According to Walter Benjamin, the more difficult a text is, the easier it is to translate. Miller writes: “Translation, he [Benjamin] says, is ‘difficult’ in inverse proportion to the ‘translatability’ of an original; thus, a high degree of ‘translatability’ means little difficulty for the translator,’ while a low degree of ‘translatability’ means that the task of the translator will be proportionately difficult” (Miller, 471). Simple texts, claims Benjamin, only frustrate.
Taking Benjamin’s premise and applying it to the Japanese tradition, Roy Andrew Miller reasons thus: Since, as Benjamin asserts, “translations [are] untranslatable,” and since the entire canon of Japanese literature (and culture and expression, for that matter) is a translation of a translation of a translation ad infinitum, not a single work from the canon of Japanese literature is translatable. Miller traces the Japanese tradition back to the early Chinese sources, namely the Ming fantasies and fictional works, and the poetry of Po Chu’I. From this starting point, he follows the importation of the Chinese tradition into Japan, marking its development through the various stages -- waka, Nō, bunraku, kabuki, and, more recently, the shinpa theater -- all of which together form what he calls the Japanese "canon of taste."
Faced with the task of making relevant the past, each generation must naturalize the foreign text for a new age. The whole Japanese tradition, Miller asserts, is in essence a retelling of Po Chui’s “Ballade of Endless Remorse.” “Japanese language, Japanese culture, and the expression of both in Japanese literature," he explains, "are all themselves ‘translations,’ and hence, in terms of Benjamin’s axiom, theoretically impervious to translation” (473).
Though the "matter" may be borrowed from a foreign literature, the fashion in which it is presented is markedly different from the earlier versions. “The overwhelming concern of the [Japanese] culture in its literary expression has been with how something is said, rather than with what is being said. It is a language, a literature, and a culture of style, form, rhetoric, genre” (473). It might be illuminating to compare this aspect of the Japanese tradition with Plato’s distinction between form and content presented in Phaedrus, in which he makes the distinction between the ‘what’ (content) and the ‘how’ (form, genre). Perhaps the lines are more blurred in the Japanese tradition than they are in the West. One might use Genji monogatari as a prime example for demonstrating this.
Literature of the modern age, Miller continues, can be seen in similar terms. While pre-modern Japanese literature was an endless retelling of the earlier Chinese classics, the development of modern Japanese literature has been a similar process of translation and re-translation -- only this time importation came from the West. Using Sōseki’s Michikusa as an example, Miller shows how post-Meiji Japanese literature has, to a large extent, been an attempt at interpreting Western literature through the retelling of it.
Summing up, Miller writes,
"Japanese literature in particular, and Japanese language and literary culture in general, are . . . a translation literature, and a literary culture of translation. Translation of translations is not difficult, it is impossible, as Benjamin has demonstrated. Japanese literature is translation; translation is Form, ‘genre,’ and so is Japanese literature. The translation of Japanese literature is therefore not difficult; it is impossible” (477).
Miller presents a convincing, if pretentious, argument, but he leaves unanswered three important questions. If, as he argues, translation of a translation (i.e., every work of Japanese literature) is impossible, then what, if any, good is it the attempt? Is he advising that we quit with translation altogether? And second, is this phenomenon of "translation," or re-telling, particular to the Japanese tradition, or has it not been a feature in the literary history of the West as well? And finally, if everything is a translation of a previous work, is tracking how each generation modifies, interprets and misinterprets its predecessors the only thing left for literary historians to do?
[Roy Andrew Miller`s essay can be found in Studies on Japanese Culture, vol II, 469-79]