint” (XVII.25), seems likely to have been one of the things he would have wished to avoid. He says, for instance, “I shall expound the comforting Path of Purification, pure in expositions, relying on the teaching of the dwellers in the Great Monastery” (I.4; see also epilogue), and again “Now, as to the entire trustworthiness (samantapāsādikatta) of this Samantapāsādika: the wise see nothing untrustworthy here when they look—in the chain of teachers, in the citations of circumstance, instance and category [in each case], in the avoidance of others’ standpoints, in the purity of [our] own standpoint, in the correctness of details, in the word-meanings, in the order of construing the text, in the exposition of the training precepts, in the use of classification by the analytical method—which is why this detailed commentary on the Vinaya … is called Samantapāsādika ([Vin-a] epilogue). And then: “The commentary on the Pātimokkha, which I began at the request of the Elder Soṇa for the purpose of removing doubts in those uncertain of the Vinaya, and which covers the whole Sinhalese commentarial system based upon the arrangement adopted by the dwellers in the Great Monastery, is finished. The whole essence of the commentary and the entire meaning of the text has been extracted and there is no sentence here that might conflict with the text or with the commentaries of the dwellers in the Great Monastery or those of the Ancients” (Pātimokkha Commentary epilogue). Such examples could be multiplied (see especially also XVII.25).

There is only one instance in the Visuddhimagga where he openly advances an opinion of his own, with the words “our preference here is this” (XIII.123). He does so once in the Majjhima Nikāya Commentary, too, saying “the point is not dealt with by the Ancients, but this is my opinion” ([M-a] I 28). The rarity of such instances and the caution expressed in them imply that he himself was disinclined to speculate and felt the need to point the fact out when he did. He actually says “one’s own opinion is the weakest authority of all and should only be accepted if it accords with the Suttas” ([D-a] 567–568). So it is likely that [39|xxxvii]he regarded what we should call original thinking as the province of the Buddha, and his own task as the fortification of that thought by coordinating the explanations of it. However, not every detail that he edited can claim direct support in the Suttas.

The following considerations lend some support to the assumptions just made. It has been pointed out^[10]that in describing in the Vinaya Commentary how the tradition had been “maintained up to the present day by the chain of teachers and pupils” ([Vin-a] 61–62) the list of teachers’ names that follows contains names only traceable down to about the middle of the 2nd century CE, but not later. Again, there appear in his works numbers of illustrative stories, all of which are set either in India or Sri Lanka. However, no single one of them can be pointed to as contemporary. Stories about India in every case where a date can be assigned are not later than Asoka (3rd cent. BCE). Many stories about Sri Lanka cannot be dated, but of those that can none seems later than the 2nd century CE. This suggests that the material which he had before him to edit and translate had been already completed and fixed more than two centuries earlier in Sri Lanka, and that the words “present day” were not used by him to refer to his own time, but were already in the material he was coordinating. This final fixing, if it is a fact, might have been the aftermath of the decision taken in Sri Lanka in the first century BCE to commit the Pali Tipiṭaka to writing.

Something now needs to be said about the relation of the Visuddhimagga to the other books. This author’s work is characterized by relentless accuracy, consistency, and fluency of erudition, and much dominated by formalism. Not only is this formalism evident in the elaborate pattern of the Visuddhimagga but also that work’s relationship to the others is governed by it. The Visuddhimagga itself extracts from the Tipiṭaka all the central doctrines that pivot upon the Four Noble Truths, presenting them as a coherent systematic whole by way of quotation and explanation interspersed with treatises on subjects of more or less relative importance, all being welded into an intricate edifice. The work can thus stand alone. But the aim of the commentaries to the four main Nikāyas or Collections of Suttas is to explain the subject matter of individual discourses and, as well, certain topics and special doctrines not dealt with in the Visuddhimagga (many passages commenting on identical material in the Suttas in different Nikāyas are reproduced verbatim in each commentary, and elsewhere, e.g., MN 10, cf. DN 22, Satipaṭṭhāna Vibhaṅga, etc., etc., and respective commentaries). But these commentaries always refer the reader to the Visuddhimagga for explanations of the central doctrines. And though the Vinaya and Abhidhamma (commentaries are less closely bound to the Visuddhimagga, still they too either refer the reader to it or reproduce large blocks of it. The author himself says: “The treatises on virtue and on the ascetic’s rules, all the meditation subjects, the details of the attainments of the jhānas, together with the directions for each temperament, all the various kinds of direct-knowledge, the exposition of the definition of understanding, the aggregates, elements, bases, and faculties, the Four Noble Truths, the explanation [40|xxxviii]of the structure of conditions (dependent origination), and lastly the development of insight, by methods that are purified and sure and not divergent from Scripture—since these things have already been quite clearly stated in the Visuddhimagga I shall no more dwell upon them here; for the Visuddhimagga stands between and in the midst of all four Collections (Nikāyas) and will clarify the meaning of such things stated therein. It was made in that way: take it therefore along with this same commentary and know the meaning of the Long Collection (Dīgha Nikāya)” (prologue to the four Nikāyas).

This is all that can, without unsafe inferences, be gleaned of Bhadantācariya Buddhaghosa himself from his own works (but see below). Now, there is the Mahāvaṃsa account. The composition of the second part (often called Cūḷavaṃsa) of that historical poem is attributed to an Elder Dhammakitti, who lived in or about the thirteenth century. Here is a translation of the relevant passage:

“There was a Brahman student who was born near the site of the Enlightenment Tree. He was acquainted with the arts and accomplishments of the sciences and was qualified in the Vedas. He was well versed in what he knew and unhesitant over any phrase. Being interested in doctrines, he wandered over Jambudīpa (India) engaging in disputation.

“He came to a certain monastery, and there in the night he recited Pātañjali’s system with each phrase complete and well rounded. The senior elder there, Revata by name, recognized, ‘This is a being of great understanding who ought to be tamed.’ He said, ‘Who is that braying the ass’s bray?’ The other asked, ‘What, then, do you know the meaning of the ass’s bray?’ The elder answered, ‘I know it,’ and he then not only expounded it himself, but explained each statement in the proper way and also pointed out contradictions. The other then urged him, ‘Now expound your own doctrine,’ and the elder repeated a text from the Abhidhamma, but the visitor could not solve its meaning. He asked, ‘Whose system is this?’ and the elder replied, ‘It is the Enlightened One’s system.’ ‘Give it to me,’ he said, but the elder answered, ‘You will have to take the going forth into homelessness.’ So he took the going forth, since he was interested in the system, and he learned the three Piṭakas, after which he believed, ‘This is the only way’ ([M] I 55). Because his speech (ghosa) was profound (voice was deep) like that of the Enlightened One (Buddha) they called him Buddhaghosa, so that like the Enlightened One he might be voiced over the surface of the earth.

“He prepared a treatise there called Ñāṇodaya, and then the Atthasālinī, a commentary on the Dhammasaṅgaṇī. Next he began work on a commentary to the Paritta.^[11] When the Elder Revata saw that, he said, ‘Here only the text has been preserved. There is no commentary here, and likewise no Teachers’ Doctrine; for that has been allowed to go to pieces and is no longer known. However, a Sinhalese commentary still exists, which is pure. It was rendered into the Sinhalese tongue by the learned Mahinda with proper regard for the[41|xxxix] way of commenting that was handed down by the three Councils as taught by the Enlightened One and inculcated by Sāriputta and others. Go there, and after you have learnt it translate it into the language of the Magadhans. That will bring benefit to the whole world.’ As soon as this was said, he made up his mind to set out.

“He came from there to this island in the reign of this king (Mahānāma). He came to the (Great Monastery, the monastery of all true men. There he stayed in a large workroom, and he learnt the whole Sinhalese Commentary of the Elders’ Doctrine (theravāda) under Saṅghapāla.^[12] He decided, ‘This alone is the intention of the Dhamma’s Lord.’ So he assembled the Community there and asked, ‘Give me all the books to make a commentary.’ Then in order to test him the Community gave him two stanzas, saying ‘Show your ability with these; when we have seen that you have it, we will give you all the books.’ On that text alone he summarized the three Piṭakas together with the Commentary as an epitome, which was named the Path of Purification (Visuddhimagga). Then, in the precincts of the (sapling of the) Enlightenment Tree (in Anurādhapura), he assembled the Community expert in the Fully Enlightened One’s system, and he began to read it out. In order to demonstrate his skill to the multitude deities hid the book, and he was obliged to prepare it a second time, and again a third time. When the book was brought for the third time to be read out, the gods replaced the other two copies with it. Then the bhikkhus read out the three copies together, and it was found that there was no difference between the three in either the chapters or the meaning or the order of the material or the phrases and syllables of the Theravāda texts. With that the Community applauded in high delight and again and again it was said, ‘Surely this is (the Bodhisatta) Metteyya.’ “They gave him the books of the three Piṭakas together with the Commentary. Then, while staying undisturbed in the Library Monastery, he translated the Sinhalese Commentary into the Magadhan language, the root-speech of all, by which he brought benefit to beings of all tongues. The teachers of the Elders’ Tradition accepted it as equal in authority with the texts themselves. Then, when the tasks to be done were finished, he went back to Jambudīpa to pay homage to the Great Enlightenment Tree.

“And when Mahānāma had enjoyed twenty-two years’ reign upon earth and had performed a variety of meritorious works, he passed on according to his deeds”—([Mhv] XXXVII.215–47).

King Mahānāma is identified with the “King Sirinivāsa” and the “King Sirikuḍḍa” mentioned respectively in the epilogues to the Vinaya and Dhammapada Commentaries. There is no trace, and no other mention anywhere, of the Ñāṇodaya. The Atthasālinī described as composed in India could not be the version extant today, which cites the Sri Lankan Commentaries and refers to the Visuddhimagga; it will have been revised later.

The prologues and epilogues of this author’s works are the only instances in which we can be sure that he is speaking of his own experience and not only simply editing; and while they point only to his residence in South India, they neither [42|xl]confute nor confirm the Mahāvaṃsa statement that he was born in Magadha (see note 8). The Sri Lankan Chronicles survived the historical criticism to which they were subjected in the last hundred years. The independent evidence that could be brought to bear supported them, and Western scholars ended by pronouncing them reliable in essentials. The account just quoted is considered to be based on historical fact even if it contains legendary matter.

It is not possible to make use of the body of Bhadantācariya Buddhaghosa’s works to test the Mahāvaṃsa’s claim that he was a learned Brahman from central India, and so on. It has been shown already how the presumption is always, where the contrary is not explicitly stated, that he is editing and translating material placed before him rather than displaying his own private knowledge, experience and opinions. And so it would be a critical mistake to use any such passage in his work for assessing his personal traits; for in them it is, pretty certainly, not him we are dealing with at all but people who lived three or more centuries earlier. Those passages probably tell us merely that he was a scrupulously accurate and conscientious editor. His geographical descriptions are translations, not eyewitness accounts. Then such a sutta passage as that commented on in Chapter I, 86–97 of the present work, which is a part of a sutta used by bhikkhus for daily reflection on the four requisites of the life of a bhikkhu, is certain to have been fully commented on from the earliest times, so that it would be just such a critical mistake to infer from this comment anything about his abilities as an original commentator, or anything else of a personal nature about him or his own past experience.^[13]And again, the controversial subject of the origin of the Brahman caste (see [M-a] II 418) must have been fully explained from the Buddhist standpoint from the very start. If then that account disagrees with Brahmanical lore—and it would be odd, all things considered, if it did not—there is no justification for concluding on those grounds that the author of the Visuddhimagga was not of Brahman origin and that the Mahāvaṃsa is wrong. What does indeed seem improbable is that the authorities of the Great Monastery, resolutely committed to oppose unorthodoxy, would have given him a free hand to “correct” their traditions to accord with Brahmanical texts or with other alien sources, even if he had so wished. Again, the fact that there are allusions to extraneous, non-Buddhist literature (e.g. VII.58; XVI.4 n.2; XVI.85, etc.) hardly affects this issue because they too can have been already in the [43|xli]material he was editing or supplied to him by the elders with whom he was working. What might repay careful study are perhaps those things, such as certain Mahayana teachings and names, as well as much Brahmanical philosophy, which he ignores though he must have known about them. This ignoring cannot safely be ascribed to ignorance unless we are sure it was not dictated by policy; and we are not sure at all. His silences (in contrast to the author of the Paramatthamañjūsā) are sometimes notable in this respect.

The “popular novel” called Buddhaghosuppatti, which was composed in Burma by an elder called Mahāmaṅgala, perhaps as early as the 15th century, is less dependable. But a survey without some account of it would be incomplete. So here is a précis:

Near the Bodhi Tree at Gayā there was a town called Ghosa. Its ruler had a Brahman chaplain called Kesi married to a wife called Kesinī. An elder bhikkhu, who was a friend of Kesi, used to wonder, when the Buddha’s teaching was recited in Sinhalese, and people did not therefore understand it, who would be able to translate it into Magadhan (Pāḷi). He saw that there was the son of a deity living in the Tāvatiṃsa heaven, whose name was Ghosa and who was capable of doing it. This deity was persuaded to be reborn in the human world as the son of the Brahman Kesi. He learnt the Vedas. One day he sat down in a place sacred to Vishnu and ate peas. Brahmans angrily rebuked him, but he uttered a stanza, “The pea itself is Vishnu; who is there called Vishnu? And how shall I know which is Vishnu?” and no one could answer him. Then one day while Kesi was instructing the town’s ruler in the Vedas a certain passage puzzled him, but Ghosa wrote down the explanations on a palm leaf, which was found later by his father—(Chapter I).

Once when the elder bhikkhu was invited to Kesi’s house for a meal Ghosa’s mat was given to him to sit on. Ghosa was furious and abused the elder. Then he asked him if he knew the Vedas and any other system. The elder gave a recitation from the Vedas. Then Ghosa asked him for his own system, whereupon the elder expounded the first triad of the Abhidhamma schedule, on profitable, unprofitable, and indeterminate thought-arisings. Ghosa asked whose the system was. He was told that it was the Buddha’s and that it could only be learnt after becoming a bhikkhu. He accordingly went forth into homelessness as a bhikkhu, and in one month he learned the three Piṭakas. After receiving the full admission he acquired the four discriminations. The name given to him was Buddhaghosa—(Chapter II).

One day the question arose in his mind: “Who has more understanding of the Buddha-word, I or my preceptor?” His preceptor, whose cankers were exhausted, read the thought in his mind and rebuked him, telling him to ask his forgiveness. The pupil was then very afraid, and after asking for forgiveness, he was told that in order to make amends he must go to Sri Lanka and translate the Buddha-word (sic) from Sinhalese into Magadhan. He agreed, but asked that he might first be allowed to convert his father from the Brahman religion to the Buddha’s teaching. In order to achieve this he had a brick apartment fitted with locks and furnished with food and water. He set a contrivance so that when his father went inside he was trapped. He then preached to his father on the virtues of the Buddha, and on the pains of hell resulting from wrong belief. After three days his father was converted, and he took the Three Refuges. The son then opened the door and made [44|xlii]opened the door and made amends to his father with flowers and such things for the offence done to him. Kesi became a stream-enterer—(Chapter III).

This done, he set sail in a ship for Sri Lanka. The Mahāthera Buddhadatta^[14] had set sail that day from Sri Lanka for India. The two ships met by the intervention of Sakka Ruler of Gods. When the two elders saw each other, the Elder Buddhaghosa told the other: “The Buddha’s Dispensation has been put into Sinhalese; I shall go and translate it and put it into Magadhan.” The other said, “I was sent to go and translate the Buddha-word and write it in Magadhan. I have only done the Jinālaṅkāra, the Dantavaṃsa, the Dhātuvaṃsa and the Bodhivaṃsa, not the commentaries and the sub-commentaries (ṭīkā). If you, sir, are translating the Dispensation from Sinhalese into Magadhan, do the commentaries to the Three Piṭakas.” Then praising the Elder Buddhaghosa, he gave him the gall-nut, the iron stylus, and the stone given him by Sakka Ruler of Gods, adding, “If you have eye trouble or backache, rub the gall-nut on the stone and wet the place that hurts; then your ailment will vanish.” Then he recited a stanza from his Jinālaṅkāra. The other said, “Venerable sir, your book is written in very ornate style. Future clansmen will not be able to follow its meaning. It is hard for simple people to understand it.”—“Friend Buddhaghosa, I went to Sri Lanka before you to work on the Blessed One’s Dispensation. But I have little time before me and shall not live long. So I cannot do it. Do it therefore yourself, and do it well.” Then the two ships separated. Soon after they had completed their voyages the Elder Buddhadatta died and was reborn in the Tusita heaven—(Chapter IV).

The Elder Buddhaghosa stayed near the port of Dvijaṭhāna in Sri Lanka. While there he saw one woman water-carrier accidentally break another’s jar, which led to a violent quarrel between them with foul abuse. Knowing that he might be called as a witness, he wrote down what they said in a book. When the case came before the king, the elder was cited as a witness. He sent his notebook, which decided the case. The king then asked to see him—(Chapter V).

After this the elder went to pay homage to the Saṅgharāja,^[15]the senior elder of Sri Lanka. One day while the senior elder was teaching bhikkhus he came upon a difficult point of Abhidhamma that he could not explain. The Elder Buddhaghosa knew its meaning and wrote it on a board after the senior elder had left. Next day it was discovered and then the senior elder suggested that he should teach the Order of Bhikkhus. The reply was: “I have come to translate the Buddha’s Dispensation into Magadhan.” The senior elder told him, “If so, then construe the Three Piṭakas upon the text beginning, ‘When a wise man, established well in virtue…’” He began the work that day, the stars being favourable, and wrote very quickly. When finished, he put it aside and went to sleep. Meanwhile Sakka, Ruler of Gods, abstracted the book. The elder awoke, and missing it, he wrote another copy very fast by lamplight then he put it aside and slept. Sakka abstracted that [45|xliii]too. The elder awoke, and not seeing his book, he wrote a third copy very fast by lamplight and wrapped it in his robe. Then he slept again. While he was asleep Sakka put the other two books beside him, and when he awoke he found all three copies. He took them to the senior elder and told him what had happened. When they were read over there was no difference even in a single letter. Thereupon the senior elder gave permission for the translating of the Buddha’s Dispensation. From then on the elder was known to the people of Sri Lanka by the name of Buddhaghosa—(Chapter VI).

He was given apartments in the Brazen Palace, of whose seven floors he occupied the lowest. He observed the ascetic practices and was expert in all the scriptures. It was during his stay there that he translated the Buddha’s Dispensation. When on his alms round he saw fallen palm leaves he would pick them up; this was a duty undertaken by him. One day a man who had climbed a palm tree saw him. He left some palm leaves on the ground, watched him pick them up, and then followed him. Afterwards he brought him a gift of food. The elder concluded his writing of the Dispensation in three months. When the rainy season was over and he had completed the Pavāraṇā ceremony, he consigned the books to the senior elder, the Saṅgharāja. Then the Elder Buddhaghosa had the books written by Elder Mahinda piled up and burnt near the Great Shrine; the pile was as high as seven elephants. Now that this work was done, and wanting to see his parents, he took his leave before going back to India. Before he left, however, his knowledge of Sanskrit was queried by bhikkhus; but he silenced this by delivering a sermon in the language by the Great Shrine. Then he departed—(Chapter VII).

On his return he went to his preceptor and cleared himself of his penance. His parents too forgave him his offences; and when they died they were reborn in the Tusita heaven. He himself, knowing that he would not live much longer, paid homage to his preceptor and went to the Great Enlightenment Tree. Foreseeing his approaching death, he considered thus: “There are three kinds of death: death as cutting off, momentary death, and conventional death. Death as cutting off belongs to those whose cankers are exhausted (and are Arahants). Momentary death is that of each consciousness of the cognitive series beginning with life-continuum consciousness, which arise each immediately on the cessation of the one preceding. Conventional death is that of all (so-called) living beings.^[16]Mine will be conventional death.” After his death he was reborn in the Tusita heaven in a golden mansion seven leagues broad surrounded with divine nymphs. When the Bodhisatta Metteyya comes to this human world, he will be his disciple. After his cremation his relics were deposited near the Enlightenment Tree and shrines erected over them—(Chapter VIII).

It has already been remarked that the general opinion of European scholars is that where this imaginative tale differs from, or adds to, the Mahāvaṃsa’s account it is in legend rather than history.

Finally there is the question of the Talaing Chronicles of Burma, which mention an elder named Buddhaghosa, of brahman stock, who went from Thatōn [46|xliv](the ancient Buddhist stronghold in the Rāmaññadesa of Burma) to Sri Lanka (perhaps via India) to translate the Buddha-word into Talaing and bring it back. It is hard to evaluate this tradition on the evidence available; but according to the opinion of the more reliable Western scholars another elder of the same name is involved here.^[17]

What can be said of the Visuddhimagga’s author without venturing into unfounded speculation is now exhausted, at least in so far as the restricted scope of this introduction permits. The facts are tantalizingly few. Indeed this, like many scenes in Indian history, has something of the enigmatic transparencies and uncommunicative shadows of a moonlit landscape—at the same time inescapable and ungraspable.

Some answer has, however, been furnished to the two questions: why did he come to Sri Lanka? And why did his work become famous beyond its shores? Trends such as have been outlined, working not quite parallel on the Theravāda of India and Sri Lanka, had evolved a situation favouring a rehabilitation of Pali, and consequently the question was already one of interest not only to Sri Lanka, where the old material was preserved. Again the author possessed outstandingly just those personal qualities most fitted to the need—accuracy, an indefatigable mental orderliness, and insight able to crystallize the vast, unwieldy, accumulated exegesis of the Tipiṭaka into a coherent workable whole with a dignified vigorous style, respect for authenticity and dislike of speculation, and (in the circumstances not at all paradoxically) preference for self-effacement. The impetus given by him to Pali scholarship left an indelible mark on the centuries that followed, enabling it to survive from then on the Sanskrit siege as well as the continuing schism and the political difficulties and disasters that harassed Sri Lanka before the “Second Renascence.” A long epoch of culture stems from him. His successors in the Great Monastery tradition continued to write in various centres in South India till the 12th century or so, while his own works spread to Burma and beyond. Today in Sri Lanka and South East Asia his authority is as weighty as it ever was and his name is venerated as before.

Besides the books in Sinhala Bhadantācariya Buddhaghosa names as available to him (which have all disappeared) there was also a manual (existing now only in a Chinese translation of the 6th century CE), presumed to have been written in Pali. Bhadantācariya Buddhaghosa himself makes no mention of it; but his commentator, Bhadantācariya Dhammapāla (writing perhaps within two centuries of him), mentions it by name (see Ch. III, n.19). The Visuddhimagga refutes a certain method of classifying temperaments as unsound. The Elder Dhammapāla ascribes the theory refuted to the Vimuttimagga. The theory refuted is actually found in the Chinese version. Then other points rejected by the Visuddhimagga are found in the [47|xlv]Vimuttimagga. Some of these are attributed by the Elder Dhammapāla to the Abhayagiri Monastery. However, the Vimuttimagga itself contains nothing at all of the Mahāyāna, its unorthodoxies being well within the “Hīnayāna” field.

The book is much shorter than the Visuddhimagga. Though set out in the same three general divisions of virtue, concentration, and understanding, it does not superimpose the pattern of the seven purifications. Proportionately much less space is devoted to understanding, and there are no stories. Though the appearance in both books of numbers of nearly identical passages suggests that they both drew a good deal from the same sources, the general style differs widely. The four measureless states and the four immaterial states are handled differently in the two books. Besides the “material octads,” “enneads” and “decads,” it mentions “endecads,” etc., too. Its description of the thirteen ascetic practices is quite different. Also Abhidhamma, which is the keystone of Bhadantācariya Buddhaghosa’s exegesis, is not used at all in the Vimuttimagga (aggregates, truths, etc., do not in themselves constitute Abhidhamma in the sense of that Piṭaka). There is for instance even in its description of the consciousness aggregate, no reference to the Dhammasaṅgaṇī’s classification of 89 types, and nothing from the Paṭṭhāna; and though the cognitive series is stated once in its full form (in Ch. 11) no use is made of it to explain conscious workings. This Vimuttimagga is in fact a book of practical instructions, not of exegesis.

Its authorship is ascribed to an Elder Upatissa. But the mere coincidence of names is insufficient to identify him with the Arahant Upatissa (prior to 3rd cent. CE) mentioned in the Vinaya Parivāra. A plausible theory puts its composition sometime before the Visuddhimagga, possibly in India. That is quite compatible with its being a product of the Great Monastery before the Visuddhimagga was written, though again evidence is needed to support the hypothesis. That it contains some minor points accepted by the Abhayagiri Monastery does not necessarily imply that it had any special connections with that centre. The source may have been common to both. The disputed points are not schismatical. Bhadantācariya Buddhaghosa himself never mentions it.

The doctrines (Dhamma) of the Theravāda Pali tradition can be conveniently traced in three main layers. (1) The first of these contains the main books of the Pali Sutta Piṭakas. (2) The second is the Abhidhamma Piṭaka, notably the closely related books, the Dhammasaṅgaṇī, Vibhaṅga, Paṭṭhāna. (3) The third is the system which the author of the Visuddhimagga completed, or found completed, and which he set himself to edit and translate back into Pali (some further minor developments took place subsequently, particularly with the 12th century (?) Abhidhammatthasaṅgaha, but they are outside the present scope). The point at issue here is not the much-debated historical question of how far the Abhidhamma books (leaving aside the Kathāvatthu) were contemporary with the Vinaya and Suttas, but rather what discernible direction they show in evolution of thought.

(1) The Suttas being taken as the original exposition of the Buddha’s teaching, (2) the Abhidhamma Piṭaka itself appears as a highly technical and specialized systematization, or complementary set of modifications built [48|xlvi]upon that. Its immediate purpose is, one may say, to describe and pin-point mental constituents and characteristics and relate them to their material basis and to each other (with the secondary object, perhaps, of providing an efficient defence in disputes with heretics and exponents of outsiders’ doctrines). Its ultimate purpose is to furnish additional techniques for getting rid of unjustified assumptions that favour clinging and so obstruct the attainment of the extinction of clinging. Various instruments have been forged in it for sorting and re-sorting experience expressed as dhammas (see Ch. VII, n.1). These instruments are new to the Suttas, though partly traceable to them. The principal instruments peculiar to it are three: (a) the strict treatment of experience (or the knowable and knowledge, using the words in their widest possible sense) in terms of momentary cognizable states (dhamma) and the definition of these states, which is done in the Dhammasaṅgaṇī and Vibhaṅga; (b) the creation of a ”schedule” (mātikā) consisting of a set of triple (tika) and double (duka) classifications for sorting these states; and (c) the enumeration of twenty-four kinds of conditioning relations (paccaya), which is done in the Paṭṭhāna. The states as defined are thus, as it were, momentary “stills”; the structure of relations combines the stills into continuities; the schedule classifications indicate the direction of the continuities.

The three Abhidhamma books already mentioned are the essential basis for what later came to be called the “Abhidhamma method”: together they form an integral whole. The other four books, which may be said to support them in various technical fields, need not be discussed here. This, then, is a bare outline of what is in fact an enormous maze with many unexplored side-turnings.

(3) The system found in the Commentaries has moved on (perhaps slightly diverged) from the strict Abhidhamma Piṭaka standpoint. The Suttas offered descriptions of discovery; the Abhidhamma map-making; but emphasis now is not on discovery, or even on mapping, so much as on consolidating, filling in and explaining. The material is worked over for consistency. Among the principal new developments here are these. The “cognitive series” (citta-vīthi) in the occurrence of the conscious process is organized (see Ch. IV, n.13 and Table V) and completed, and its association with three different kinds of kamma is laid down. The term sabhāva (“individual essence,” “own-being” or “it-ness,” see Ch. VII, n.68) is introduced to explain the key word dhamma, thereby submitting that term to ontological criticism, while the samaya (“event” or “occasion”) of the Dhammasaṅgaṇī is now termed a khaṇa (“moment”), thus shifting the weight and balance a little in the treatment of time. Then there is the specific ascription of the three “instants” (khaṇa, too) of arising, presence and dissolution (uppāda-ṭṭhiti-bhaṅga) to each “moment” (khaṇa), one “material moment” being calculated to last as long as sixteen “mental moments” (XX.24; [Dhs-a] 60).^[18]New to the Piṭakas are also the rather unwieldy enumeration of concepts (paññatti, see Ch. VIII, n.11), and the [49|xlvii]handy defining-formula of word-meaning, characteristic, function, manifestation, and proximate cause (locus); also many minor instances such as the substitution of the specific “heart-basis” for the Paṭṭhāna’s “material basis of mind,” the conception of “material octads,” etc., the detailed descriptions of the thirty-two parts of the body instead of the bare enumeration of the names in the Suttas (thirty-one in the four Nikāyas and thirty-two in the Khuddakapāṭha and the Paṭisambhidāmagga), and many more. And the word paramattha acquires a new and slightly altered currency. The question of how much this process of development owes to the post-Mauryan evolution of Sanskrit thought on the Indian mainland (either through assimilation or opposition) still remains to be explored, like so many others in this field. The object of this sketch is only to point to a few landmarks.

The notes to this translation contain many quotations from the commentary to the Visuddhimagga, called the Paramatthamañjūsā or Mahā-ṭīkā. It is regarded as an authoritative work. The quotations are included both for the light they shed on difficult passages in the Visuddhimagga and for the sake of rendering for the first time some of the essays interspersed in it. The prologue and epilogue give its author as an elder named Dhammapāla, who lived at Badaratittha (identified as near Chennai). This author, himself also an Indian, is usually held to have lived within two centuries or so of Bhadantācariya Buddhaghosa. There is nothing to say that he ever came to Sri Lanka.

The Visuddhimagga quotes freely from the Paṭisambhidāmagga, the commentary to which was written by an elder named Mahānāma (date in the Middle Period and place of residence uncertain). Mostly but not quite always, the Elder Dhammapāla says the same thing, when commenting on these quoted passages, as the Elder Mahānāma but in more words.^[19] He relies much on syllogisms and logical arguments. Also there are several discussions of some of the systems of the “Six Schools” of Brahmanical philosophy. There are no stories. This academic writer is difficult, formalistic, and often involved, very careful and accurate. Various other works are attributed to him.

[50|xlviii]The Visuddhimagga is probably best regarded as a detailed manual for meditation masters, and as a work of reference. As to its rather intricate construction, the List of Contents is given rather fully in order to serve as a guide to the often complicated form of the chapters and to the work as a whole. In addition, the following considerations may be noted.

Chapters I and II, which deal with virtue as the practice of restraint, or withdrawal, need present no difficulties. It can be remarked here, though, that when the Buddhist ascetic goes into seclusion (restrains the sense doors), it would be incorrect to say of him that he “leaves the world”; for where a man is, there is his world (loka), as appears in the discourse quoted in VII.36 (cf. also [S] IV 116 as well as many other suttas on the same subject). So when he retreats from the clamour of society to the woods and rocks, he takes his world with him, as though withdrawing to his laboratory, in order to better analyze it.

Chapters III to XI describe the process of concentration and give directions for attaining it by means of a choice of forty meditation subjects for developing concentration. The account of each single meditation subject as given here is incomplete unless taken in conjunction with the whole of Part III (Understanding), which applies to all. Concentration is training in intensity and depth of focus and in single-mindedness. While Buddhism makes no exclusive claim to teach jhāna concentration (samatha = samādhi), it does claim that the development of insight (vipassanā) culminating in penetration of the Four Noble Truths is peculiar to it. The two have to be coupled together in order to attain the Truths^[20]and the end of suffering. Insight is initially training to see experience as it occurs, without misperception, invalid assumptions, or wrong inferences.

Chapters XII and XIII describe the rewards of concentration fully developed without insight.

Chapters XIV to XVII on understanding are entirely theoretical. Experience in general is dissected, and the separated components are described and grouped in several alternative patterns in Chapters XIV to XVI.1–12. The rest of Chapter XVI expounds the Four Noble Truths, the centre of the Buddha’s teaching. After that, dependent origination, or the structure of conditionality, is dealt with in its aspect of arising, or the process of being (Ch. XVII; as cessation, or Nibbāna, it is dealt with separately in Chapters XVI and XIX). The formula of dependent origination in its varying modes describes the working economics of the first two truths (suffering as outcome of craving, and craving itself—see also Ch. XVII, n.48). Without an understanding of conditionality the Buddha’s teaching cannot be grasped: “He who sees dependent origination sees the Dhamma” ([M] I 191), though not all details in this work are always necessary. Since the detailed part of this chapter is very elaborate (§58–272), a first reading confined to §1–6, §20–57, and §273–314, might help to avoid losing the thread. These four chapters are “theoretical” because they contain in detailed form what needs to be learnt, if only in outline, as “book-learning” [51|xlix](sotāvadhāna-ñāṇa). They furnish techniques for describing the total experience and the experienceable rather as the branches of arithmetic and double-entry bookkeeping are to be learned as techniques for keeping accurate business accounts.

Chapters XVIII to XXI, on the contrary, are practical and give instructions for applying the book-knowledge learnt from Chapters XIV to XVII by analyzing in its terms the meditator’s individual experience, dealing also with what may be expected to happen in the course of development. Chapter XVIII as “defining of mentality-materiality” (first application of Chapters XIV to XVI) and Chapter XIX as “discerning conditions” (first application of Chapter XVII) are preparatory to insight proper, which begins in Chapter XX with contemplation of rise and fall. After this, progress continues through the “eight knowledges” with successive clarification—clarification of view of the object and consequent alterations of subjective attitude towards it—till a point, called “conformity knowledge,” is reached which, through one of the “three gateways to liberation,” heralds the attainment of the first supramundane path.

In Chapter XXII, the attainment of the four successive supramundane paths (or successive stages in realization) is described, with the first of which Nibbāna (extinction of the craving which originates suffering) is ‘seen’ for the first time, having till then been only intellectually conceived. At that moment suffering as a noble truth is fully understood, craving, its origin, is abandoned, suffering’s cessation is realized, and the way to its cessation is developed.^[21] The three remaining paths develop further and complete that vision.

Finally, Chapter XXIII, as the counterpart of Chapters XII and XIII, describes the benefits of understanding. The description of Nibbāna is given at Chapter VIII, §245ff., and a discussion of it at Chapter XVI, §66ff.

The pitfalls that await anyone translating from another European language into his own native English are familiar enough; there is no need for him to fall into them. But when he ventures upon rendering an Oriental language, he will often have to be his own guide.

Naturally, a translator from Pali today owes a large debt to his predecessors and to the Pali Text Society’s publications, including in particular the Society’s invaluable Pali-English Dictionary. A translator of the Visuddhimagga, too, must make due acknowledgement of its pioneer translation^[22] U Pe Maung Tin. [52|l]The word pāḷi is translatable by “text.” The pāḷi language (the “text language,” which the commentators call Magadhan) holds a special position, with no European parallel, being reserved to one field, namely, the Buddha’s teaching. So there are no alien echoes. In the Suttas, the Sanskrit is silent, and it is heavily muted in the later literature. This fact, coupled with the richness and integrity of the subject itself, gives it a singular limpidness and depth in its early form, as in a string quartet or the clear ocean, which attains in the style of the Suttas to an exquisite and unrivalled beauty unreflectable by any rendering. Traces seem to linger even in the intricate formalism preferred by the commentators.

This translation presents many formidable problems. Mainly either epistemological and psychological, or else linguistic, they relate either to what ideas and things are being discussed, or else to the manipulation of dictionary meanings of words used in discussion.

The first is perhaps dominant. As mentioned earlier, the Visuddhimagga can be properly studied only as part of the whole commentarial edifice, whose cornerstone it is. But while indexes of words and subjects to the PTS edition of the Visuddhimagga exist, most of its author’s works have only indexes of Piṭaka words and names commented on but none for the mass of subject matter. So the student has to make his own. Of the commentaries too, only the Atthasālinī, the Dhammapada Commentary, and the Jātaka Commentary have so far been translated (and the latter two are rather in a separate class). But that is a minor aspect.

This book is largely technical and presents all the difficulties peculiar to technical translation: it deals, besides, with mental happenings. Now where many synonyms are used, as they often are in Pali, for public material objects—an elephant, say, or gold or the sun—the “material objects” should be pointable to, if there is doubt about what is referred to. Again even such generally recognized private experiences as those referred to by the words “consciousness” or “pain” seem too obvious to introspection for uncertainty to arise (communication to fail) if they are given variant symbols. Here the English translator can forsake the Pali allotment of synonyms and indulge a liking for “elegant variation,” if he has it, without fear of muddle. But mind is fluid, as it were, and materially negative, and its analysis needs a different and a strict treatment. In the Suttas, and still more in the Abhidhamma, charting by analysis and definition of pin-pointed mental states is carried far into unfamiliar waters. It was already recognized then that this is no more a solid landscape of “things” to be pointed to when variation has resulted in vagueness. As an instance of disregard of this fact: a greater scholar with impeccable historical and philological judgment (perhaps the most eminent of the English translators) has in a single work rendered the cattāro satipaṭṭhāna (here represented by “four foundations of mindfulness”) by “four inceptions of deliberation,” “fourfold setting up of mindfulness,” “fourfold setting up of starting,” “four applications of mindfulness,” and other variants. The PED foreword observes: “No one needs now to use the one English word ‘desire’ as a translation of sixteen distinct Pali words, no one of which means precisely desire. Yet this was done in Vol. X of the Sacred Books of the East by Max Müller and Fausböll.” True; but need one go to the other extreme? How without looking up the Pali can one be sure if the same idea is [53|li]referred to by all these variants and not some other such as those referred to by cattāro iddhipādā (“four roads to power” or “bases of success”), cattāro sammappadhānā (“four right endeavours”), etc., or one of the many other “fours”? It is customary not to vary, say, the “call for the categorical imperative” in a new context by some such alternative as “uncompromising order” or “plain-speaking bidding” or “call for unconditional surrender,” which the dictionaries would justify, or “faith” which the exegetists might recommend; that is to say, if it is hoped to avoid confusion. The choosing of an adequate rendering is, however, a quite different problem.

But there is something more to be considered before coming to that. So far only the difficulty of isolating, symbolizing, and describing individual mental states has been touched on. But here the whole mental structure with its temporal-dynamic process is dealt with too. Identified mental as well as material states (none of which can arise independently) must be recognizable with their associations when encountered in new circumstances: for here arises the central question of thought-association and its manipulation. That is tacitly recognized in the Pali. If disregarded in the English rendering the tenuous structure with its inferences and negations—the flexible pattern of thought-associations—can no longer be communicated or followed, because the pattern of speech no longer reflects it, and whatever may be communicated is only fragmentary and perhaps deceptive. Renderings of words have to be distinguished, too, from renderings of words used to explain those words. From this aspect the Oriental system of word-by-word translation, which transliterates the sound of the principal substantive and verb stems and attaches to them local inflections, has much to recommend it, though, of course, it is not readable as “literature.” One is handling instead of pictures of isolated ideas or even groups of ideas a whole coherent chart system. And besides, words, like maps and charts, are conventionally used to represent high dimensions.

When already identified states or currents are encountered from new angles, the new situation can be verbalized in one of two ways at least: either by using in a new appropriate verbal setting the words already allotted to these states, or by describing the whole situation afresh in different terminology chosen ad hoc. While the second may gain in individual brightness, connections with other allied references can hardly fail to be lost. Aerial photographs must be taken from consistent altitudes, if they are to be used for making maps. And words serve the double purpose of recording ideas already formed and of arousing new ones.

Structural coherence between different parts in the Pali of the present work needs reflecting in the translation—especially in the last ten chapters—if the thread is not soon to be lost. In fact, in the Pali (just as much in the Tipiṭaka as in its Commentaries), when such subjects are being handled, one finds that a tacit rule, “One term and one flexible definition for one idea (or state or event or situation) referred to,” is adhered to pretty thoroughly. The reason has already been made clear. With no such rule, ideas are apt to disintegrate or coalesce or fictitiously multiply (and, of course, any serious attempt at indexing in English is stultified). [54|lii]One thing needs to be made clear, though; for there is confusion of thought on this whole subject (one so far only partly investigated).^[23]This “rule of parsimony in variants” has nothing to do with mechanical transliteration, which is a translator’s refuge when he is unsure of himself. The guiding rule, “One recognizable idea, one word, or phrase to symbolize it,” in no sense implies any such rule as, “One Pali word, one English word,” which is neither desirable nor practicable. Nor in translating need the rule apply beyond the scope reviewed.

So much for the epistemological and psychological problems.

The linguistic problem is scarcely less formidable though much better recognized. While English is extremely analytic, Pali (another Indo-European language) is one of the groups of tongues regarded as dominated by Sanskrit, strongly agglutinative, forming long compounds and heavily inflected. The vocabulary chosen occasioned much heart-searching but is still very imperfect. If a few of the words encountered seem a bit algebraical at first, contexts and definitions should make them clear. In the translation of an Oriental language, especially a classical one, the translator must recognize that such knowledge which the Oriental reader is taken for granted to possess is lacking in his European counterpart, who tends unawares to fill the gaps from his own foreign store: the result can be like taking two pictures on one film. Not only is the common background evoked by the words shadowy and patchy, but European thought and Indian thought tend to approach the problems of human existence from opposite directions. This affects word formations. And so double meanings (utraquisms, puns, and metaphors) and etymological links often follow quite different tracks, a fact which is particularly intrusive in describing mental events, where the terms employed are mainly “material” ones used metaphorically. Unwanted contexts constantly creep in and wanted ones stay out. Then there are no well-defined techniques for recognizing and handling idioms, literal rendering of which misleads (while, say, one may not wonder whether to render tour de force by “enforced tour” or “tower of strength,” one cannot always be so confident in Pali).

Then again in the Visuddhimagga alone the actual words and word-meanings not in the PED come to more than two hundred and forty. The PED, as its preface states, is “essentially preliminary”; for when it was published many books had still not been collated; it leaves out many words even from the Sutta Piṭaka, and the Sub-commentaries are not touched by it. Also—and most important here—in the making of that dictionary the study of Pali literature had for the most part not been tackled much from, shall one say, the philosophical, or better, epistemological, angle,^[24]work and interest having been concentrated till then almost exclusively on history and philology. For instance, the epistemologically unimportant word vimāna (divine mansion) is given more than twice the space allotted to the term paṭicca-samuppāda (dependent origination), a difficult subject of central importance, the article on which is altogether inadequate and misleading (owing partly to misapplication of the “historical method”). Then gala (throat) has been found more [55|liii]glossarialy interesting than paṭisandhi (rebirth-linking), the original use of which word at [M] III 230 is ignored. Under nāma, too, nāma-rūpa is confused with nāma-kāya. And so one might continue. By this, however, it is not intended at all to depreciate that great dictionary, but only to observe that in using it the Pali student has sometimes to be wary: if it is criticized in particular here (and it can well hold its own against criticism), tribute must also be paid to its own inestimable general value.

Current standard English has been aimed at and preference given always to simplicity. This has often necessitated cutting up long involved sentences, omitting connecting particles (such as pana, pan’ettha, yasmā when followed by tasmā, hi, kho, etc.), which serve simply as grammatical grease in long chains of subordinate periods. Conversely the author is sometimes extraordinarily elliptic (as in XIV.46 and XVI.68f.), and then the device of square brackets has been used to add supplementary matter, without which the sentence would be too enigmatically shorthand. Such additions (kept to the minimum) are in almost every case taken from elsewhere in the work itself or from the Paramatthamañjūsā. Round brackets have been reserved for references and for alternative renderings (as, e.g., in I.140) where there is a sense too wide for any appropriate English word to straddle.

A few words have been left untranslated (see individual notes). The choice is necessarily arbitrary. It includes kamma, dhamma (sometimes), jhāna, Buddha (sometimes), bhikkhu, Nibbāna, Pātimokkha, kasiṇa, Piṭaka, and arahant. There seemed no advantage and much disadvantage in using the Sanskrit forms, bhikṣu, dharma, dhyāna, arhat, etc., as is sometimes done (even though ”karma” and “nirvana” are in the Concise Oxford Dictionary), and no reason against absorbing the Pali words into English as they are by dropping the diacritical marks. Proper names appear in their Pali spelling without italics and with diacritical marks. Wherever Pali words or names appear, the stem form has been used (e.g. Buddha, kamma) rather than the inflected nominative (Buddho, kammaṃ), unless there were reasons against it.^[25]

Accepted renderings have not been departed from nor earlier translators gone against capriciously. It seemed advisable to treat certain emotionally charged words such as “real” (especially with a capital R) with caution. Certain other words have been avoided altogether. For example, vassa (“rains”) signifies a three-month period of residence in one place during the rainy season, enjoined upon bhikkhus by the Buddha in order that they should not travel about trampling down crops and so [56|liv]annoy farmers. To translate it by “lent” as is sometimes done lets in a historical background and religious atmosphere of mourning and fasting quite alien to it (with no etymological support). “Metempsychosis” for paṭisandhi is another notable instance.^[26]

The handling of three words, dhamma, citta, and rūpa (see Glossary and relevant notes) is admittedly something of a makeshift. The only English word that might with some agility be used consistently for dhamma seems to be “idea”; but it has been crippled by philosophers and would perhaps mislead. Citta might with advantage have been rendered throughout by “cognizance,” in order to preserve its independence, instead of rendering it sometimes by “mind” (shared with mano) and sometimes by “consciousness” (shared with viññāṇa) as has been done. But in many contexts all three Pali words are synonyms for the same general notion (see XIV.82); and technically, the notion of “cognition,” referred to in its bare aspect by viññāṇa, is also referred to along with its concomitant affective colouring, thought and memory, etc., by citta. So the treatment accorded to citta here finds support to that extent. Lastly “mentality-materiality” for nāma-rūpa is inadequate and “name-and-form” in some ways preferable. “Name” (see Ch. XVIII, n.4) still suggests nāma’s function of “naming”; and “form” for the rūpa of the rūpakkhandha (“materiality aggregate”) can preserve the link with the rūpa of the rūpāyatana, (“visible-object base”) by rendering them respectively with “material form aggregate” and “visible form base”—a point not without philosophical importance. A compromise has been made at Chapter X.13. “Materiality” or “matter” wherever used should not be taken as implying any hypostasis, any “permanent or semi-permanent substance behind appearances” (the objective counterpart of the subjective ego), which would find no support in the Pali.

The editions of Sri Lanka, Burma and Thailand have been consulted as well as the two Latin-script editions; and Sinhalese translations, besides. The paragraph numbers of the Harvard University Press edition will be found at the start of paragraphs and the page numbers of the Pali Text Society’s edition in square brackets in the text (the latter, though sometimes appearing at the end of paragraphs, mark the beginnings of the PTS pages). Errors of readings and punctuation in the PTS edition not in the Harvard edition have not been referred to in the notes.

For the quotations from the Tipiṭaka it was found impossible to make use of existing published translations because they lacked the kind of treatment sought. However, other translation work in hand served as the basis for all the Piṭaka quotations.

Rhymes seemed unsuitable for the verses from the Tipiṭaka and the “Ancients”; but they have been resorted to for the summarizing verses belonging to the Visuddhimagga itself. The English language is too weak in fixed stresses to lend [57|lv]itself to Pali rhythms, though one attempt to reproduce them was made in Chapter IV.

Where a passage from a sutta is commented on, the order of the explanatory comments follows the Pali order of words in the original sentence, which is not always that of the translation of it.

In Indian books the titles and subtitles are placed only at, the end of the subject matter. In the translations they have been inserted at the beginning, and some subtitles added for the sake of clarity. In this connection the title at the end of Chapter XI, “Description of Concentration” is a “heading” applying not only to that chapter but as far back as the beginning of Chapter III. Similarly, the title at the end of Chapter XIII refers back to the beginning of Chapter XII. The heading “Description of the Soil in which Understanding Grows” (paññā-bhūmi-niddesa) refers back from the end of Chapter XVII to the beginning of Chapter XIV.

The book abounds in “shorthand” allusions to the Piṭakas and to other parts of itself. They are often hard to recognize, and failure to do so results in a sentence with a half-meaning. It is hoped that most of them have been hunted down.

Criticism has been strictly confined to the application of Pali Buddhist standards in an attempt to produce a balanced and uncoloured English counterpart of the original. The use of words has been stricter in the translation itself than the Introduction to it.

The translator will, of course, have sometimes slipped or failed to follow his own rules; and there are many passages any rendering of which is bound to evoke query from some quarter where there is interest in the subject. As to the rules, however, and the vocabulary chosen, it has not been intended to lay down laws, and when the methods adopted are described above that is done simply to indicate the line taken: Janapada-niruttiṃ nābhiniveseyya, samaññaṃ nāti-dhāveyyā ti (see XVII.24).