BuddhaSasana Home Page English Section

The Home of Pali

U Razinda
Dept. of Ancient Indian & Asian Studies, 
Nalanda, India

From: "The Light of Majjhimadesa" - Volume (1), U Chandramani Foundation, 2001

Source: http://www.rakhapura.com

Pali, in which only the Buddha delivered his noble messages, appears to have been hallowed as the text of the Buddhavacana. The language of the Buddhavacana is called Pali or Magadhi and sometimes Suddha-Magadhi, presumably in order to distinguish it from Ardha-Magadhi, the language of Jaina Canons. Magadhi means the language or dialect current in the Magadha. In Pali Lexicon, the definition of Pali is given thus: 1 pa paleti, rakkhati ' ti pali. Since it preserves the Buddhavacana (words) in the form of the sacred text, it is called Pali. In fact, the word Pali signifies only "text" "sacred text". 2

According to the tradition current in Theravada Buddhist countries, Pali is Magadhi, Magadhanirutti, Magadhikabhasa, that is to say, the language of the region in which Buddhism had arisen. The Buddhistic tradition makes the further claim that the Pali Tipitaka is composed in the language used by the Buddha himself. 3 For this reason Magadhi is also called Mulabhasa 4 as the basic language in which the words of the Buddha were originally fixed. However, for Pali now arises the question, which region of India was the home of that language which was the basis of Pali.

Westergrd 5 and E.Kuhn 6 consider Pali to be the dialect of Ujjayini, because it stands closest to the language of the Asokan-inscriptions of Girnar (Guzerat), and also because the dialect of Ujjayini is said to have been the mother-tongue of Mahinda who preached Buddhism in Ceylon (Sri Lanka). R.O. Franke had a similar opinion by different means 7; and he finally reached the conclusion that the original home of Pali was "a territory, which could not have been too narrow, situated about the region from the middle to the Western Vindhya ranges". Thus it is not improbable that Ujjayini was the centre of its region of expansion. Sten Konow 8 too has decided in favour of the Vindhya region as the home of Pali.

Oldenberg (1879) 9 and E.Muller (1884) 10 consider the Kalinga country to be the home of Pali. Oldenberg thinks that Buddhism, and with it's the Tipitaka, was introduced into Ceylon rather in course of an intercourse between the island and the neighboring continent extending over a long period. However, E.MUller bases his conclusion on the observation that the oldest settlements in Ceylon could have been founded only by the people of Kalinga, the area on the mainland opposite Ceylon and not by people from Bengal and Bihar.

Maurice Winternitz 11 is of the opinion that Buddha himself spoke the dialect of his native province Kosala (Oudh) and it was most likely in this same dialect that he first began to proclaim his doctrine. Later on, however, he wandered and taught in Magadha (Bihar) he probably preached in the dialect of this province. When in course of time the doctrine spread over a large area, the monks of various districts preached each in his own dialect. It is probable that some monks coming from Brahmin circles also attempted to translate the speeches of Buddha into Sanskrit verses. However, the Buddha himself absolutely rejected it, and forbade learning his teachings in any other languages except Magadhi. Here it is related 12, how two Bhikkhus complained to the Master that the members of the order were of various origins, and that they distorted the words of the Buddha by their own dialect (Sakaya niruttiya). They, therefore, proposed that the words of the Buddha should be translated into Sanskrit verses (Chandaso). The Buddha, however, refused to grant the request and added: Anujanamibhikkhave sakaya niruttiya buddhavacanam pariyapunitum. Rhys Davids and Oldenberg 13 translate this passage by "I allow you, oh brethren, to learn the words of the Buddha each in his own dialect". This interpretation, however, is not accorded with that of Buddhaghosa, according to whom it has to be translated by "I ordain the words of the Buddha to be learnt in his own language (i.e., in Magadhi, the language used by Buddha himself)". In fact, the explanation given by Buddhaghosa is more acceptable, because neither the two monks nor Buddha himself have thought of preaching in different dialects in different cases.

Magahi or Magadhi 14 is spoken in the districts of Patna, Gaya, Hazaribagh and also in the western part of Palamau, parts of Monghyr and Bhagalpur. On its eastern frontier Magahi meets Bengali. Dr.Grierson called the dialect of this region Eastern Magahi (Magadhi). He (Dr.Grierson) has named western Magadhi speeches as Bihari. In this time he includes three dialects, Magahi (Magadhi), Maithili and Bhojpuri. Dr.Grierson, after a comparative study of the grammars of the three dialects, had decided Maithili, Magahi and Bhojpuri as three forms of a single speech. There are four reasons for terming them as Bihari, viz.,

  1. Between Eastern Hindi and Bengali have certain characteristics, which are common to the three dialects.
  2. It becomes a provincial language like Gujarati, Punjabi, Marathi, etc.
  3. The name is appropriate from the historical point of view. Bihar was so named after so many Buddhist Viharas in the state. Ancient Bihar language was probably the language of early Buddhists and Jainas.
  4. It is not a fact that in Bihar there is no literature. In Maithili we have extant ancient literature.

Though Hindi is highly respected as a literary language in Biharyetthe Maithili, Magahi and Bhojpuri languages are deeply entrenched in the emotions of the people. The fact is that Bihari is a speech distinct from Eastern Hindi and has to be classified with Bengali, Oriya and Assamese as they share common descent from Magadhi, Prakrit and Apabhransha. It is clear 15 that an uneducated and illiterate Bihari when he goes to Bengal begins to speak good Bengali with little effort but ordinarily it is not easy for an educated Bihari to speak correct Hindi. Dr.Grierson has inclined to decide that Magadhi was a dialect of Magadha (Bihar) and some parts of West Bengal and Uttar Pradesh.

The area covered by the Buddha's missionary activities included Bihar and Uttar Pradesh including the Nepal Tarai. So it may be assumed that the Buddha spoke in a dialect or dialects current in those regions. Welhelm Geiger 16 considers that Pali was indeed on pure Magadhi, but was yet a form of the popular speech which was based on Magadhi and which was used by Buddha himself. It may be imagined that the Buddha might choose a widespread language which was used or understood by common people in the region, because through which he could propagate his noble teachings to the common people. Thus, Pali or the dialect of Magadha was more probably the language of the common people and also a lingua franca of a large region including mainly Magadha (Bihar).


(1). Dhamma Annual - Vol. 19, No. 10-11 (June-1995).
(2). Cf. the expression iti pi pali, eg., th 2 co. 618, where pali=patho. Further, pali "sacred text" as distinct from attha katha, Dpvs.=20-20; Mhvs.=33-100; Sdhs. (Saddhammasamgaha, ed. by Saddhananda); Jpts. (Journal of the Pali Text Society) 1890, p. 535.
(3). Cf. Buddhaghosa: etha saka nirutti nama sammasambuddhena vuttappakaro magadhiko voharo, Comm. to Cullavagga V 33-1. see samantapasadika, ed. by Saya U Pye, IV416.(10)
(4). Sdhs. (Saddhammasamgaha, ed. by Saddhananda). Jpts. (Journal of the Pali Text Society) 1890, pp.55(23), 56(21), 57(19).
(5). Uber dem altesten Zeitram der indischen Geschi chte, p. 87.
(6). Beitr., p. 6ff. Cf. Mur, original Sanskrit texts, II, p. 356.
(7). Pali and Sanskrit, p. 131 ff. By Pali I, of course, always understand what has been called "Literary PAW' by Franke.
(8). The home of Paiuaci, ZDMG (Zeitschrift der deutschen Morgenlandischen Desell-schaft).
(9). The Vinaya Pitaka I, London 1879, p. L ff.
(10). Simplified Grammar of the Pali language, London 1884, p. III.
(11). History of Indian Literature, Vol. II, p. 13.
(12). Cullavagga V. 33 1=Vin. II, 139.
(13). Vinaya Texts III=Sacred Books of the East, XX, p.151.
(14). The Comprehensive History of Bihar, Vol. I, Part I, p. 91, edited by Dr. Bindeshwari Prasad Sinha.
(15). Ibid. P. 89-90
(16). Pali Literature and Language, p. 6.

The Pali Language and Literature

From: Pali Text Society, http://www.palitext.com/ 

Pali is the name given to the language of the texts of Theravada Buddhism, although the commentarial tradition of the Theravadins states that the language of the canon is Magadhi, the language spoken by Gotama Buddha. The term Pali originally referred to a canonical text or passage rather than to a language and its current use is based on a misunderstanding which occurred several centuries ago. The language of the Theravadin canon is a version of a dialect of Middle Indo-Aryan, not Magadhi, created by the homogenisation of the dialects in which the teachings of the Buddha were orally recorded and transmitted. This became necessary as Buddhism was transmitted far beyond the area of its origin and as the Buddhist monastic order codified his teachings.

The tradition recorded in the ancient Sinhalese chronicles states that the Theravadin canon was written down in the first century B.C.E. The language of the canon continued to be influenced by commentators and grammarians and by the native languages of the countries in which Theravada Buddhism became established over many centuries. The oral transmission of the Pali canon continued for several centuries after the death of the Buddha, even after the texts were first preserved in writing. No single script was ever developed for the language of the canon; scribes used the scripts of their native languages to transcribe the texts. Although monasteries in South India are known to have been important centres of Buddhist learning in the early part of this millennium, no manuscripts from anywhere in India except Nepal have survived. Almost all the manuscripts available to scholars since the PTS (Pali Text Society) began can be dated to the 18th or 19th centuries C.E. and the textual traditions of the different Buddhist countries represented by these manuscripts show much evidence of interweaving. The pattern of recitation and validation of texts by councils of monks has continued into the 20th century.

The main division of the Pali canon as it exists today is threefold, although the Pali commentarial tradition refers to several different ways of classification. The three divisions are known as pi.takas and the canon itself as the Tipitaka; the significance of the term pitaka, literally "basket", is not clear. The text of the canon is divided, according to this system, into Vinaya (monastic rules), Suttas (discourses) and Abhidhamma (analysis of the teaching). The PTS edition of the Tipitaka contains fifty-six books (including indexes), and it cannot therefore be considered to be a homogenous entity, comparable to the Christian Bible or Muslim Koran. Although Buddhists refer to the Tipitaka as Buddha-vacana, "the word of the Buddha", there are texts within the canon either attributed to specific monks or related to an event post-dating the time of the Buddha or that can be shown to have been composed after that time. The first four nikayas (collections) of the Sutta-pitaka contain sermons in which the basic doctrines of the Buddha's teaching are expounded either briefly or in detail.

Buddhism: Language and Literature

Peter Friedlander

Source:  "Buddhist Studies - Lecture Notes", School of Social Sciences, La Trobe University, Australia, http://www.latrobe.edu.au/asianstudies/Buddha/index.html


This is the last chapter on the pre-Mahayana in this book. It covers a period from around the 6th century BCE to the 2nd century CE. Within the scope of this chapter I will attempt to simply sketch out various key aspects of Buddhist language and literature over a period of eight centuries. This will be rather more an investigation of the issues raised by the these topics than a detailed study.


Three key terms which we need to consider are Sanskrit, Pali and Prakrit. We will also need to consider the terms Magadhi and Ardha-Magadhi. What does Sanskrit mean? It has a root meaning which does not actually refer to a language as such but to the concept of something being refined or purified. The term Sanskrit can be found in Buddhist texts used in the sense of meaning that which is refined as opposed to that which is natural which is called Prakrit. Likewise in Samkhya the principle of Prakriti is nature, hence Prakrit is that which is natural. So in a sense then Sanskrit does not refer to a language as such but to that which is refined or purified speech.

The languages in which the Vedas are written are not quite the same as classical Sanskrit which was standardised by Panini in about the 2nd century BCE. Despite the variations in the linguistic forms from the Rig Veda, which is considerably different from classical Sanskrit, the languages of the majority of Indian high cultural texts are all in forms of Sanskrit. Some of the later texts, such as the Puranas and the Epics are often not in very refined Sanskrit, but they are still in Sanskrit. Also from around the second century BCE onwards Buddhist texts began to be produced in Sanskrit. These texts are often in a kind of Sanskrit mixed with vernacular forms and which is often referred to as ‘Buddhist Hybrid Sanskrit’. They are hybrid as they are a mix of Sanskrit and Prakrit. So you should bear in mind that the term Sanskrit does not simply refer to the classical standard form of the language but rather to a group of related language forms which share a common heritage in grammar, vocabulary and syntax.

In a similar manner the term Prakrit, which means ‘natural [speech]’ refers to a group of language forms. Indeed Prakrits appear in Sanskrit texts. For instance, classical Sanskrit dramas, such as Kalidasa’s ‘Little clay cart’ include speeches by different characters in various forms of Prakrit. For instance, whereas the cook speaks in a ‘cook’s Prakrit’ and monkeys speak in a Prakrit appropriate for monkeys, the king the leading characters and the narrator speak in Sanskrit. This is similar to the modern linguistic situation in India where within a single environment or location a variety of language forms are spoken by different people. For instance in a monastery in Bodh Gaya, the cooks and workers will speak in varieties of local dialect, but the monks will speak in standard Hindi as well as their mother tongues, and the leading figures will also be able to converse in English. In other words the use of multiple languages according to social register is a common feature in South Asia.

There are also three other elements which need to be considered. First, there is the Dravidian element in the language situation in India. This term refers to a completely different language group nowadays spoken in the forms of Tamil, Malayalam, Telegu and Kannada in the Southern states of India. There is also an isolated pocket of the Dravidian languages in the Brahui language of modern Pakistan. This language group is based on a quite distinct vocabulary and grammar. Second, it should be noted, for completeness sake, that there are also a variety of ‘tribal’ languages spoken in India which belong to various other language groups again. These include the languages of the tribal groups in Bihar, such as Santhali and Gond. Third there are also languages from distinct language groups spoken in the Himalayan and Burmese border regions of South Asia.

The situation at the time of the Buddha was probably very similar with a wide variety of languages being spoken in the area in which he lived. The dominant Prakrit language of his period in the area where he was active was called Magadhi, as is the present Hindi dialect of the area. This name is also preserved in the name given to the Prakrit of many of the Jaina scriptures. These were compiled from oral sources based on traditions active mainly in the Magadh area and the language of these scriptures is called ‘Adha-Magadhi’, that is ‘Half-Magadhi’. It is a form of cleaned up Magadhi, half way between everyday speech and a ‘pure’ language.

Language and meaning

The most important reason to consider any of this is that we need to consider how the Buddha would have addressed his audiences. He would have needed to speak in such a manner as would have been comprehensible to his audience. Clearly is a situation of such linguistic diversity he would have had to modulate his forms of speech according to the audience he was addressing. Speaking to a king and to a gang of street children, you need to speak in different ways.

Also we should consider that modern mass-education and media have been rapidly erasing the differences between dialects but that the situation in pre-modern cultures is one in which language forms vary considerably over short distances. There is a Hindi saying that after every three villages the language (that is the dialect) changes. So in that the Buddha was born on the Nepali border then his own language would not have been the same as that of Rajgir in Magadh or Sarnath in Uttar Pradesh. There are elements in the texts of the Pali canon which can be regarded as indicative of slight differences in language perhaps reflective of these ancient dialect differences.

Surely when the Buddha was addressing King Bimbasara he would have expressed himself in a different register than when he was addressing an ascetic who was visiting from another part of India, such as Bahiya who had come from Maharashtra to visit the Buddha. I would speculate that a skilled orator would express even the same notion to both audiences in different ways in order to get the teaching across as well as possible. If then you had been listening to both speeches you would have heard two versions of the same teaching. Were you then to be asked ‘which was the genuine teaching of the two’ you would have had to say that both were genuine, although they were different in exact wording as they carried the same teaching.

The question of how to teach and the languages in which to teach is indeed addressed in the Pali canon. It is said that the Buddha was asked when teaching in different areas should the teachings be in a single language or adapted to the local language. The Buddha said that the teachings should be made in the language of the area. So disciples of the Buddha would have been teaching in a variety of languages according to the contexts in which they found themselves, so that people could understand them.

The Buddha is also said to have favoured natural language, Prakrit, over refined speech, Sanskrit, as the latter would not have been comprehensible to the general public. So what then is the relationship between Prakrit and Pali?

In a sense the term Pali, like Sanskrit, does not refer to a language at all. Richard Gombrich pointed out that it actually means ‘sacred scripture’ and is a descriptive term for the Theravada scriptures and the language they are in. It is a standardised and consistent language based on earlier dialects. It is not exactly what the Buddha said, it is a standardised form of what the Buddha said. It is close to the Prakrit Magadhi languages that the Buddha probably spoke in, but it is not identical to them.

The Pali canon features long set phrases which are repeated countless times in identical terms. Such as the formulation of the Noble Truths and set descriptions like ‘he saluted the Buddha and sat down at one side of him’. These set doctrinal phrases and stock descriptive elements are, however, normally contextualised within passages which are each in a sense unique.

It seems to therefore be appropriate to point out that we have no way of knowing when the tradition of explaining the Pali canon with further commentorial material began. The textual traditions now extant always feature the main texts and subsidiary commentaries. It is known than that this tradition goes back in Sri Lanka to the time of the introduction of Buddhism, when it is said that commentaries explaining the texts were introduced along with the texts themselves. (I am using the term text here to refer to a spoken text, not a written text). This pattern of text and commentary is common in South Asian literature. It is also a feature of non-Buddhist Indian literature and a Sutra (Skt) or Sutta (Pali) means a ‘string’ or ‘thread’ and is the condensed essence of a text onto which a commentary should be stung.

The repetition of set phrases and material to contextualise and explain them is a feature which is typical of texts with commentaries. Part of the motivation for this is clearly that it is no good giving a teaching in a language nobody understands, it has to be accessible. Likewise even if the main teaching is linguistically comprehensible it will probably need an explanation to contextualise and make the meaning clear to the particular audience which is being addressed. Thus the issue of what constitutes ‘the speech of the Buddha’ (Buddhavacana) is further complicated here by the possibility that we may have multiple versions of reported versions of what the Buddha said, all genuine, but all slightly different.

There is also an issue which is raised by a reference in the canon to two Brahmin brothers who had become monks and remembered the teachings and asked for permission to chant them in the manner of a Vedic chant. But the Buddha said that this was not appropriate. Despite this the Buddhist texts are chanted, but the manner and styles of their chanting do not conform to the Vedic patterns for the chanting of texts.

If we entertain the notion that the Pali texts are not the actual speech of the Buddha, but standardised versions of what he said, what then would be the relationship of the Sanskrit versions of the texts to the Pali versions? The Sanskrit versions are also standardised versions and would stand in similar relationship to the original utterances.

If we put aside the Theravada claim that the Pali texts are literal word of the Buddha then we have to consider this possibility. The existence of other Prakrit versions also seems to point to the same truth. None of the extant versions are simply ‘the literal words’ of the Buddha, all textual traditions are, in one way or another, standardised versions of the words of the Buddha. The canon itself contains references to how it is important to understand the intended meaning of the text and not get caught up in the literal meaning.

In the present day the various Sanskrit and Prakrit versions of the canon are not all perfectly preserved. There are large sections of the canons of a number of Nikaya Buddhist traditions extant in their original language forms and, fortunately, more extensive translations of these texts into Chinese. Therefore it is possible to compare the Sanskrit, Prakit and Pali versions of some texts.

An instance of this is the Dhammapada. This is available in Pali, Sanskrit, two Prakrits, Chinese and Tibetan translations. The various traditions do not have exactly the same text. The number of verses vary, the order of the verses vary and the texts of the verses vary and to some extent even the meaning of individual verses vary.

The common endeavour behind all of this was clearly a constant effort by different people in diverse locations to keep the Buddha’s teachings comprehensible. For some people it seemed that Pali was the best, for some Prakrits, for some a widely know standard language, Sanskrit, seemed the most appropriate. For some it was necessary to translate the texts into totally new languages, such as Chinese. In the article by (I have forgotten his name) on the translation of the Lotus Sutra into Chinese there is a fine description of this translation process. It needed one or more Indian Pandits and one or more central Asian and Chinese Pandits who would sit together. The India Pandit reads out the text to the Chinese Pandit who writes it down and then it is compared for meaning by the various people involved. In the particular case that was being studied in this article it is argued that although the text is described as being in Sanskrit, the Indian Pandit was apparently reading it out in Prakrit based on the evidence of the kinds of mistakes that were being made in translation. So this suggests that not only do we need to consider the languages of the written forms of the texts but of the spoken forms of exposition which were employed. We must remember then that the text consists of the text, the expounder and the listener.

Linguistic change and the Growth of Buddhism

One other point about the linguistic changes in the growth of the canon. It seems to reflect to some extent the geographical spread of Buddhism. By the second century CE Buddhism had spread throughout South Asia and into Central Asia and China. Therefore the issue of how to give the teachings must have been of prime concern in the Buddhist world. The common consensus was clearly that the texts needed to be translated into languages appropriate for the peoples of the areas in which Buddhism was active. But at the same time there is of course the overarching need to maintain the meaning of the teachings while the form of expression varies. Within the North Indian linguistic area is was possible to maintain key terms in forms which were commonly employed, sukha, dukkha, dharma, karma, nirvana, samsara, etc.

But, once the texts started being translated into Chinese a new set of problems was apparent. Just as terms such as dharma, nirvana, samsara present problems for translation into English, so to is there a problem when translating such terms into Chinese. There was, for instance, no common view of reincarnation samsara as a given truth in Chinese.

Interestingly enough the first school of Chinese translation, the old school, translated by finding the most similar Chinese terms available for Indic terms, normally finding terms from Taoism that were equivalent. Thus the Buddha became a teacher of the Tao rather than the dharma. This translation approach was standard from the beginnings in the 1st/2nd century CE up to around the 5th century. At this point the translators revised their views and retranslated the texts again using Chinese equivalent versions of the Indic terms rather than Taoist equivalents.

Buddhist literature

So was the canon of the Nikaya Buddhist traditions exactly the same for all the traditions? I have indicated above that in the case of the Dhammapada there were variations between the different traditions. Variations in the number of verses and verses that are common to all traditions and unique to individual traditions. You cannot simply say that one version is the original version, yet it is desirable to consider how the versions related to each other. It is likely that non of the extant versions are the original version as oral traditions are often more fluid than textual traditions. So rather than saying that any one textual version it might be better to propose that all the versions are but windows onto an earlier oral tradition. There are in the case of other texts instances where the Pali versions of texts seem more developed than other versions. For instance the Pali Mahaparinibbana sutta seems more complex than the version translated into Chinese from the Sanskrit Sarvastivada tradition. The latter having a more simple description of the funeral rites and the former a more elaborate version.

Nikaya literature: vinaya, sutta, and abhidharma pitakas

There are basically three parts of the Nikaya Buddhist canon. The Sutta pitaka, the Vinaya pitaka and the Abhidhamma pitaka. The Sutta pitaka is fairly consistent in some parts over the various versions, in particular the Digha and Majjima, Anguttara and Samyakta Nikayas are fairly consistent in their contents, if not in the exact forms of the texts.

However the next Nikaya, Khuddaka Nikaya which in the Pali version contains 14 texts has a much greater variation in its contents. It includes the Khuddaka Pattha, a sort of early version of a collection of the chants for daily recitation and the Dhammapada, which I have already noted has considerable variations between the various versions. The next text is the Udana, which at least in the Sarva stiva da version is similar to the name given to the Dhammapada which is called the Udanavarga. There is also the Itivuttika further sayings of the Buddha and the jatakas. The number of the jatakas also varies from tradition to tradition. There are also instances of completely different works being included in this part of the canon by different traditions.

The Sarvastivada tradition included a text called the Mahavastu in the canon, a kind of life of the Buddha, but the Theravada tradition does not include this text. While the Theravada tradition included the Vimanavattu and the Pettavatu in its canon, tales on the good and bad results of giving or not giving to the samgha. These last two texts are regarded as very late by scholars. So to are the following texts called the Buddhavamsa, an account of the previous 24 Buddhas and the Cariya-pitaka which is an account of how the Buddha manifested the ten perfections in his previous lives as a Bodhisattva. The very fact that the title of the last includes the term pitaka in its title, which is a term that means basket or winnowing fan suggest that it must have come from a time when the canon could be put into baskets, clearly only possible once it had been written down.

The Pali canon was first written down in the first century BCE in Sri Lanka according to Sri Lankan sources. The traditional explanation of this is that it was due to fear of parts of the canon getting lost that led to it being written down. It is said that during a famine there was only a only a single monk left alive who knew one section of it and this was the cause of it being set down in writing. You may think it was odd that it was not previously written down, but there seems to have been a reluctance to write things down in ancient India.

To return to the contents of the canon the next part is the Vinaya pitaka which includes details on how the monks and nuns should live and stories to explain the rules of the monastic code. Even in the fifth century CE when Chinese pilgrims were visiting India and trying to get copies of the Vinaya they found it quite difficult as in many places it existed only in the form of oral tradition. The reluctance to commit to writing parts of the canon seems to have been a long standing aspect of the tradition in India. People simply preferred to remember the whole thing. It was indeed one specialisation that monks could have was to memorise entire parts of the canon, and memorisation of the Vinaya pitaka was apparently a common phenomena.

The last part of the canon is the Abhidhamma pitaka, a philosophical study of the Buddha’s teachings. This contains seven works in the Theravada version. In the Sarva stiva da version the number and nature of the works was somewhat different. Certain parts show evidences of having been based on similar earlier traditions, others are clearly distinctive contributions of the various schools. It is not clear if all schools had their own Abhidhamma pitaka traditions or they were shared in common by various traditions. The main Abhidhamma pitaka traditions seem to have been those of the Theravada and the Sarva stiva da traditions.

The different Abhidhamma pitaka traditions are acknowledged to be later parts of the canon which were not in existence at the time of the first council and they post date the Sutta and Vinaya pitakas. There are considerable variations between the different philosophical traditions. The Theravada tradition held that there were only four realities rupa, citta, cetasaka and Nibbana, whereas the Sarva stiva da tradition held that there were five realities and included space a ka sa as a fifth reality. Also whilst the Theravada tradition held that only the present moment ‘existed’ when things were perceived, the Sarva stiva da tradition held that things ‘existed’ in the past, present and future. This last view accounts for the name of the tradition which means ‘all exists’. Due to this it is natural that the philosophical texts vary in their contents. Despite sharing a common interest in philosophical analysis. Indeed the differences between the traditions form the basis for a Theravada tradition text, the Katthavattu or ‘Points of controversy’ which outlines the differences between the traditions as seen from a Theravada viewpoint.

A point of note in this is that in the Katthavattu the philosophical position on the possibility of transferring merit to deceased relatives of the Theravada tradition is put as that it is impossible in distinction from that of the other schools which say that it is possible. But, this viewpoint also conflicts with the views expressed in the Khuddaka Nikaya of the Theravada canon itself, in which the transfer of merit is clearly regarded as possible. A further twist to this issue is that in the later text called Milanda Panha a compromise is suggested that merit can be transferred to some classes of preta, and this is the current view of most Theravada tradition followers.

Nikaya and Mahayana literature

There is a further question which is worth addressing here is. ‘What parts of the Nikaya Buddhist canon are also accepted by Mahayana Buddhist traditions?’ Interestingly enough though the question becomes not really what are accepted texts, so much as what are texts that interest different traditions. The Sutta texts for instance are accepted as genuine by the Mahayana tradition, but they are of little interest to the Mahayana it seems. However, almost all the traditions agree on the importance of the Dhammapada as the essence of the Buddha’s teachings.

The Vinaya pitaka is also a commonly held part of the early canon. Although that majority of East Asian and Himalayan traditions follow the Sarva stiva da Vinaya rather than the Theravada Vinaya, however there are in theory no major differences. This is of course quite separate from the question of how the Vinaya is interpreted which evidently varies widely between the Northern and Southern traditions.

The Abhidhamma contains almost no texts which are common between Nikaya Buddhists, let alone between the Nikaya Buddhists and the Mahayana Buddhists. However, there is a similar fascination with philosophy in all the traditions.

It is also vital to realise that there is much in Theravada tradition which is unique to it and not held in common with other Nikaya Buddhist traditions. The great synthesis of teachings in the Visuddhimagga, ‘The Path of Purification’ by Buddhaghosa which was composed in the 5th century CE is distinctly Theravada in its viewpoint. It was based on a translation into Pali of the existing Singhalese commentaries on the canon and records traditions which may well go back in origin to India but had undergone centuries of evolution in Sri Lanka. Buddhaghosa himself was from North India, from near Bodh Gaya and went to Sri Lanka to translate their vernacular commentaries into Pali.

The famous Sri Lankan chronicles, such as the Mahavamsa are also distinctly Sri Lankan Theravada creations that link the history of Buddhism to that of the ruling dynasties of Sri Lanka.

There was also a continuous tradition of creating new Pali texts in South East Asia, in Burma, Thailand, Cambodia and Laos. It is interesting to note that in this case the argument for Pali as the sacred language has completely altered. The early argument for Pali it seems was, as suggested above, that it was comprehensible to the people as it was close to everyday speech. Evidently in Sri Lanka and South East Asia this was not the case. Rather it was seen as being the authentic language of the Buddha. In a sense then it has become a kind of purified language whose function is akin to that of Sanskrit in India, a kind of sacred lingua franca comprehensible over a wide area and felt to be the essence of refinement and imbued with great power and sophistication.

The Earliest Buddhist Manuscripts

Finally, as an epilogue let us consider the case of the earliest Buddhist manuscripts yet discovered. A few years ago the British Library in London was approached to find out if it was interested in acquiring what appeared to be some old manuscripts which had emerged from war torn Afghanistan. These were a collection of rolled up birch bark manuscripts. These are very difficult materials to deal with as they normally crumble into dust as you touch them. In this case they were stored in urns and they were purchased in the urns. The library spent a year and a half gradually humidifying and unrolling the manuscripts a millimetre at a time and ended up with fragile sheets of birch bark sandwiched between perspex sheets. It should be born in mind that birch bark is a bit like vellum, as long as its kept in normal conditions it is pliable and an excellent writing surface, it only become so crumbly if left to dry out in an arid environment for two thousand years. These were then photographed and digitised. They are a very exciting discovery as it has become apparent that they date from around the first century CE. They are written in a dialect of Prakrit in a script called Kharoshti, and the number of scholars it is said who can read this script are said to be merely a handful. The Kharoshti script was popular in the North Western part of India and dropped out of use by the time of the Islamic invasions of India. The group of scholars who are working on these manuscripts are still working on deciphering them.

The initial reports indicate that they are all fragments of works. This turns out to be because they are fragments of old manuscripts which had been re-copied and the old manuscripts were interred in an urn and buried as if they the body of the Buddha. This in itself is fascinating as it shows that the Buddhists buried their old manuscripts, Hindu’s also treat their manuscripts like their dead and prefer to ideally place them into rivers as they do the ashes of bodies.

The contents of the manuscripts include sections from Dhammapada, the rhinoceros verses, and verses in praise of the lake now known as Manasarover by Mount Kailash, known in Buddhist literature as lake Anavatapta. There also indications that they productions of the Dhammaguptika tradition. They contain no parts of the Vinaya or Abhidhamma pitakas and appear to be all drawn from the Sutta pitaka. However, we are still waiting for further detailed reports on their contents.


In conclusion then it is clear that the breadth and depth of Buddhist literature is hard to comprehend. Even were you to become a master of the Theravada Tipitaka you would still not have read the greater part of the literature of the other Nikaya Buddhist traditions. Also to be able to do a good comparative study of this literature in real depth you would need to know not just Pali, Prakrit and Sanskrit, but also to access the translations of the parts of the Nikaya Buddhist canons lost in Indic languages you need to learn Chinese to read these portions in translation. This is as they say in Australia ‘a big ask’, however, beginning to map out the dimensions of this issue is the first step on the road to the study of Nikaya Buddhist literature.

The Pali Language

Source: http://www.buddhamind.info

A question often asked is: "Did the Buddha speak Pali?" If so, how much of the original language has been retained? If not, how much has translation affected the accurate transmission of the teachings? There seems to be no one answer to these questions but I offer the following as the results of my investigation.

The paramount power in India for two centuries, spanning both before and after the Buddha, was the Kingdom of Kosala, of which the Buddha's birth kingdom, Magadha, was a fiefdom. Magadhi seems to be a dialect of Kosalan, and there is some evidence that this was the language that the Buddha spoke. The Pali of the Canon seems to be based on the standard Kosalan as spoken in the 6th and 7th centuries BC. The script used on the rock edicts of Asoka is a younger form of this standard. On one of the Asoka pillars (about 300 BC) there is a list of named Suttas which can be linguistically placed within the Singhalese Canon.

Sanskrit was also widely spoken and warrants discussion. It seems to have been the language of the Brahmin's, the 'spiritual' class. It is etymologically older than Pali but, as regards texts and inscriptions, the native tongue (Kosalan) was the more common or popular medium. In the Text we see the Buddha encouraging his disciples to teach in the popular language of any area. However after the Buddha's death, what were considered more 'learned' forms were gradually made use of, despite the fact that these gave a less faithful picture of the living speech. Slowly the efforts to represent the real facts of the spoken language gave way to another effort, the expression of learned phraseology, until roughly 300 AD, classical Sanskrit became used exclusively in relation to Buddhism. This trend is reflected in the scripture of later Buddhist traditions.

The use of Pali is practically confined to Buddhist subjects, and then only in the Theravada school. It's exact origin is the subject of much learned debate and from the point of view of the non-specialist, we can think of it as a kind of simplified, common man's Sanskrit. The source of the Pali Text we have lies in the North of India. It is definitely not Singhalese in origin as it contains no mention of any place in Sri Lanka, or even South India. The similes abounding in the Singhalese literature are those of a sub-tropical climate and of a great river valley rather than those of a tropical island.

Being an essentially oral language, lacking a strong literary base of its own, it adopted the written script of each country it settled in. It is clear that by the time the Text arrived in Sri Lanka, with Asoka's son Mahinda, about 240 BC, it was considered closed.


Any historical study is much like a jigsaw puzzle. Piecing together information from a scrap of parchment here, a clay tablet there; comparing various bits of antiquity, the opinions and insights of others; analysing and evaluating - and then - coming to a conclusion. The more Buddhist history books I studied, to try and determine precise information, the more opinions I ended up collecting. History, it seems, can be very much a matter of opinion.

Very few undisputed facts exist by which to prove the authenticity of the Pali Canon. Even the dates of the Buddha are questionable. The earliest reliable dates in Indian history that we have are those for Emperor Asoka's rule; 274 - 236 BC. We can also be relatively certain that the Text remained unchanged from the time it was written down, about 80 BC.

As regards the reliability of the Text I felt two items to be of greatest importance.

* Firstly: The reason that anything survives the rigours of more than 2000 years of history is that it is considered to be of great value. Presumably the reason for this evaluation was that the teaching was seen to work, i.e. to lead to the transcendence of suffering. Such a known treasure would have been well guarded and part of this protection would have been a tremendous concern for retaining the 'jewel' in its entirety, i.e. accurately.

* Secondly: After several centuries of travelling to many different lands and being translated into different languages, the disparity between the various renderings of the main Text existing today in Pali, Sanskrit, Chinese and Tibetan is typically greatest in matters of least importance. Only very rarely are differences founded on doctrinal matters. It can be seen that these works are clearly not independent compositions, being very similar in their substantive content. This "authenticity by comparison" is an important item in support of scriptural accuracy. More specifically, the Vinaya is almost without exception, identical in every Buddhist tradition.

On a more general note:

I feel that the majority of us who have come to give the Text some consideration, originally set out in search of a guide by which to find a way to resolve the root-problem of our personal existence. The process of production warrants investigation but surely the true test of any guide book is its ability to lead one to the desired destination. The whole energy behind the Buddha's teaching was the ending of suffering. If what you glean from the Text eases or ends your suffering then the teaching has been accurately transmitted. What is of greatest importance is to take the teachings that seem relevant, that feel applicable to your life, and to make them a personal reality, to turn the theory into practice.


Source: Internet

[Back to English Index]
last updated: 07-07-2005