Reflections on Translating Philosophical and Religious Texts

Paul L. Swanson[*] []
James W. Heisig[**] []


Two scholars of religion, both seasoned translators, share their thoughts on the promise and perils of translating texts from Far Eastern languages. Topics include relations between original and translation, limitations on the possibility of accurate translation, the influence of intended audience, and the readability of translations of technical language.


Dois cientistas da religião, tradutores experientes, compartilham pensamentos nas promessas e nos perigos de traduzir textos das línguas orientais. Os temas incluem relações entre originais e traduções, limitações na possibilidade de tradução exata, a influência do público desejado e a legibilidade de traduções da língua técnica.

1. Religious Texts: Paul L. Swanson

As one who has been translating religious texts (or texts about religious thought and practices)—both modern and classical, both primary texts and secondary studies, from Japanese and Chinese into English—for over twenty-five years, I will try to put together some thoughts on what is involved in the translation of these kinds of texts, and to reflect on the challenges and rewards of this enterprise.

Never having studied translation formally, and hence being all but completely ignorant of the literature on the subject, I realize I may end up repeating what is already cliché to those who know about such things. In any case, experience has taught me how slippery and ambiguous translation and crossing between languages is, how the nuances shift so quickly, how the word associations in one language can lead in different directions in another, how diverse are the implications of words and thoughts in different language contexts. I begin with what I have found to be two cardinal rules about translating words and ideas from one language into another: first, that there is no one-to-one correspondence between words of different languages; and second, there is never only one correct translation.

1. There is no one-to-one correspondence between words of different languages.[1] None, never. Beyond the limitations of having to work in specific languages,[2] the cultural background and historical development of a word gives it connotations beyond the dictionary definitions that can never be exactly replicated in another language. There may be close correspondences between words in sister languages (e.g., between French or Spanish with English), but these ties weaken as the "distance" between the languages increases, as in those of modern English and classical languages such as Latin or Greek. Still, there is some historical link between Latin or Greek (or even Sanskrit) and English. The cultural and historical gap that separates English and languages such as Chinese, Korean, or Japanese is much greater by comparison. There is also the additional complication that these languages use characters with visual impact and meaning, a factor missing in alphabetic or phonetically transcribed languages. The Chinese characters, with their pictorial and/or immediate visual impact, "work" differently from phonetic words. Besides these complications, the task of translating religious texts involves the problem of dealing with the intricacies and nuances of religious discourse. Descriptive or technical passages (travel guides or instructional manuals, for example) are more likely to have a satisfactory corresponding translation than the kinds of "slippery" subjects one finds in religious, philosophical, or literary texts.

A first corollary to the rule is that there is a great danger of misrepresentation if a given word in one language is always translated with the same word in another, a "foolish consistency" that can only be maintained by disregarding the context. Strict adherence to a "consistent" translation can lead to what one of my acquaintances has called "dictionary fundamentalism."[3] This does not necessarily advocate arbitrariness or blatant inconsistency. A reasoned consistency is a laudable goal, but only with the caveat that the translator should be open to possible exceptions depending on the context.

A second corollary to this rule is that when you are translating a text, you can never be sure how well you really "got it." You can always be certain that your translation is not perfect, but never so sure to what extent it is imperfect. Like a jig-saw puzzle that can never be finished, and with many pieces missing, you can sometimes get a good grasp of the picture as a whole, but you never have all the pieces, and often some of the pieces just don't seem to fit.

2. There is never only one correct translation. A variety of translations are possible for all texts, without having to conclude that one of them must be "correct" and all the others "wrong." It is even possible that different translations could all be "right" in different ways; some can be more correct or accurate than others. Or again, they could all be "wrong" or inadequate. A few years ago I was confronted by our copy editor at Nanzan, who pointed out that I had translated the same passage from a Chinese Buddhist text in quite different ways in two different publications, and he wanted to know which one was "right." At the time I felt a bit embarrassed, but if I had had my wits about me, I might have argued that both were "right," given their context. One was a technically precise translation used in the context of an academic essay for a Buddhist studies journal; the other, a much freer translation in a chapter for an encyclopedia intended for a general audience. Both, as far as I could tell, were appropriate for their contexts and purposes.

This ambiguity, imprecision, and multivalence of language (which, as I have said, is compounded in religious texts) is probably good cause to despair of computers ever translating religious texts reliably. There are those who believe that eventually computers will be able to take over the task of translation. I have my doubts—but then, many people said a computer could never beat a master at chess. At the same time, as one who spends much time on the mundane tasks required for translation (looking up words that I have looked up many times before, checking references, trying to remember how the word was translated previously), I must admit I look forward to the day when computers can handle some of these technical aspects and perhaps provide a preliminary translation or suggestions that one could use as one would use other reference works. But we are not there yet, at least not in my judgment. At the Nanzan Institute we have tried out a number of translation programs, but none has even the minimum sophistication for our needs. Recently we were playing with a program that translates between modern Chinese and English. A colleague of mine typed in a number of statements to test the accuracy of the program, including "Paul Swanson has a pony tail." The translation provided in Chinese was a phrase that (retranslated back into English) could mean "Paul Swanson is a small horse's ass." On second thought, perhaps computers have more insight than we give them credit for.

The Interweaving of Three Levels

To approach this matter from a different angle, we might say that in working with languages and translating texts, there are at least three different levels to consider: (1) particular words and terms; (2) more general concepts and ideas, along with their historical development and implications; and (3) the intended audience, both of the original text and of the translation.

Not so long ago I was struck by these three levels in the course of preparing a paper in English that I had originally prepared in Japanese. I discovered that one cannot give the "same" paper in two different languages. When one works in a second (or third) language, not only do the words and ideas fail to carry the same nuances as the first language, but one is pulled in different directions by the force of the words and ideas in the different languages, and by the (perhaps imagined) expectations of the intended audience.[4] Let us look at these three levels.


As I said before, and as anyone working in translation quickly realizes, there are no "exact" equivalents for translating words from one language into another. Each word has multileveled meanings and implications that can never be carried over in toto to another language. When a word is used, it carries with it layers of historical development, contextual nuances, and half-hidden associations that are often unconsciously present even to the original verbalizer.[5] Even something as concrete as a pen or a fork can have quite different nuances and carry very different implications in different languages. In English "rice" is "rice"; in Japanese, there is a distinction between komé (rice grain), gohan (cooked rice), and raisu ("rice"); what goes into your mouth is the same "thing," but as you are pouring it into boiling water it is komé, if you eat it with chopsticks from a bowl it is gohan, and if you eat it from a plate with a fork it is raisu.


At the level of ideas, as with individual words, one is often led in different directions by what seem to be near-equivalent terms. For example, if one uses the concept of "scripture" or "canon" for what appear to be somewhat equivalent words in Chinese or Japanese, one is immediately suggesting ideas, connotations, and implications that derive from the use and development of these ideas in the English context, some of which have Judeo-Christian implications that would not be applicable in, for example, a Buddhist or Taoist context. On the other hand, using a term such as tripitaka or daizōkyō would be confusing to readers not familiar with the technical vocabulary of the field, as well as failing to convey the similarities that these terms do share with English terms such as "canon" or "scripture."[6]

Another recent example of this issue is the well-known debate over how to translate the term "believing mind").[7] Some argue that there is sufficient overlap with the English word "faith," with its rich history and multivalence, to justify translating shinjin as "faith"; others argue that "faith" in a religious context implies belief in an almighty God (among other things) and that use of the word would pull the hearer in a direction that would be misleading for the Buddhist context. Those in the second camp use the transliteration shinjin, in the hope that it will eventually enter English on its own, keeping all its original implications.

Again, in dealing with the term "mind," Herbert Guenther warns that language "is a treacherous instrument":

If it is already difficult to know what we mean by these terms 'mind' and 'mental' in our own language, it will be readily admitted that it is still more difficult to ascertain the meaning of what is translated by 'mind' or 'mental' from Eastern texts. The question, whether the authors of the original texts actually meant the same as we do by those words about whose meaning we ourselves are not quite clear, should always be present, not only when translating texts but still more when dealing with a systematic presentation of Eastern philosophies.[8]

In the case of explicitly religious texts, terms can pull the author in a certain direction, sometimes in a way that the translator cannot figure out quite what it is that is guiding the flow of the text or the direction of thought. An awareness of this process may help clarify, or at least relieve anxiety over, passages in which it seems there is no consistent line of thought, or where the argument seems to jump over itself. In any case, some things may simply be lost irretrievably in the past, and it is best to keep this possibility in mind.


Finally, and not unrelated to the above levels, is the influence the intended audience has on a text. As mentioned above, I found that preparing a presentation in Japanese for a Japanese academic audience of Buddhologists, and preparing the "same" paper in English for a more general but Western academic audience, affected the content and flow of the paper. Before a Japanese audience one can assume a certain level of knowledge of technical terms that one cannot always assume for a Western audience. On the other hand, one can assume a greater interest among a Western audience in things like general hermeneutical issues, or the history of Buddhism beyond the Sino-Japanese developments. This colors not only the details one chooses to include, but also the direction one's train of thought takes.

With a translation, there is not only the question of the intended audience of the original text, but also the intended audience of the translation. How much knowledge does one assume on the part of the reader? Does one aim for a strict, literal rendering to remain "true" to the text, or does one aim for a rendering that reads smoothly and meaningfully in the "host" language. How much "extra" information needs to be provided to make the English rendering as intelligible as the original was to its intended audience?

I have often come across passages in my translation of the Mo-ho chih-kuan where the author, Chih-i, refers to analogies or texts with short, cryptic phrases that do not make any sense until one is familiar with the original source behind them. For example, Chih-i's analysis of a certain meditative state (Mo-ho chih-kuan, T. 46.12c) closes with the following exhortation:

If people do not cultivate such a method [of meditation], they forfeit immeasurable, valuable treasures, and [this is a cause for] both humans and gods to grieve. [Their loss] is as if a person with a stuffy nose sniffed sandalwood and could not smell it, or is like a rustic man who [ignorantly] offers [only] one ox for a [price-less wish-fulfilling] mani jewel. (T 46.13a21-23)

Both similes—the person with a stuffy nose and the rustic man—are references to a series of analogies found in the Pratyutpanna-samadhi-sūtra, and can only be fully appreciated by referring to the original source. Chih-i seems to have assumed that his audience would immediately recognize and understand his images, much the same as a modern audience could be expected to supply the emotional and imaginative context needed to understand phrases such as "crying wolf", "finger in the dike", "barking up the wrong tree", "a material girl", or "Butt-head". But when faced with phrases such as "a rustic man offering an ox" or "seeing seven jewels and one's relatives in a dream and rejoicing", a modern reader cannot make much sense of these without some help.

This leads to a further question. When Chih-i summarizes, or picks up certain phrases and omits others, does he pick up only what he thinks is important, or does he assume that his readers or listeners are familiar with the context and will know how to fill in the details on their own? Is he deliberately emphasizing certain points, or does he intend his summary to stand metonymously for the whole? In some cases, such as the passages cited above, it is obvious that he is using a kind of shorthand for a fuller context known to his audience. But this is not always the case. In either case, the modern reader is likely to be at sea without additional information to understand and interpret the text. In such cases, a merely "accurate" literal translation captures at best only the surface meaning, and at worst leaves only a meaningless jumble of words.

It is not always a simple question of "right" and "wrong." The translators' choices are made through a combination of a number of factors: consistency with previous choices in translating these or similar terms, maintaining a balance between literal meaning and clear English rendering, judgments on how much to rely on explanations through footnotes, perceptions of the needs and wants of the audience, and so forth. It is the translator's art to make these choices elegantly.

I conclude this short essay with a few remarks on the difficulties, joys, and rewards of translating religious texts, and on the importance of the task. I am convinced of its value and the need for more translation to be done. It is, after all, what I spend a great deal of my time on. At the same time, if you have not already guessed, translating is often a dreary and difficult task, overshadowed by the constant realization that perfection is beyond one's grasp. It means long hours of sifting through the dry dust of ancient texts, never quite sure even about how much is being understood (or is possible to understand), always aware that one is perceiving only a partial, warped, and hazy reflection in a darkened glass.

And yet there is so much to be done, so many important religious texts that remain to be translated, so many puzzling words and phrases and ideas that need clarification, so many treasures waiting to be "exhumed." Translating religious texts is, after all, much like an archaeological dig: many hours of sifting through the dust with often meager results to show for one's efforts. The results are often uncertain and ambiguous, the work often frustrating and onerous. Nevertheless, the goal—to create successful and meaningful translations—is not hopeless or futile. Translations are possible wherein we can be confident that the original is accurately conveyed (if not fully, at least satisfactorily). The process offers special moments filled with the joy of discovery, and the results, I still hope and believe, offer the reward that the accomplishments are worth pursuing.

2. Philosophical Texts: James W. Heisig

Perhaps the main reason philosophical texts are not widely read in Japan is that they are not written to be widely read. Quite the contrary, they are written to be classified as sound philosophy or as solid contributions to the history of philosophy. The keepers of the classification are the older generation, who were so classified by the generation that preceded them. Its journals are for specialists and as such mirror the every-increasing narrowness of specialization. Simply put, the system is self-closed by definition, and maintains its vitality in proportion as it increases its closure and exclusiveness. Like the uroboros that swallows its own tail, institutional philosophy feeds off itself, as if in the effort to grow as small as possible and eventually disappear.

All indications are that it is succeeding. In universities across the country departments of philosophy are shrinking or simply being absorbed as curricular specializations in other departments. Despite the swing towards generalized education in liberal arts programs, the amount of philosophy read in classrooms has declined dramatically. Opportunities for being guided in the reading and discussion of the classics of philosophy continue to dwindle. And even in philosophical curricula proper breadth of exposure to the richness of the tradition continues to lose ground to the fetish of concentration on particular thinkers.

The maldistribution of philosophy feeds its growing elitism, much the same as the maldistribution of food and clothing among the poor transforms those who have it in abundance into an elite. What ought to be common possessions have become luxuries. In short, what we have here is a transgression against the basic spirit of philosophy. And because the transgression is institutionalized, the fault does not lie principally with individual students of philosophy, but with the dominant myth that they have inherited. The only hope is in some form of de-institutionalization. Institutions, after all, have a way of growing to the point that they actually begin to work against their founding purpose. If it is the case that the philosophical establishment in Japan has crossed that critical threshold and is actually promoting the ignorance of philosophical thinking, then nothing short of a demystification of the dominant myth can restore its original spirit. If philosophy has fallen into a rut in Japan and failed to produce sufficient numbers of original thinkers capable of making an impact on the general modes of thought of the age, surely the bulk of the explanation lies within the general perception of the philosophical vocation itself.

The circumstances of philosophy in Japan today are nothing new to the history of Western philosophy. There is hardly a single major movement from the pre-Socratics to the present day that has not had to contend with accusations of elitism or snobbery for its peculiar and unintelligible use of language.[9]The reasons often ride on the shirttails of other complaints about social insignificance, political naïveté, and the like. But the complaint about the failure of philosophers' language to communicate can be pulled away for a closer look.

There are two distinct but related questions here: how bad writing corrupts thinking, and how esoteric language inoculates thinking against criticism by outsiders. In the case of Japan, where imported philosophy outweighs native production, these questions immediately draw us in to asking about how philosophical texts are translated, and this is the standpoint from which I would like to think about them here. In fact, many of the problems with the decline of philosophy may begin from the fact that this is so little discussed, or at least that the discussion has so little influence on the young generation of translators. To be fair, I know of no encyclopedia of philosophy, in any language, that treats the problem of actual translation of texts as a philosophical problem. At any event, in Japan's academic world, translation is seen as a technical issue, not a proper philosophical question. Footnotes and glosses in translations about the subtleties of the original text typically are great in content and show an admirable grappling with the content of the text. But they rarely go beyond arguing for the rendition of some term or other or demonstrating the translator's competence and trustworthiness. More than that is not asked, and it is almost impossible to judge what if any translation theory is at work. As far as I can tell, this is typical of the genre of philosophical translations as a whole.

Against this background, I would like to argue the case for a radical liberalization of the standards of philosophical translation in Japan. It is time great numbers of aspiring philosophers were set free to err on the side of creativity and rhetorical elegance, which have been longstanding victims of the largely tacit but powerful assumptions regarding translation. The step is an audacious one only because it is unfamiliar. Once taken, however, I am convinced that it will help to free the thinking of the young generation of philosophical minds who typically begin their careers with translating texts, and at the same time increase the reading public of philosophy. Accordingly, the object of my argument here will be the sacred cow of fidelity to the original text.

The idea that texts are more beautiful, or at least richer, in the original is a truism that no translator of philosophy would dare challenge in public, but it does not settle well for either readers or translators. No doubt the absence of translation is by far the more compelling reason to read texts in Western languages. Communicating with scholars from abroad and publishing one's papers in foreign journals is one thing. Grappling with philosophical ideas in one's own is another. Even where one has a fairly good mastery of the languages, the associations, connections, and reflections prompted by reading in Japanese far exceed the stimulus of a text in a Western language. The question is why the valuation of translation does not take these more into account.

Even before we talk of liberalizing the idea of translation, it has to be recognized that Japanese translations of Western philosophical texts are full of mistakes that can be traced back to an insufficient understanding of the original language. Examples of failure to understand grammar and idiomatic usage as well as the historical echoes of particular words and metaphors are commonplace. Ordinary language gets converted into technical jargon and technical terms lose their links with other branches of learning, and as a result even the aim of being faithful to the original, independently of where the Japanese reads "naturally" or not, is not met. The distinction between elegant prose and bad prose is erased; the flowing stream of James and Bergson are made to read like the clotted prose of Adorno and Heidegger.[10]

Before you accuse me of gross exaggeration, let me state another, equally obvious fact: there is nothing particularly Japanese about this. Western philosophy has been producing its share of bad writing and bad translations for centuries, and has never been without its critics for doing so. (Even the word translation is a mistranslation.[11]) I find no reason to single Japan out here for a slap on the wrists, and have no doubt that a solid counter-argument could be made about the translations of Eastern philosophical texts by Western scholars. If there is any difference, it is that the prolonged alienation of philosophy from the intellectual mainstream has hardened its stylistic habits into a grounds for self-identity. It is hardly my place to issue a call for repentance. All I can do, with one foot in Japan and one foot outside, is try to identify the philosophical reasons why this state of affairs is allowed to continue.

In doing so, I mean to resist the temptation to lay the bulk of the blame on the psychological, social, and educational deficiencies of Japan's linguistic culture. The web of dominant conventions that govern an academic career in Japan are part of this culture, and the translation of Western philosophical texts cannot be under-stood without it looking at standing demands that really have nothing to do with the content of the translations. Not even the contents of the philosophy being translated are likely to overturn the myth or break through it, because it is prior to the translator's work and gives it a place in the social relationships. This myth is a kind of arché whose criticism amounts to a kind of anarchism. The standards of professional certification will, of course, have to loosen and change before philosophy enters the mainstream of Japanese intellectual life. But nothing I have to say about the matter, and certainly nothing in a talk as short as this, is likely to advance the process. I therefore choose to look at the execution of translation as a philosophical choice rather than a mythical one.

I have no intention here of trying to make any contribution to "translation theory." To do so would be to stray from the far simpler objective of arguing for the liberalization of philosophical translations. I would only note in passing the growing awareness during the twentieth century that translating from one language to another needs to be understood in the wider context of what George Steiner called "inner translation," that is, the semiotics of hearing what people say and saying what one thinks. Consciousness creates a certain disequilibrium with the world. Reflection processes the world not as the fact of what is but as what it might be; we are always reading into what perception gives us, and this builds up a pressure of frustration as the world resists our hopes for it. Speech is our way of keeping that pressure from exploding.[12] While the need for speech—the translation of what we say to ourselves into what can be communicated to others—is universal to consciousness, its definition, both in amount and in content, is cultural and temporal. The cultural difference is well known to easterners who have lived in western countries and vice-versa. What is too often overlooked is the fact that a similar disequilibrium comes into play when I read something written before I was born. I translate it, even if it is in my native tongue. In fact the past is a foreign country, whose distance from us is perhaps even more than that which separates the con-temporary language of Europe from that of Japan.

The point for us here is that, when translation between languages is not aware of these prior levels of translation, in effect it projects all the pressures towards the foreignness of the foreign language, which greatly contributes to its gaining an inviolable character. An extensive enterprise of philosophical translation like Japan's should do something to heighten the awareness of these questions. If linguistic theory stops at the relation between thought and expression, and translation is seen as largely a technical question, the bridge between translating and thinking is weakened. In the same sense in which Goethe's poetry is unthinkable without his efforts at translation from Romance, Slavic, Iranian, and Germanic tongues, I believe Japanese philosophy will never mature until it becomes more self-aware of what is going on when it translates.

I am not going to be detained here by arguments that translation from one language to another is out and out impossible. As Ortega y Gasset rightly notes, translation without interpretation is a naïve fantasy, and surely not everything is translatable. But interlingual translation is no more impossible than the transition from ideas to speech, where what is held in silence is important to understand what is communicated, but which we negotiate all the time in varying degrees of success.[13] Formal arguments against the translatability between languages have accumulated at least since the fifteenth century, and while there is good antidote there to mechanical theories of translation, the level at which the final position is true is uninteresting to philosophy.

Self-criticism is the soul of philosophy. And as Whitehead used to tell his students, "to be refuted in every century after you have written is the acme of triumph."[14] I would add: to be refuted in several languages only sweetens the victory. Nothing finite is self-supporting and philosophical problems are no exception. Translators who enshrine a philosophical text in the contingencies of its birthplace in the effort to give it an infinity beyond the reach of the time and culture of the language they are writing in are claiming an infinity for it that will only kill it in the end.

Complaints about bad writing have, as I said, long accompanied philosophy. But only rarely is the nature of translated philosophy taken up as a serious part of self-criticism. There are, of course, those who champion dense and halting style almost as a philosophical virtue.[15] By far, however, the majority of great philosophers who have bothered to write about style have done so to applaud clarity and berate obscurity.

Such opinions are much more acceptable when directed at original texts, but somehow translations have had a privilege of exemption in philosophical circles that they have never enjoyed in literary ones. Vague and confused translation language is assumed to be the fault of the translated language, and there the matter ends. The problem is, the reader of the translation almost never trips over the style at the same places as the native reader of the original. In other words, the translator's policy of "non-interference" and "objectivity" can only be based on a lack of understanding of the enormous amount of interpretation that goes on in translating between languages. This is so independently of the quality of the original style. Indeed, reproducing the same quality of bad writing in languages as different as Japanese and European languages, would take the highest literary skills, perhaps even higher than carrying over the flavor of a translucent, flowing style. Few if any translators of philosophical texts possess this, and it is not reasonable to ask it. But neither is it reasonable to swallow without criticism the idea that translations that are tough to plow through are the result of either a flawed original or the distance between the two languages.

A translator sanctifies the text out of misplaced respect for the author. The amount of effort that goes into producing a translation only heightens the respect, and few translators would affront common sense with the arrogance to stand shoulder to shoulder with the text with the thought of improving it. This posture of enchantment before the original text is precisely the cause of the disenchantment of readers with the resultant translation. When a text is difficult to understand, it is assumed that the original is difficult. To the extent that the translation stumbles and grates on one's native sensitivities, there is no repressing the feeling that the translation is flawed, but even this does not bring the original into question. If anything, the flaws in the reproduction make the original shine all the more, like a distant and unapproachable star. This seems to me getting things backwards. Any sense of reverence communicated through a translation that tolerates irreverence towards one's own language and one's own demand for clarity is simply misplaced. And this can happen only because of the shared assumption that the work of translation was done in an objective, non-interfering manner. What is more, it all but removes the possibility of translation leaving a mark on literary style, the way, say, translations of Shakespeare left an indelible mark on the German language and introduced his name into classical German literature, or even the way nineteenth-century Japanese had to make grammatical adjustments in order to accommodate translations of foreign texts into the language.

In classifying this as a kind of sacralization, I mean that philosophical texts are being misclassified. Homer's epics and the Koran are good examples of quasi-sacred texts, whose translation merits the kind of respect it seems to me Japan accords ordinary philosophical works, and also from the comatose state texts are reduced to in order to be translated "faithfully." Their very survival across time sets them off from ordinary historical discourse. The appropriate form of translation for this is literal, the belief that the word-for-word technique is the ideal way of submitting oneself to the original text and eliciting the full meaning of the text. Very little, if any, classical Western philosophy belongs in the category of the sacred text in this sense. For the translator to take it as such is to make a fundamental hermeneutical mistake. I have the impression, however, that young students of philosophy in Japan, hoping to make a career in the discipline, take this sacralization as a matter of common sense. It further seems to me that this fixed idea of what constitutes a "faithful" reproduction of a text not only does not broaden the reading audience for philosophical texts—which is, after all, the point of translation—but actually stimulates philosophy's appetite for swallowing its own tail.

Based on what has been said, desacralizing philosophical texts means adjusting the current notions of what constitutes "fidelity" in translation. For purity of argument, let us assume an accomplished translator—that is, someone who does not need the translation. He can read the original with relative comfort. Such a person knows there are better ways to come to grips with a text than the arduous and often unrewarded task of translating. Aside from earning credibility as a translator, the point of the translation is to make it accessible to those who would not otherwise have access to it, or at least who would prefer reading a work in their own language, even a clumsily worded version of it, to reading the original, even though they may occasionally return to the original to confirm critical passages or check an oddity in translation. This being so, it is only natural that the translator's idea of fidelity should coincide with the fidelity expected by the reader: an accurate reproduction of the surface of the text in a second language that can stand up to the critical eye of those who compare it with the original. Interpretation and paraphrase, it is assumed, should be left as far as possible to the reader. The greatest fear of the reproducer is that he will not feel as comfortable in the text he has traveled to as he would like to be, and that under the obligation not to leave anything behind, he will carry his misreadings back to the native soil of his own language, often unaware of the mistakes he is making.

But this is not the only reason a translation can go bad, though the fear of erring in this respect is so real that it often obscures other, equally important reasons. One can also be unfaithful to one's readers by presenting them with a hybrid prose on the assumption they will be able to see through to the alien grammar behind it and then chalk up the offense done to their native language as a necessary evil. The catalog of such sins makes interesting reading—especially for the Japanese student of philosophy who is likely to have his own list ready to hand—but repentance is seen to be unrealistic. Why bother, when there are no serious consequences to one's reputation as a specialist in philosophy for not doing so? Individual consequences, perhaps not. But consequences for the way philosophical texts are read in the intellectual mainstream, and hence for the future study of philosophy itself, enormous. Besides, it is unfaithful to the original. When one wrestles with a translated text, one is at least doing what one does when one struggles to grasp the connections, the flow of the argument, the association of ideas, and the subtle implications that do not reach the surface of the text in an untranslated original in one's own language. To be denied this is to forfeit even the minimum expectations one has when writing one's own philosophical prose. Willy-nilly, the impression can hardly avoid building up over time that philosophy is something cut off from the way language works in general.

This leads to a third, but somewhat subtler form of infidelity in translation. Here we have to do less with the particular text at hand or the readers who will be handed it in their own language than with the failure to see how questions of translation are themselves fundamental philosophical problems. Simply put, as I have been insisting, the translation of a philosophical text is faithful to philosophy itself to the degree that it is aware of the role of language in communicating thought; and to the extent that it is not aware, or does not allow its awareness to interfere with the translation process, it is unfaithful.

The range of problems that language presents to the expression of philosophical thought is broad, but here I would like to consider how dealing with them affects the actual work of translation. To begin with, there is an awareness of what is at stake in killing off polysemic elements. All neologism—be it by distorting language into nonsense, combining existing languages, or creating new terms—is aimed at controlling polysemy, which is something natural to language. At the opposite end of the spectrum there is the developmentally rich polysemy of humor, irony, and sarcasm, without which a great deal of the classics of philosophy gets glossed over. Here polysemy is a form of hermeticism that includes the reader, and a translation that is unaware of the fact is likely to reproduce it as an exclusive hermeticism, that has the neologism's effect of restricting access to the initiated. The multiplicity of meanings can be hidden in a term, in a phrase, or in the flow of the argument. Which is primary will depend on the context, but without attention to all three, the layers of meaning are likely to be lost more often than preserved. To put it radically, insofar as one can read a philosophical translation and reconstruct the original from the surface of the text, the original has not been understood and that translation is incomplete.

Second, there is the problem of leaning on existing translations from a third language, increasing the possibility of repeating mistakes. This is very common in Japanese translations, especially of classical texts but also including philosophical works. Time and again I have found mistakes in translation that could not have come from the original but only from a misunderstanding of a peculiar English usage. Einstein said that a genius is someone who is good at concealing his sources. I suspect that this applies to not a few of those in the pantheon of Japan's great translators. In any case, I think we have to look at the assumptions behind this use of other translations for the assumption that everything open on one's desk is somehow removed from the living stream of language and that attention to the surface of the text is adequate. (When it comes to the concrete question of how this affects Japanese prose style, I find myself often standing out in the margins looking in, and must therefore defer to those who can move more freely between the lines of the text. Though even my limited acquaintance is enough to give me a sense of discomfort, often enough I trip over language simply because it is too good for me.)

Though I am highly cautious of consulting existing translations in a third language, there is one case in which it is most helpful, namely to serve as a supplementary lexicon for individual terms that cause difficulty. Dictionaries themselves are another matter. They are the daily bread of the translator, but they are not idols. They are to be devoured, not worshipped. I have the impression that trust in their omniscience, or at least irrefutable authority, is the closest thing to original sin in the world of Japanese philosophical translations, though they are perceived as a via salvationis for those wandering aimlessly in the forest of words. That said, I think that problems of infidelity to the linguistic dimension of philosophical thought are exacerbated not because one relies too much on dictionaries, but that one relies too much on too few of them. To the native, words are always more than the sum of dictionary definitions. To the translator, always less. One way to compensate for the imbalance in the way the translated language and the translating language face a text, to break free of belief in the infallibility of the bi-lingual dictionary is to tem-per their use with etymological and historical dictionaries of both languages being studied. But even this is not enough. To assume that, given the suitable capacity, anything from two centuries ago can be captured in one's own native language leads not to accurate translation but to the paralysis of style. Language, after all, is not dead—unless you kill it, and then it is no longer language. To all appearances, philosophy in Japan is a mass grave of such executions.

In this same regard, I find appalling the growing habit of introducing foreign words into a text as a solution to apparently untranslatable key terms. This belongs to the general failure to appreciate the style of the original. The fact is, Western philosophers often write badly and use strange terminology to cover their faults, but this is no excuse for writing barbarous prose in one's native language out of a sense of "faithfulness" to the original. The translated text of a Western philosophical work is, after all, a new language. It is not simply an "equivalent" rendition of one language into another. The struggle to find everyday, intelligible expressions for alien idioms and grammatical usage is a contribution to language. Just as children, the oppressed, the excluded, minorities, and so forth rebel against dominant forms of language, so is the introduction of a foreign thought into one's own linguistic world an interruption of the status quo. To ignore this, or pretend it is not happening, is to displace language from the only place where it can live and breathe.

I began speaking of fidelity with the assumption that the translation is not needed for the translator. Actually it is, in three very different senses. First, the majority of translators only really read and understand the book, even in a surface sense, once they have translated it and re-read it in their own language. There is a difference between reading 10 pages in 10 or 20 minutes, which a fluent reader would do, to reading 10 pages in 10 or 20 hours, which I suspect a high percentage of translators do. This being the case, it is unreasonable to expect that even the minimal "feel" of the flow of the text can be translated. The river flows so slowly it is virtually frozen. This is part of the reason why only a fraction of philosophical translation is great, most of it passable, and a solid mass of it downright awful.

The conclusion I draw from this is alarmingly simple. I am not suggesting that one subtract anything from the translation, leaving out what is unclear or too difficult to render. Neither am I suggesting that one add phrases and sentences along the way to clarify the meaning. I find both these practices appalling. The addition and subtraction I have in mind is of a different, less invasive sort.

First, I would stress the need to add the stage of radically editing a completed translation for readability. Much translation is not bad because it is inaccurate in a first sense, but because it is incomplete, a first draft that deserves to be poured over and rethought with the same care that a good writer gives his own prose. This is a courtesy to the readers and also, as I have been insisting, a courtesy to the original text.

Secondly, there is a need to subtract the style of translated philosophy from one's own writing style when composing one's own philosophical texts. The permanent temptation in philosophy, a temptation which I stated at the outset is fast becoming a chronic condition in Japan, is that its idiom becomes a kind of obsolete dialect. The tendency of philosophers to focus their efforts on dealing with each other's writings rather than with the fundamental problems of philosophy has to be resisted as part of the devotion to self-criticism. I do not mean to suggest by the foregoing that all infidelity in translation is destructive. There are also mistakes and misreadings of texts that make possible entire new ways of seeing and thinking.

Faithful translation, at least as I am understanding it here, always involves some balance of mimesis and poesis, between the attempt to preserve the original vitality of the text by trying to enter in and repeat the experience of the author, and the attempt creatively to read it from one's own point in time (what Nietzsche called erdicten). While it is a matter of philosophical style how one strikes the balance, both are different again from the mere mechanical reproduction of the surface I have criticized above. Original philosophical texts are always closer to a musical score than they are to a bouquet of flowers. The music can be played again and again, with varying degrees of interpretation but never purely. The only kind of flowers that can be safely translated across time are dry flowers, and this is because they have been cut off from their roots.

In either case, translation creates waste; it always diminishes the original, even when the style is an improvement in a literary sense. This is not simple falsification, but belongs to the same drive towards the future that makes all translation necessary, beginning with the translation of one's own thoughts and desires to oneself and others. There is always "more than words can tell," a mythical element in all logos. Mistranslation is one kind of lie; good translations are another. But both fragment and destroy in order to rebuild. The attempt to avoid all such deformation, or pretend that it can be avoided, is by far the greater lie.

That said, translation is tempted by two forms of betrayal, each of which is a form of linguistic madness. On the one hand, there is the belief that too much is forfeited for it to be done, and the perfect translation would be to teach people to read the original. The extreme case of this is Borges's Menard, who struggles so long with the text of Don Quixote that he ends up reproducing it word for word in the original. On the other hand, there is the belief that the text belongs to the translator and his age, that its native context is no longer relevant. In the extreme, the loss is ignored and the book read as a contemporary work. The text becomes like the prisoner in Paul Valéry's Histoires briseés who is exiled to a land where everyone knows him as someone he is not, and whose only salvation is to forget who he really is. Most translation falls somewhere in between.

When it comes to philosophical texts, surely some writers suffer in the translation more than others. For example, I have argued that the writings of the Kyoto-school philosophers Nishida, Tanabe, and Nishitani, do not suffer in the translation, surely nothing of the scale of what great stylists like Bergson and James suffers in Japanese. What is more, there is a sense in which for their permanent contribution to philosophy to be secured they must be read in translation, and these readings must be allowed to reflect back critically on the readings of those who work with the original texts. Despite all my complaints, I am persuaded that what philosophies lose in translation is generally trivial compared to what they gain. There are translations so bad that nothing happens at all, except that it is ignored. But a mostly competent translation is an event at least as important as the fact that the books are still read. The real issue of translation does not require the ability to do the work. It is self-evident or it is esoteric. I believe it is the former, and that twentieth-century Japanese philosophy, particularly the philosophy of Nishida and his leading disciples is one of the clearest examples of this.

If philosophy were only the history of philosophy, perhaps the need to desacralize our translations would not be so great. But insofar as philosophical texts excite the mind to connections not previously seen and enlighten aspects of the present that would otherwise go unnoticed, to pretend that their translation is no more than a crutch for the linguistically impaired is to forfeit the soul of the translator's vocation. Translation is not just memory, it is also anticipation. And are these not the two impulses that combine to pull us out of animal consciousness?


BLANSHARD, B. On Philosophical Style. New York, Greenwod Press, 1969.

CLARKE, J.J. Jung and Eastern Thought: A Dialogue with the Orient. London and New York: Routledge, 1994.

GAUNTLETT, D. Media, Gender, and Identity. London: Routledge, 2002.

GÓMEZ, L., Shinran's Faith and the Sacred Name of Amida. In: Monumenta Nipponica 38, 1 (1983): 73-84.

GUENTHER, H. Tibetan Buddhism in Western Perspective. Berkeley: Dharma Publishing, 1989.

KASULIS, T. Review of Letters of Shinran. In: Philosophy East and West 31, 2 (1981): 246-8.

ORTEGA Y GASSET, J. Miseria y esplendor de la traducción. Obras completas. Madrid: Revista de Occidente, 1970, vol. 5.

STEINER, G. After Babel: Aspects of Language and Translation. London: Oxford University Press, 1975.

SWANSON, P.L. What's Going on Here? Chih-i's Use (and Abuse) of Scripture. In: Journal of the International Association of Buddhist Studies 20, 1 (1997): 1-30.

SWANSON, P.L. Apocryphal Texts in Chinese Buddhism: T'ien-t'ai Chih-i's Use of Apocryphal Scriptures. In: Canonization and Decanonization. Leiden: Brill, 1998. 245-55.


[*] PAUL SWANSON is a permanent research fellow and Director of the Nanzan Institute for Religion and Culture in Nagoya, Japan, where he has been since 1986. He is currently engaged in an English translation of the Chinese Buddhist classic "Mo-ho chih-kuan" (The Great Cessation-and Contemplation).

[**] JAMES HEISIG is a permanent research fellow at the Nanzan Institute for Religion and Culture in Nagoya, Japan, where he has been since 1979. He is currently engaged in the preparation of a sourcebook on Japanese philosophy, a field on which he has written widely.

[1] It could be pointed out that there is never an exact one-to-one correspondence between different words of the same language, but this would bring us into the broader realm of linguistics and meaning rather than the specific question of translation between languages.

[2] J. J. Clarke writes, "As the American logician Quine has reminded us, there lies at the heart of any attempt to translate from one language to another, a radical and inescapable indeterminacy, for we have no standpoint outside of language from which to judge the adequacy of the procedure, and no access to 'meaning' other than through specific languages. This question is especially urgent in the translation of Eastern philosophical texts…" Jung and Eastern Thought, p. 38

[3] Some characteristics of "dictionary fundamentalism" are the commitment always to use the same word to translate the same term regardless of the context, and to reject the use of a word (or neologism) because "it ain't in the dictionary."

[4] I have attempted to address these points in a previous essay; What's Going on Here? In: Journal of the International Association of Buddhist Studies, pp. 1-30.

[5] Specific examples of this are given from my attempted translation of Chih-i's Mo-ho chih-kuan in my article "What's Going on Here?" 25-7.

[6] I have addressed the question of applying such terms as "canon," "scripture," and "apocryphal" to the con-text of Chinese Buddhism elsewhere: Apocryphal Texts in Chinese Buddhism. In: Canonization and Decanonization, pp. 245-55.

[7] See, for example, L.GÓMEZ's review article, Shinran's Faith and the Sacred Name of Amida. In: Monumenta Nipponica, pp. 73-84 (especially pp. 81-4), and T. KASULIS's review of Letters of Shinran. In: Philosophy East and West, pp. 246-8, and the reply of the translators in the same journal 31/4 (1981): 507-11.

[8] H. GUENTHER, Tibetan Buddhism in Western Perspective, pp. 37-8.

[9] The bulk of my attention here is to the present day, but even the great vernaculizers of the Middle Ages who broke with the convention of writing only in Latin, were aware of this tendency. Did not Dante identify with the souls suffering in the first terrace of purgatory because of his pride of learning, his tendency, as Giovanni Villani comments five centuries later "to be rude, as philosophers are, and not know how to speak with the unlearned"? Cronica di Giovanni Villani, ed. by F. Gherardi Deagomanni, 4 vols. (Florence, 1844-5), IX, 136. Or again, when we see Ramon Llull a century before translating his own books between Arabic, Latin, and Catalan, composing abbreviated and simplified versions of his own complex texts, and alerting his reader to the different levels within a single book (such as in the opening remarks to the Llibre del gentil i dels tres savis) it was precisely because of the ill repute in which philosophers were held by ordinary people.

[10] There is, of course, the argument that elegance impedes clear philosophical thinking, so that someone like Brand BLANSHARD can come down hard on Nietzsche, Wittgenstein, and Kierkegaard for the fact that their style cloaks unclear thinking. On Philosophical Style.

[11] Leonardi Bruni (1369-1444) misread a line in the Noctes Atticae of Aulus Gellius where traducere meant "introduce, lead into" as "carrying over" and hence "translating." The etymological mistake carried over to French and Italian in the fifteenth century and was simply repeated in English but covered over in the German Übersetzung.

[12] George STEINER's After Babel is a masterly review of the field and has probably influenced my remarks here far more than I shall credit him. The delicious irony of the Japanese translation is that the most telling examples of the book, which show a often brilliant attention to detail, are virtually nonsensical in Japanese, not through any fault of the translator but because Steiner requires a knowledge of French, German, and English for his argument to be followed.

[13] "A being incapable of renouncing the saying of many things would be incapable of speaking. Every language has a different equation of manifestations and silences. Every people keeps silence on certain things in order to be able to say others. For everything would be unsayable. Hence the enormous difficulty of translation: in it one tries to say in one idiom precisely what the language tends to silence." J.ORTEGA Y GASSET, "Miseria y esplendor de la traducción." Obras completas (Madrid: Revista de Occidente, 1970) vol. 5: p. 444.

[14] Science and Philosophy, 122.

[15] The American gender theorist, Judith Butler's appeal to Adorno in her defense is a contemporary case. Though I esteem her ideas highly, I have a certain sympathy for a recent critic when he writes: "Her prose is unnecessarily dense and long-winded, and almost never fails to use jargon even where much more accessible vocabulary is available.… However, although Butler's writing is like an explosion in a dictionary factory, if one takes time to dig through the rubble one finds that her ideas are actually quite straightforward." D.GAUNTLETT, Media, Gender, and Identity, ch. 7.