ArticlePDF Available

"Do the Tibetan Translations of Indian Buddhist Texts Provide Guidelines to Contemporary Translators?"



Problems in the translation of Indian Buddhist texts in Pāli into Classical Tibetan
Sean Gaffney, SOAS, London University
Article Title:
"Do the Tibetan Translations of Indian Buddhist Texts
Provide Guidelines to Contemporary Translators?"
The translation of Indian Buddhist texts has posed an
array of problems of a philosophical, conceptual and
cultural nature, to any translator during all periods of
history. In particular the large number of technical terms
utilised in these Indian texts, which are crucial to their
understanding, present by far the biggest challenge to any
translator of these texts. Beginning with the earliest
extant translations from Buddhist texts into Chinese
(Mizuno 1982: 45) by the translator Lokakshema in the
second century AD followed by the Tibetan translations
commencing in the eigth century AD, and the Western
academic translations of the last one and a half centuries,
there has been a continuing quest for reliable translations
of original Indian Buddhist texts. From as far afield as
Gandhâra, Sogdiana, the Central Asian regions of Khotan,
Turfan or Tun-huang, China and Tibet, successive
generations of translators have been engaged with the
many problematic issues connected with the translation of
Buddhist texts. There have been a variety of approaches
to the translation process in these different regions, some
translating literally, some by means of paraphrasing, but
the ultimate aim of them all has been to produce faithful
and comprehensible renderings of the Indian originals in
their own respective languages.
The attempts at the translation of Indian Buddhist
texts are illustrative of some of the problems involved in
the translation of religious works generally, though these
issues also arise with the translation of any ideological,
scientific or literary work. For example the translation of
the Ma'o-tse-tung-dgongs-pa (Thoughts of Chairman Mao)
from Chinese into Tibetan necessitated the creation of an
entirely new range of Tibetan vocabulary. In order to
represent such Marxist concepts as ‘proletarian’,
‘bourgeoisie’, ‘means of production’ and ‘class-struggle’,
to name but a few, the Tibetans coined new words. The
translators of Indian Buddhist texts into Chinese or Tibetan
were presented with the specific problems of how to
render the religio-philosophical terminology, and the
Indian literary forms, into their own languages without
distorting the meaning or intention of those original texts.
During the early period of translation activity in China and
Tibet each translator would use their own interpretations
of these terms, leading to an inconsistency of terminology
when comparing texts prepared by other translators.
Some of the Chinese translators often translated a
Buddhist technical term with a Taoist technical term, so
importing totally inappropriate philosophical assumptions
into the text, while others merely transliterated the terms.
The method of dealing with these issues was gradually
refined over a period of centuries until an Indian Buddhist
text could be translated into Chinese or Tibetan with a
high degree of accuracy in terms of its representation of
the original intention of the text and the consistency of its
philosophical terminology.
The problems associated with the translation of
Indian Buddhist texts is by no means confined to ancient
history. Since the beginnings of Western academic
philological and historical studies of Buddhism,
commencing with Csoma de Koros in the mid nineteenth
century, the philological, philosophical and cross cultural
problems inherent in the translation of ancient Indian texts
have been studied and debated. Even down to the present
day there is no universally agreed method for rendering
even some of the most basic and frequently met technical
terms found in Tibetan and Sanskrit into English, or for
that matter into any other European language. Therefore
any contemporary scholar dealing with a Tibetan
translation of an Indian work has several choices in their
method of approach to the work of translation, and in the
vocabulary used to translate technical terms into English.
Whatever approach is adopted a knowledge of the
historical background to the text, if available, and the
philosophical assumptions of the text is of course a
prerequisite. When the history of a text can be traced it
may provide evidence of the distinct types of literature
that have been utilised in the composition of the text.
One text, the Pâli Jâtakanidâna (Introduction to the
Birth Stories), which is still extant in its Indian original and
that has a history that can be traced with some precision;
whereas for its Tibetan translation its date and place of
translation, and names of its translators are known. This
text and its Tibetan translation are the outcome of both
oral and written traditions spanning many centuries. The
text of the Jâtakanidâna is a composite work containing as
it does material from the earliest strata of the Pâli canon,
dating to about the fourth century BC, as well as the old
oral commentaries accompanying them, and finally
commentarial material from as late as the fifth century
AD. This text is one of the few Pâli canonical texts whose
date of composition can be put within narrow parameters.
From the introductory verses (Fausbøll 1897: I, 1) we also
learn that the text was composed at the request of three
theras (elders) of the Buddhist monastic community,
probably from the Mahâvihâra monastery in
Anuradhapura, in Ceylon during the early fifth century AD.
The authorship of this text is uncertain, being ascribed by
the Singhalese tradition to the great Indian commentator
and translator Buddhaghosa (Norman 1983:128), though
this is disputed by many on stylistic and other grounds. To
say that the text had an author in the contemporary
sense of the term is misleading, for as has already been
mentioned the Jâtakanidâna is made up of layers of early
canonical materials, particularly gâthâs (verses),
interspersed with both early and later prose commentary.
The early commentarial material contained in the
Jâtakanidâna has a history almost as old as the canonical
literature itself, being collected together at a very early
date in Buddhist history. The oral traditions preserving the
Buddhist canonical teachings were taken to Ceylon when
Buddhism arrived there in the third century BC. At that
time or very soon after (Frauwallner 1956: 18), the oral
commentaries on the canonical texts were also taken to
Ceylon. The canonical works being in Pâli while the
commentaries appear to have been in a north Indian
Prâkrit (Norman 1991: 38) that was very closely related to
Pâli. In Ceylon the canonical works were maintained in an
oral tradition in Pâli but the commentaries, with the
exception of any canonical gâthâs (verses) contained in
them which were retained in Pâli, were soon translated
into Old Sinhalese Prâkrit in order to make them more
easily accessible to the Sinhalese. So from a very early
period the canonical material in Ceylon was preserved
orally in Pâli, with the Indian commentaries being
preserved orally in Old Sinhalese Prâkrit, until these
commentaries were committed to writing at an early
though unknown date. These Old Sinhalese commentaries
came to be known collectively as the Atthakathâs
(Commentaries) and were considered as authoritative
sources for elucidating the topics contained in the
canonical works.
With the committing of the Pâli canon to writing in
Ceylon during the first century BC the ability to preserve
and disseminate canonical texts was greatly improved.
From this time up to the fifth century AD the Old Sinhalese
commentaries were retained as they were, with other
newer commentaries being added to them over time. Thus
by the fifth century AD a mass of commentarial literature
in Sinhalese Prâkrit, ancient and medieval, had built up
around the canonical sources. As the original canonical
literature was preserved in the now sacred language of
Pâli a movement arose aimed at translating, or
retranslating, all of the commentarial materials into that
language. This had a twofold objective, the first being to
clarify the meaning of these works, as the Sinhalese
language had changed significantly since their translation
from Pâli into Old Sinhalese, so much so that the
interpretation of some of them was uncertain. The second
task was to edit the texts, eliminating duplications and
spurious texts, and expounding a teaching consistent with
that of the Mahâvihâra monastery. Whichever of these
Indian translators composed the Jâtakanidâna, it is evident
that they relied heavily on these Old Sinhalese
commentaries for the traditions explaining the canonical
gâthâs (verses) included in that canonical texts.
There were several Indian translators from this period
working in Ceylon, who were associated with this
translation and editing work, the most important figure
being that of Buddhaghosa, though whether he composed
the Jâtakanidâna is a contentious issue. He is the
undisputed compiler of a great many Pâli commentarial
works, his method was to read many different
commentaries, paraphrasing the essential points from
them into a new text; his method was therefore not a
straightforward translation. Unfortunately since none of
the Old Sinhalese commentaries are currently extant it is
impossible to ascertain what relationship his works have
to their original texts.What is significant is the fact that
the Jâtakanidâna itself is the product of numerous levels of
canonical text, translation of old commentary and the
commentary of its compiler. Parts of it being initially
compiled in a north Indian dialect, followed by a
retranslation into Old Sinhalese Prâkrit and finally a
retranslation into Pâli; while other parts of it were
composed in Pâli during the fifth century AD.
The Jâtakanidâna was translated into Tibetan using a
system of translation that had been evolved by the
Tibetans over a five hundred year period. The Tibetans
actually had a larger amount of texts translated into
Tibetan from Indian works than was contained in the
entire corpus of Tibetan literature at that time. Much of
what is currently known about the content and scope of
Buddhist literature, produced in India from the early
centuries BC up to the thirteenth century AD, comes not
from any of the Indian languages themselves, but from
translations made into the Tibetan language. The
historical reason for this is that very few of the Buddhist
works written in India survive in their originals, whether in
Classical Sanskrit, Buddhist Sanskrit or one of the Prâkrit
languages. The Tibetans became prolific translators of
Indian Buddhist texts from the seventh century AD and
continued this process as long as it was possible for them
to obtain texts from India, from Tibetans who visited India
or from Indian teachers, up to about the early fourteenth
century AD. It is for this reason that the Tibetan
translations of Indian texts are so important for a
complete understanding of the development of Indian
Buddhist philosophy during this period of lost Indian
The written Tibetan language was a relatively late
creation, being ‘invented’ in the seventh century AD by
Thon-mi-sam-bhota, a minister of king Song-tsen-gam-po
(c.609-649), who created the Tibetan script after years in
India studying various Indian scripts and grammar. He
composed the first works explaining Tibetan grammar and
was also the first recorded person to translate a text from
Sanskrit into Tibetan. Therefore at the time of the Tibetans
initial contact with Indian literature the amount of
indigenous written Tibetan material was really very small.
This material falls into two main categories, the earliest
being the rDo-rings (Pillar Inscriptions) of the various
Tibetan kings. The second consisting for the most part of
historical works, the earliest of which is the bSam-yas-lo-
rgyus (Annals of Samye Monastery) (Vostrikov 1970: 24),
that are principally concerned with the genealogy of the
royal line. There are also early written materials from the
Silk-Road region, most notably the textual materials
discovered at Tun-huang, which excepting the large
amount of Buddhist material, are mainly of an
administrative nature consisting of communiques between
Lhasa and the various outposts on the edge of the newly
created Tibetan empire. The period in question then has
no works that we would recognise as being literature in its
broadest sense, though here were of course oral traditions
that later came to be committed to writing. The earliest
extant written Tibetan materials are of a utilitarian nature,
that is to say they make known the kings’ intentions, their
lineage; or orders for food, reinforcements and objects of
trade from military commanders on the borders of the new
Tibetan empire.
The Tibetan world view at the time of the creation of
written Tibetan had no connection whatever to the
primarily Indian Buddhist world view that the Tibetans
were gradually encountering as they began to import
Buddhism and Buddhist ideas and concepts from India.
This lack of a common world view would itself seem to be
a serious obstacle to the making of accurate and
intelligible translations into Tibetan. However the Tibetans
were able to avoid any inconsistencies between the two
world views by a wholesale importation of the Indian
Buddhist world view. The Indian Buddhist world view
operated initially only within the confines of the
intellectual world of the Tibetan translations of Indian
texts, while the indigenous Tibetan world view provided
the actual viewpoint of Tibetan culture in general. Only
after some centuries was the Indian Buddhist world view
fully assimilated within Tibet itself, and then with some
important modifications that have their origin in the
indigenous Tibetan world view. For a considerable period
of time there were in effect two entirely separate views of
the world in Tibet, one literary and theoretical belonging
to the higher culture of the literary world; the other being
the actual everyday world view of the majority of the
Tibetan population.
Prior to the translation of Indian Buddhist texts into
Tibetan there is no evidence, or at least any surviving
evidence, of any text being translated into Tibetan from
any other language. This lack of any discernable tradition
of translation into Tibetan makes the Tibetan
achievements in the sphere of translating Indian Buddhist
works even more remarkable. Though the process of
developing a system of translation was not without its
problems. Very early on it was recognised that in
translating Buddhist texts from Sanskrit into Tibetan,
different translators were using disparate Tibetan terms to
translate the same Sanskrit technical terms. This could
lead to confusion and misunderstandings, as many of the
Sanskrit texts were of an extremely technical nature and
so required a method of rendering Sanskrit technical
terms with great precision and in a standardised manner.
This problem was addressed after about one hundred and
fifty years of translation work when an authoritative
Tibetan-Sanskrit dictionary, called simply the Mahâvyutpatti
(Great Dictionary) or Bye-brag-tu-rtogs-par-byed-pa-chen-mo in
Tibetan, was compiled at the command of king Thri-tsug-
de-tsen (c.815-838). Following this all earlier translations
were retranslated and a new system of translation
terminology and methodology was introduced which
attempted to standardise both the terminological
consistency and the semantic accuracy of the translations.
Within two hundred years of the creation of the Tibetan
language, it had become the medium for translating some
of the most refined and complex concepts and ideas of
Indian Buddhism.
This is the background to the beginnings of the
process of translation of Indian Buddhist texts into
Tibetan. Though it must be mentioned that while the
Mahâvyutpatti provided authoritative guidelines for
translation terminology, it must not be thought that only
forms found in that dictionary were used in later
translations. Modern lexical studies (Chandra 1982) based
on an analysis of a variety of Tibetan translations, of
Indian and Buddhist works, have noted that different
Tibetan terms are in fact often employed in different texts
to represent the same Sanskrit terms. In other words the
Mahâvyutpatti was not treated by later generations of
Tibetan translators as being completely prescriptive in all
cases, but rather as providing the basis for a
standardisation of translation terminology. In the centuries
after the compilation of the Mahâvyutpatti the Tibetan
translation tradition continued to develop, adding new
Tibetan calques as new texts and vocabulary were
The scale of the translation work carried out by
Tibetans is phenomenal, thousands of Indian Buddhist
texts, on philosophy, logic, grammar, medicine and
poetics, and other subjects, were rendered into a
specialised form of the Tibetan language. This specialised
form of the language, that was solely used for the
translation of Indian texts, differed markedly from the way
in which Tibetan was used in all other spheres of writing.
Furthermore many of these Indian Buddhist texts relied
upon either an oral or a written commentary, that may or
may not have been available to the Tibetans, in order to
make their import clear. The conceptual and grammatical
difficulties of the Indian originals were fully understood by
the Tibetans, and as well as creating a special translation
language they devised a rigorous translation method
aimed at capturing the precise meaning of the texts. The
process developed for the translation of each text
consisted of a collaborative effort between one or more
Tibetan translators (lo-tsâ-ba) and one or more Indian
panditas (scholars) or âcâryas (teachers). The Indian
scholars could bring out the full import of the text at any
sections where the Tibetans were in doubt as to the
intended meaning. By having scholars from both the
Indian and Tibetan traditions the Tibetans attempted to
create translations that were as faithful as possible to their
source texts.
The language that was evolved for this translation
process has been called an ‘artificial language’ (Inaba
1985:iii) in that, in its very literalness, grammar,
vocabulary and style, it is totally alien to any of the
indigenous written Tibetan styles. The Tibetans also
created tens of thousands of loan translations or calques
in order to be able to accurately represent the rich
vocabulary of the Sanskrit language. The Tibetan
language itself simply did not have the range of
vocabulary necessary to be able to supply a different word
for each of the technical terms, synonyms and nuances
employed throughout the Sanskrit works. Therefore the
Tibetans simply created a whole new range of vocabulary
by compounding existing Tibetan words together, to
represent Sanskrit technical terminology, and by the
addition of suffixes and affixes to verbs to represent the
various Sanskrit tenses. The result is a language that
could not easily be read by a Tibetan who was not familiar
with the vocabulary and syntax of this specialised
translation language.
The Tibetan translation of the Jâtakanidâna is from
Pâli, one of the north Indian Prâkrits or colloquial dialects
(Lüders 1954: 5) used by the early Buddhist schools,
which is very close to Sanskrit in both its grammar,
vocabulary, and technical terminology. The Tibetan
translation was carried out on the principles laid down for
the translation of Indian texts. The Pâli text abounds in
gâthâs (verses), extensive use of similies, long nominal
compounds, and complex sentences. Of the thousands of
Indian texts preserved in Tibetan less than twenty Pâli
texts (Skilling 1993: 73) are recorded as having been
translated into Tibetan. The Tibetan colophons of thirteen
of these texts, including the Jâtakanidâna, inform us that
they were translated by the same two Buddhist monks,
the Tibetan translator Nyi-ma-rgyal-mtshan-dpal-bzang-po
and the Singhalese scholar Anandasri. One authoritative
fourteenth century Tibetan historical work (Bu-ston 1988:
206) compiled by Bu-ston, a student of Nyi-ma rgyal-
mtshan-dpal-bzang-po, provides the information that the
translations were carried out in Nepal during the mid-
fourteenth century.
The most striking feature of the Tibetan translation of
the Jâtakanidâna is the extremely literal way in with which
the Tibetans have translated much of the text. Only in a
few instances, usually similes or metaphors, is any kind of
paraphrasing employed. This literalist approach to
translation is in keeping with the underlying aim of the
Tibetan tradition (Ruegg 1973: 249), namely to present as
accurate and faithful a translation of the original source
text as possible. Of course it may be objected, with some
justification, that literalness is not always equivalent to
accuracy; but it does have the virtue of presenting just the
bare content of the text free from any exegetical
interpolations or glosses on certain words. Any
standardised and systematic translation method, such as
that established by the Tibetans, can only lay down the
basic principles for the translator to adhere to. It is an
implicit assumption of this method that the literal
translation of a text will ipso facto be a faithful
representation of the original text. The system of literal
translation is a reasonably coherent way to convey the
semantic content of a text but the style may leave much
to be desired. The result of such translation principles
may appear as something of a contrived creation, it does
however allow for a thorough presentation of the content
and meaning of a text with a minimum amount of
In many respects a modern translator of Indian
Buddhist texts, such as the Jâtakanidâna, is in a position
similar to that of the early Tibetan translators, that is to
say the world view of the contemporary translator is as far
from that of ancient India as was the case for the early
Tibetan translators. This similarity is reflected in the way
in which many contemporary translators of the Tibetan
translations of Buddhist texts employ an almost artificial
use of the English language, and newly created English
terminology in their work. This has interesting parallels to
the Tibetan method of creating calques, or loan translation
words, to convey the precise meaning of the original
Indian term without importing any presuppositions or
connotations from the Tibetan language. A very similar
method is currently employed in America by what might
be termed the ‘Hopkins school’ initiated by professor J.
Hopkins of the University of Viginia. When applied
universally this method leads to a good literal translation
but with the disadvantage that the English is of a
decidedly peculiar nature (Griffiths 1981: 20), and the
reader has to learn a new vocabulary of English loan
translations in order to understand the text. Very often
with this type of translation it takes a reader already
versed in Sanskrit, Prâkrit and Tibetan to be able to
understand the meaning of the English translation, the
meaning being obscure to a non-specialist reader.
Translation of course presupposes an audience, and
one has to make some decision when translating as to
who precisely is one's audience. A translation for the
specialist is an entirely different thing than a translation
for the non-specialist. One widely used method used in
translations for the specialist is to leave the technical
terminology in its original Sanskrit or Pâli form, with no
English rendering, but on the first occurrence of each term
a footnote is added explaining the term and giving the
translators preferred choice of English equivalents. This
method has the advantage of being able to discuss or
paraphrase the term without adding material to the actual
text of the translation itself. Both of the methods
mentioned do have the drawback of making a somewhat
unwieldy finished translation. Either one is faced by a
mass of English loan translations of various technical
terms, or the translation becomes heavy with explanatory
footnotes with untranslated Sanskrit or Pâli terms left in
the translation. The real root of the problem of this Indian
technical vocabulary is not that vocabulary itself, but the
inability of English to convey these terms, without using
paraphrase or multiple word translations, in an accurate
and definitive manner.
The method of translating Tibetan and Indian texts
by paraphrase has been used by a number of French
scholars, but is strangely out of favour in the English
speaking world. For the non-specialist reader this provides
the most satisfactory results, the text being
unencumbered by footnotes or unexplained loan words.
Some of these translators by paraphrase have often
incorporated materials from later commentaries into their
paraphrases of canonical texts, though for the textual
purist this can be annoying. Interestingly this paraphrase
method was used by the Chinese in some of their early
translations of Buddhist texts, though it was superceded
by the more literal approach. The use of these various
methods is indicative of the translator's dilemma, which is
centrered at it most basic level on the problem of
conveying meaning from one language and cultural
background into another, without adding to or diminishing
it by inapt translation terms.
What emerges out of the attempts to render the
Tibetan translations of Indian texts into English are those
very same problems encountered by the Tibetans
themselves. Namely, how could they bridge the
conceptual gap between Tibetan, and Sanskrit or Pâli. The
problem is the same for the contemporary Western
scholar trying to find the right English term with which to
translate one of these Indian technical terms or long and
complex sentences. These terms are unique to India,
Indian culture and the Indian languages, there is very
often no English term that could be safely employed that
would not distort the sense of the original, or more
seriously import some cultural or philosophical nuances
that are absent in the original. One is reminded of the
adage that 'the translator is a traitor', that is, the
translator in the very act of interpreting one language in
terms of another is hoplessly enmeshed within the limits
and usages of that language into which they are
attempting to translate. The original Indian texts were
composed in a distant time, where the compilers and
audience shared the same mental outlook. Overcoming
the problems posed by these kinds of questions of how to
achieve symantic accuracy, and a reasonably normal
English style are the universal aims of translation; where
the translator is trying to elucidate the text and express its
ideas and world view in a language that is totally alien to
the original.
Works Cited:
Bu ston, 1988, Bu-ston-chos-’byung-gsung-rab-rin-po-che’i-
mdzod, (Bu- ston’s History of Religion and Treasury of
Precious Sayings), Qinghai, China Tibetology Publishing
Chandra, L., 1982, Tibetan-Sanskrit Dictionary, 13 vols, New
Delhi, International Academy of Indian Culture, 1959-61,
rpt. Kyoto: Rinsen Book Company.
Fausbøll, V., ed., 1877-96, The Jâtaka Together with its
Commentary, 6 vols, London: Pâli Text Society.
Frauwallner, E., 1956, The Earliest Vinaya and the Beginnings
of Buddhist Literature, Rome, Serie Orientale Roma.
Griffiths, P.J., 1981, “Buddhist Hybrid English: Some Notes
on Philology and Hermaneutics for Buddhologists”, Journal
of the International Association of Buddhist Studies, 4, pp. 17-32.
Inaba, S., et al, 1985, The Tibetan Tripitaka Peking Edition:
Catalogue and Index, Kyoto, Rinsen Book Company.
Lüders, H., 1954, Beobachtungen über die Sprache des
buddhistischen urkanons, ed. E. Waldschhmidt, Berlin,
Academie Verlag.
Mizuno, K., 1982, Buddhist Sutras: Origin, Development,
Transmission, Tokyo, Kosei Publishing.
Norman, K.R., 1983, Pâli Literature, Weisbaden, Otto
Norman, K.R., 1991, Collected Papers, vol 1, Oxford, Pâli
Text Society.
Skilling, P., 1993, “Theravâdin Literature in Tibetan
Translation”, Journal of the Pâli Text Society, vol 19, Oxford,
pp. 69-201.
Vostrikov, A.I., 1970, Tibetan Historical Literature, tr. H.C.
Gupta, Calcutta, Soviet Indology Series, No. 4.
Ruegg, D. Seyfort, 1973, “On Translating the Buddhist
Canon”, Studies in Indo-Asian Art and Culture, vol. 3, ed. P.
Ratnam, pp. 243-261.
... To understand this transformation from colloquial to technical meanings, we have to look at how medical terms were (and still are) created in the Tibetan language. Tibetans have a history of adopting and creating technical terms from different languages and cultural backgrounds (Gaffney 2000). The Tibetan language is mono-syllabic in nature, and meaning is basically syllabic. ...
... Such debates are not new. Tibetans have a history of adopting and creating technical terms from different languages and cultural backgrounds (Gaffney 2000). A historical and textual study of the creation of Tibetan medical terminology would require a careful comparison of Sanskrit originals and their Tibetan translations, including their commentaries, 3 as well as research into medical terminology of the Dunhuang manuscripts and Zhang Zhung medical literature. ...
Full-text available
Gerke, B. 2010. Correlating biomedical and Tibetan medical terms in amchi medical practice. In Medicine Between Science and Religion: Explorations on Tibetan Grounds, edited by V. Adams, M. Schrempf and S. Craig. Oxford, New York: Berghahn Books, 127-152.
ResearchGate has not been able to resolve any references for this publication.