Historical Lessons for Translating Buddhism

The first issue is whether to translate at all or to leave the material in the original languages. There are three basic situations to consider – scriptural texts, ritual practices and debates. The historical decisions, even within one culture, have been different for each. The Buddhists of Sri Lanka, Myanmar, Thailand, Cambodia and Laos all read and recite the Theravada canonical texts in their original Pali language version. The practitioners of each of these countries write the scriptures in their own alphabets, but when they chant the texts aloud, they can sing in unison. This factor allows for a feeling of unity among all Theravada practitioners, regardless of their ethnic origin.

For ease of comprehension, the earliest commentarial literature in Sri Lanka, up until the first century of the common era, was either translated into or composed in the local language, Old Sinhalese. At the end of the fourth century, the masters began to translate part of the Pali canon as well into this tongue. In the fifth century, however, perhaps due to increased contacts with southern India, the Sri Lankans decided to codify and standardize all the Buddhist literature for their country by returning to the original Indian language of the scriptures, Pali. They stopped translating the canon into the vernacular and even translated into Pali the commentaries written in Old Sinhalese, often combining several of them to form standard grand commentaries to specific texts. In Myanmar, on the other hand, by the nineteenth century, the masters translated the entire canon into the modern tongue, Burmese. But still, none of the Southeast Asian Theravada traditions use the local vernacular in ritual ceremonies in which they chant the scriptures.

The Tibetans translated all their canonical texts into their own language, for the most part from Sanskrit, but with an occasional few from Chinese when the original Sanskrit was no longer extant. Except for mantras, which they left in the original Indian languages, the Tibetans recite all their ritual practices and chant the scriptures in their own tongue, never in Sanskrit. Many of these rituals, however, they composed themselves. Although they adopted the debate tradition from India, the Tibetans have always conducted their debates in their own language.

The Mongols translated all their scriptural texts from Tibetan. They chant some of their rituals in Tibetan and some in their own language. They debated at first in the Mongolian language, but as more and more monks studied in Tibet, they began to favor using Tibetan both for debating as well as for ritual practice. But since the Mongols, lacking the sounds in their own language, pronounced many of the Tibetan phonemes in exactly the same manner, their debates in Tibetan often became unintelligible and confusing. Therefore, in debate they began to intersperse Mongolian words into a basic Tibetan language framework. Mongol scholars wrote most of their commentaries to Tibetan texts in the Tibetan tongue.

The Chinese translated their texts from Sanskrit and other Indian dialects. They perform all their ritual practice in their own language. The Japanese, Koreans and originally the Vietnamese use the Chinese texts and rituals, written in Chinese characters, but pronounce the characters in their own distinctive manner. For many centuries, however, the Vietnamese have been chanting their rituals in their own language. They had been forced to use the Chinese characters by their Chinese conquerors, whereas the Koreans and Japanese adopted them by choice. In the twentieth century, for study purposes, the Koreans translated the canon into their own colloquial tongue.

The Manchus, on the other hand, translated the Chinese Buddhist texts into their own language, but organized their canon in the Tibetan manner. Although they followed basically the Tibetan form of Buddhism, they were more familiar with translating from Chinese due to their close political relation and assimilation with China. For ritual practice, they used Tibetan or Manchu. The northern Chinese practitioners who, from the thirteenth century, follow the Tibetan Buddhist tradition similarly perform their rituals in Tibetan or Chinese. 

The Central Asian Sogdian, Uighur and Tangut peoples, from the ninth century onwards, translated their scriptural texts primarily from Chinese. Their close economic relations with China made them familiar with this language, and many people were bilingual. The Uighurs and Tanguts also translated some texts from Tibetan. Since there was such a bewildering array of scripts used in Central Asia, the Tibetan alphabet was often used in the eighth and ninth centuries to transcribe Buddhist texts in Chinese, Uighur and Tangut. People from different ethnic backgrounds could thereby all chant the texts and rituals together.

We can learn several lessons from this historical survey. Even when cultures have kept the Pali or Chinese scriptures as their canonical texts, despite their native language being neither of these, many have eventually produced translations into their local vernacular. Thus, to facilitate study, it will be necessary for each Western country in which Buddhism becomes established eventually to translate the Buddhist scriptures into its modern language. At first, as in the case of the Pali and Chinese texts in Southeast and East Asia, it may suffice temporarily to use a version of the texts in a more universally known language such as English or Russian. But for deeper study by less linguistically gifted people, translations into Spanish, French, Estonian, Polish and so on must be made.

In certain East European countries, for instance Bulgaria, people have more respect for texts written in English or Russian than in their own language. The people of the ancient Central Asian kingdom of Khotan similarly had less veneration for Buddhist scriptures in their own tongue than those in Indian languages. But the masters, seeing how little the people understood the Sanskrit texts, encouraged translations into Khotanese. Likewise, the benefits of having intelligible scriptures in one’s own colloquial language far outweigh the disadvantages and shyness of writing in a less known tongue of the world. If these vernacular translations are not yet poetically elegant, however, one might still chant the canon in its original Asian languages for devotional purposes, as do the Burmese.

Most cultures translated their Buddhist texts from several language sources and traditions. Often, they translated versions that were themselves translations and not the originals. Thus, it is not necessary for scholars from every Western language background to translate all their texts from Asian languages. As in the case of Central Asians and Manchus using Chinese, there should be no embarrassment at translating the majority from English, for instance, into other Western languages when there are more scholars and translators trained in English than in Sanskrit, Pali, Tibetan or Chinese. This is provided, of course, that the English translations are accurate. But just as the Mongols checked the Tibetan they were translating against the original Sanskrit when there were dubious passages, likewise modern scholars can check the English if they have any doubts.

Some present-day Tibetan masters recommend that foreign Buddhists chant all their practices in the original Tibetan. One might imagine that the historical precedent for this is the Pali tradition of Southeast Asia; however, the situation is not quite analogous. Theravada Buddhists chant only the scriptures during their rituals, whereas the Tibetans chant mostly their own original compositions or Tibetan translations of prayers written by Indian masters. 

The foreign Asian cultures that have adopted Buddhism from Tibet have not always chanted their rituals in the original Tibetan. The Mongols, Manchus and northern Chinese sometimes chanted in Tibetan and at other times in their own tongues. Those among them who studied in Tibet and knew the language of that country favored doing their practices in Tibetan. For others who lacked such experience and knowledge, it was easier and more appropriate to employ their own native tongue.

It might be skillful, therefore, for Western practitioners of Tibetan Buddhism to have both alternatives available – chanting in either Tibetan or their own language. Those with linguistic ability who have studied with the Tibetans and learned their language might naturally favor using Tibetan. 

Others may find chanting in Tibetan alien and not very meaningful. But so long as poetic translations do not exist in their own language, even people who do not know Tibetan might prefer chanting in this foreign tongue since the practices in Tibetan are so rhythmic and melodic. Furthermore, Tibetan may still be preferred in international gatherings as a unifying factor so that everyone present can chant together.

Since most people will not know the Tibetan script, however, it is necessary for there to be a standard transcription system in Roman letters so that anyone knowing the Roman alphabet will pronounce it the same and be able to chant together. Standard systems are also necessary for those who share other scripts, such as Cyrillic. The historical precedent was the use of the Tibetan alphabet to transcribe various Central Asian Buddhist scripts.

The problem is that many of the Tibetan phonemes are not found in Western languages, and several of the Roman letters are pronounced differently in various Western tongues. The Tibetan script has the same problem. Its pronunciation varies with dialects, and it does not contain many of the Chinese, Uighur or Tangut phonemes. It is not clear how the Central Asians of medieval times solved this problem, but nowadays we will need to make internationally accepted conventions for the Roman script. The same is necessary for transcribing Chinese and Pali Buddhist chants.

Since people will need special training to read the international transcription systems correctly, the present variety of systems used in the various Western countries is still helpful for local use, especially in translations and for regional or private chanting. This follows the historical model of the Pali language texts being written in different scripts. 

Pali, Chinese and Tibetan Buddhist chants have a certain element of uniformity. Many of the tunes are very similar. In the West as well, when great masters transpose a classical opera from one European language to another, they merely translate the words, but retain the music.

We can do likewise when translating the Buddhist chants into our own languages. We do not need to invent something totally new. There is no benefit in changing the traditional melodies and style so that a Buddhist prayer sounds irreverently like an American folk song or piece of rock music. The accompanying musical instruments, however, would naturally alter from Asia to the West, as they did from India to Tibet. The greatest creative challenge in the process will be to translate Buddhist verse in Asian languages into poetic verse in our own tongues.

Of course, different monasteries in Tibet use slightly different melodies, and the rhythms of Tibetan or Chinese metrics are completely different from those of Sanskrit or Pali. Yet there is a unifying flavor of solemnity in all these styles. We must be very sensitive to retain this same flavor even if, out of necessity to Western metrics, we need to modify the rhythms and melodies for chanting in our own languages. 

We must avoid our chanting sounding either too Christian or too New Age, both of which would add an irrelevant and perhaps distracting and inappropriate flavor to the chants. Just as Gregorian or Hebrew chants uplift and impart a sense of religious identity to even those of their faiths who are not strong believers, likewise Buddhist chants need to fulfill that purpose for Buddhists in the West. The best way to do that is to model ourselves after the traditional idioms and forms.

Local Buddhist poetic forms have emerged in most Asian societies as joyous, spontaneous expressions of faith. There are the dohas in India, the songs of Milarepa in Tibet, Chan poetry in China and Buddhist haiku in Japan. Similar poetic expressions of Buddhism in local idioms will undoubtedly appear in Western cultures as well. Some of these may even be set to music in local styles. But these are forms of religious literature and music auxiliary to the standard texts. They are not scriptural canon chanted in temples. They serve a different purpose and therefore need not follow traditional forms.

During the course of certain Tibetan tantric rituals, several bars or lines of music of long and short horns, cymbals, bells and drums are played as offerings of beautiful sound. His Holiness the Fourteenth Dalai Lama has said that if such music were to be omitted, the rituals would not suffer the fault of being incomplete. If one wishes to play Tibetan instruments or even substitute for them saxophones, snare drums and trombones, either is perfectly all right, but neither are absolutely necessary. Just as one imagines offering exquisite fragrances and delicacies, one can also imagine offering magnificent music. It is not even necessary to represent these offerings with complicated hand gesture mudras. As His Holiness the Fourteenth Dalai Lama has also said, we can dispense with the mudras as well, except in the third class of tantra, yoga tantra, in which they form an indispensable part of the practice.

The Mongolian experience offers many suggestions for the process of adopting Tibetan Buddhist debate in the West. It is possible, in general, to debate in languages other than Tibetan. Afterall, the original debates in the Indian monasteries were held in Sanskrit. But since many Westerners are studying debate nowadays in Tibetan monasteries and using the Tibetan language, they will naturally prefer to continue debating in Tibetan once they return to their native lands. 

They and their students must be careful, however, to learn to pronounce Tibetan correctly. Otherwise, their debates in Tibetan will be unintelligible, as was the case with most Mongols. It is probably wisest for there to be two debate traditions in the West – one in Tibetan for those trained in language and for international meets, and the other in each country’s vernacular for more common use.

Root texts, especially those written by the great Nalanda masters of India were generally written in an extremely terse style, often leaving the referents of pronouns and other expressions unspecified. This allows the same verses to serve as a root from which different commentarial explanations can grow. Different commentators may therefore understand the same words according to different tenet systems or explain them on different levels, without necessarily regarding the other interpretations as invalid.  

This presents a special challenge when translating a root text. If we make its wording more explicit by filling in what “this,” “that” or another ambiguous expression refers to according to one commentary, we may make the translation easier to understand, but we also restrict the verse to that one interpretation. Ideally, a translation needs to remain sufficiently open to accommodate the different valid commentarial explanations while still making sense on its own. When greater specificity is necessary, additions should be clearly marked and the commentary on which they are based identified. 

Another issue is whether to use transcribed Sanskrit for certain technical terms or translate everything into one’s own language. The Chinese first used the former method, but in later times tended more towards the latter, while still maintaining a limited number of Sanskrit terms. The Tibetans translated almost all technical words into their own language, usually in the form of compound words made up of several syllables, each having its own meaning. The Mongols used a mixed technique of using some transcribed Sanskrit words, some purely Mongol expressions and some exact syllable-for-syllable literal translations of Tibetan terms. Let us examine the possible reasons for these choices.

The Chinese at first used transcribed Sanskrit words since they had not yet found Chinese equivalents. As they developed their own terminology, they replaced some of these foreign-derived words. But since people had become very used to many of them, they kept a certain number.

The Mongols encountered Buddhism first among the Uighurs, who, in preparing translations of Buddhist texts from Chinese into their own language, used many transcribed Sanskrit terms they had previously learned from other Central Asian languages that had borrowed them directly from India. Commissioned by Genghis Khan, the Uighurs adapted their script for writing Mongolian. Uighur scholars prepared the first Mongolian translations of Buddhist texts. Although they translated exclusively from Tibetan, they retained their custom of using Sanskrit technical terms as they did in their translations from Chinese into Uighur. Later Mongol translators followed this precedent, since most Mongol Buddhists were already familiar with the Sanskrit.

The Tibetans, on the other hand, from the beginning translated almost all Sanskrit terms into their own language. Although the Tibetans derived their alphabet in Kashmir from the Khotanese, and Khotanese translations used many Sanskrit terms, the Tibetans did not follow their precedent. This was perhaps because, unlike the example of the Mongols with the Uighurs, the Tibetans themselves prepared the first translations into their own language. 

Even though the Tibetans translated primarily from Sanskrit, they retained hardly any original terms. The Khotanese experience undoubtedly also influenced their decision. The Khotanese had preferred their Buddhist texts in Sanskrit, but few could understand them correctly. Because the people were so familiar with Sanskrit terms, the Khotanese translators had to retain a large number in their written works. But to clear up the worst misunderstandings, they devised the method of translating certain key terms with several words that expand the meaning. The Tibetans, probably wishing to avoid similar misunderstandings by their own people, adopted this Khotanese custom of translating words by expanding their meaning and employed it for almost all technical terms. Unlike the Mongols and Chinese, the Tibetans never had a precedent of Sanskrit words interspersed in their own earlier translations.

There is already more than a century of history of Western translations of Buddhist texts in which Sanskrit or Pali terminology has been liberally interspersed. Although most ordinary people will not know any of these foreign technical terms and will therefore have great difficulty understanding Buddhist texts containing them, our present situation is like that of the Chinese and the Mongols. We cannot discard all the Sanskrit terms even if we develop working equivalents. But like the Mongols, we can replace some with expressions in our own languages so as to make the texts more accessible to a wider audience, not just to academics. We can also follow the Mongol example of using the Khotanese and Tibetan technique of rendering normally misunderstood terms with several words to clarify the meaning. But certain of the more familiar Sanskrit words will need to be retained. It is too confusing to use Western language equivalents for words such as Buddha and bodhisattva. 

The early Mongolian translations often transcribed the pronunciation of Tibetan names, but according to the Mongol way of reading them. These transcribed Mongol pronunciations of Tibetan names, as well as the literal translations of polysyllabic Tibetan terms, were for the most part, however, unintelligible since hardly any Mongol knew Tibetan. But as people became more educated in Tibetan, they used more Tibetan words not only in their debates but also in their translations of texts. At this stage, they mostly transliterated into their own alphabet the spelling of these Tibetan words rather than transcribe their pronunciation. The transcribed Mongol pronunciations of Tibetan names were kept only for the personal Tibetan names that many Mongols took. 

The Mongolian solution to the issue of how to deal with Tibetan names can be instructive for our present times. The ninth-century Tibetan treatise on translating Sanskrit and the eighteenth-century Mongolian one for Tibetan that was based on it give the basic guidelines. We can translate names of persons, deities, lands, realms, flowers and so on if their meaning is understandable and it would enhance the readability of the text. But we should add the appropriate word “king,” “pure land,” “flower” and so on. If the meaning is doubtful, however, we should leave the name in Tibetan or the original Sanskrit. We must add to these guidelines the Mongol experience that if the Sanskrit or Tibetan name is already well-known and widely used, we should also leave it in the original. Otherwise, it is too confusing for the average reader who will not know to whom or to what the translated name is referring. Also, it makes sense to translate the names of deities and realms that are extremely long and cumbersome.

It also seems best to follow the Mongolian example and transliterate the spelling of the Tibetan names that we retain in the original, rather than transcribe their pronunciation. Those who are educated in Tibetan culture will want to know the spelling and will not be able to identify the name from the transcription of the pronunciation. Although the transliteration of the spelling is not user-friendly to those who do not know how to pronounce it, nevertheless people who have read about Tibetan Buddhism will have already encountered the convention adopted by previous scholars to use transliteration. They will only be confused by a transcription. If they are interested to pursue Tibetan studies, they will learn how to pronounce the language. As with allowing Sanskrit terms and Mongolized Tibetan personal names that have become widely popular, we also need, however, to keep those transcribed Tibetan words and names, such as lama, geshe, tulku, Kagyu, dzogchen, Tenzin and so on, that have been accepted in common Western use.

The main purpose for translating a text is to communicate its meaning. There can also be an additional devotional purpose of preserving exactly each word of a sacred text. The Tibetans used an extremely literal style for translating Sanskrit and even reproduced Sanskrit word order despite the fact that this violated the grammar and syntax of the Tibetan language and would be unintelligible to uneducated people if it were read aloud. Nevertheless, as a result, the Sanskrit could be, for the most part, reconstructed from the Tibetan translations when many of the original texts were lost. However, because Tibetan lacks the large scope of grammatical inflections of verbs and nouns that Sanskrit has and also lacks the large vocabulary, many of the nuances were lost.

Faithfully preserving the Sanskrit, however, was not the intended purpose when the Tibetans adopted this policy of translating literally. The Sanskrit texts were only lost several centuries later. The Tibetans wanted to systematize their earlier efforts and minimize the possibility of mistranslation.

Only two classes of pre-modern Tibetans learned how to read: government officials and monastics. The former group was extremely small, while the latter was nearly one-fifth of the population. Since hardly any monks or nuns would have been literate when they entered the monasteries or nunneries, they were taught the scriptural language as they learned to read. Thus, for most people, literacy and learning the scriptural language went hand in hand. The fact that the Tibetan translations of Sanskrit texts were unintelligible to the illiterate masses did not really matter. Those unable to read engaged in Buddhist practices that did not rely on scriptural study, such as reciting mantras, circumambulating sacred monuments and so on.

The situation in modern Western countries is very different from that of pre-modern Tibet. Almost everyone is literate, and the study of the Buddhist texts is therefore open to all. Thus, any translation must be readable, and the language must not be an obstacle for understanding. The preservation of the Pali, Sanskrit, Tibetan or Chinese original texts by means of exact, literal translations does not have great relevance or priority to the present-day situation of computer banks of information. Although such styles of translation may help the language student understand the texts in their original tongue, the resulting “Translationese” is extremely difficult for the average reader to comprehend.

If the only legitimate purpose for an extremely literal translation that reproduces every grammatical construction of the original is, as was the case with the Tibetans, to systematize the style of the scriptures and avoid mistranslation as much as possible, there are other ways of accomplishing those aims. If we look at the history of the Chinese translations of Buddhist texts, many texts were merely summarized. The most popular and widely used translations, however, have been those that were the most elegant and beautiful to read, such as those by the Central Asian scholar Kumarajiva, not those that were the most literal. Therefore, at present we need to make translations into modern languages that are both literarily pleasing as well as accurate. The two are not mutually exclusive. The main criterion for a good translation must always remain ease of comprehension.

The language and style should not be overly colloquial. If they are, as was the case with the collected sayings of the eleventh- and twelfth-century Kadam masters of Tibet, they will be extremely difficult for people of later centuries to understand. It is always better to use a more proper style that can stand as a classic over the centuries. Since certain languages such as English and Spanish have many stylistic and vocabulary differences in the various countries in which they are used, we must try, when possible, to adopt a culturally neutral style.

The technical vocabulary chosen in our own languages is also extremely crucial. At first the Chinese used the technique called “matching terms.” They employed expressions from their native Confucian and Daoist philosophical traditions to render Buddhist technical terms that seemed to be the closest in meaning. The result was much confusion and misunderstanding of the Buddhist concepts. Buddhist philosophy became merely an extension of native Chinese as ensuing authors continued metaphysical speculation about the same issues, such as non-being, as had concerned their non-Buddhist predecessors, confusedly thinking that non-being and voidness were exactly the same. After several centuries, the Chinese replaced many of these technical terms with expressions that did not already have a philosophical connotation from another system.

When certain Western translators attempt a modern version of “matching terms,” they would do well to learn a lesson from the Chinese experience. Translating Buddhist terminology with words from a modern Western system of philosophy or theology can only lead to confusion.

Just as the Chinese and Tibetans revised their technical vocabulary, we must not hesitate to begin a “New Translation Period” in the West. We need to clear our translations of words with strongly misleading, moralistic Christian connotations or those that have become meaningless jargon. “Virtue,” “merit,” “non-virtue,” “sin,” “heavens” and “hells” can be better replaced with “constructive behavior,” “positive potential,” “destructive behavior,” “negative potential,” “divine realms” and “joyless realms.” “To go for refuge” does not have any meaning in English: it is pure jargon. “To put a safe and positive direction in our life” communicates more. Furthermore, we must not hesitate to continually revise our terms when we find more accurate ways of translating. For this, it is best to go back to the original Sanskrit terms that the Tibetan and derive our terms from their definitions and etymologies. 

The translators into each modern language need to be sensitive to misleading or jargon terms in their own mother tongue. “Mitleid” in German, for example, connotes pity, not compassion. “Protectore” in Italian is a personal saint of a city who intervenes between the local people and Jesus Christ or Virgin Mary. It has nothing to do with Buddhist protectors.

A great number of these misleading terms were coined by the early Buddhist translators, many of whom were influenced by Victorian missionary thought. Although we must be grateful for their pioneering efforts, we should not perpetuate their inadequacies simply because we have become habituated to their terms. Familiarity and habit are not appropriate reasons here as they are for justifying the continuing use of Sanskrit expressions that have already become well-established. The Sanskrit words, for the most part, do not have negative or incorrect connotations. When they do, such as the word “guru” in Switzerland, which implies there the head of a dangerous Asian cult, we must replace the Sanskrit terms as well.

We must be careful, when using technical terms in Sanskrit or our own languages, to correct any imprecise usages, especially when they may cause confusion in newcomers’ understanding and practice. “Sangha,” as one of the Three Jewels of Refuge, refers to the community of arya practitioners who have attained non-conceptual cognition of voidness. The community of Buddhist monks and nuns represent the Sangha but is not the actual Sangha refuge.

Many Buddhists in the West, however, extend the meaning and use the word “Sangha” to refer to any person who belongs to their Dharma center, as if the term were the equivalent of a “congregation” of a Christian church. This is very confusing for newcomers at their center who may think they are being asked to take refuge in some of the emotionally unstable members they find there. Granted that there are plenty of psychologically disturbed monks and nuns, yet their wearing of robes and keeping of vows is worthy of respect. Although we can derive great comfort and support in our practice from the members of our local Buddhist community, if we call them “Sangha,” we are lowering the value of that term. It might be more appropriate to call them our “Dharma brothers and sisters.”

Likewise, the word “retreat” as a Buddhist technical term refers to a specific program of long-term intensive meditational practice. Nowadays many Westerners use it even when speaking of a one-day Dharma event at which participants do not leave the premises for lunch. Such loose usage makes a mockery of those who do a three-year retreat. If the word “retreat” in Western languages means simply a period when one retires from one’s ordinary activities, it is perhaps best to use it on its own only for weekend or week-long Dharma programs at which the participants eat and sleep at the center. We can expand the term to “intensive meditational retreat” when it refers to a traditional program of four sessions of practice a day over a long period of time during which, for instance, one accumulates hundreds of thousands of repetitions of a mantra. As Confucius said, we need a “rectification of names.”

These translation issues are very complex. Only after considerable experimentation with different terms and styles, and an unbiased assessment of students’ response, will we begin to be able to arrive at some conclusions. We must be sure that our final conventions are capable of equally handling the diverse contexts of written and oral translations, chanted prayers, lectures and informal discussions. We may never have a religious king, as in Tibet or Manchu China, who will gather all translators and establish conventions by decree. But just as the huge diversity of Buddhist centers must eventually siphon down to form a few Western sects in order for Buddhism to survive in its foreign context, at least a similar process must take place with respect to the wide variety of translation terms and styles.