One's Loss, Another's Gain:

The Challenges of Translating "I" in Japanese Academic Texts

1.0 Introduction

n translations of Japanese to English, issues of loss and gain are especially pertinent. Whether in translating honorific and humble forms or absent and inferred words, translators of Japanese are often faced with decisions on how to best fill in the large gap between the two languages, especially when translated text genres have not yet been clearly defined. This paper will discuss how cultural expectations of academic writing in both the Source Language (SL) and Target Language (TL) should affect how voice is constructed in translation. The translation of language with implicit relational or cultural cues as well as loss/gain issues will be discussed, particularly the problem of subject inference in Japanese and to what extent passive sentence construction should be used to translate such inferred subject constructions. The paper will first present model translations of problematic structures and discuss how these translation methods are or are not successful, and then discuss tactics that were used to overcome the same cultural problems in a translation done by this author. Methodology will focus mainly on the importance of genre considerations and a nuanced understanding of culture and genre when translating.

2.0 Background

This paper will discuss the translation of an academic report written in Japanese regarding English language instruction at the elementary school level in Niigata City, Japan. The work was contracted by the author of the paper, a Japanese elementary school teacher who hoped to present both a Japanese and English translation of his report to the Board of Education in Niigata as well as his own circle of Japanese elementary school teachers. The text was to be presented and published at a prefecture-wide seminar on elementary English instruction in Niigata prefecture. The translation was commissioned by the author to help add a sense of weight and legitimacy to the paper (as it was about English education). None of the other three papers in the journal was translated into English.

The full text can be found in the Appendix.

The key cultural problem with this translation lies in what is an appropriate register for a paper written for this specific collection of reports. Although the Japanese format was seemingly clear to the author, as no other English translations were produced (or former translations available), there was very little guidance available as to how formal translations of this type of essay were expected to be. Academic writing is widely translated from Japanese to English, as are personal narratives, but these translations differ greatly in register (as do their originals). The paper in question was unique because it shared elements of both academic and narrative writing. How the paper might be most accurately written in terms of register was most significant.

3.1 Loss in Inferred Subject Constructions

Several grammar structures in Japanese are notoriously difficult to render in English. One is the Japanese tendency to drop subjects from sentences when the subject can be clearly inferred. For example, if someone were to say, '寿司が好きです' the literal rendering would be, 'sushi (article) like. ' The Japanese listener would understand the subject 'I' is implied and the translator would render the sentence, 'I like sushi.' In Japanese, the exclusion of pronouns is frequent and often produces a sense of distance, humility and respect. Constantly referring to oneself as 'I' and one's companion as 'you' might sound conceited and confrontational, while excluding these words when they are not needed is much more tactful. This subject-exclusion doesn't create a problem of accuracy in Japanese-English translations (as the subject can often be easily inferred), but in the register of the translation.

Although simple constructions, such as the sushi example above, might not produce a great deal of debate over acceptable translations (few would argue that 'Sushi is liked' would be the most acceptable and natural rendering of the Japanese), in academic texts like the one discussed in this paper, the complete absence of the personal pronoun 'I' creates a much larger genre question of how to render the register of the text. With these subject-excluded constructions occurring again and again, any choice made by the translator will likely affect the whole register of the piece. For this genre, is it more appropriate to completely avoid the use of the personal pronoun 'I' through passive constructions, even if they prove to be unnaturally laborious and negatively affect the register? Or, should the translator sacrifice the register of the text and add the gain of a personal pronoun resulting in a more 'folksy' or conversational construction?

The following abstracts (and their translations) taken from the Japanese 'The Journal of Japanese Association of Occupational Therapists: Vol. 25, No. 6' [sic] provide good examples of these two paths.

Here, the translator (likely the author of the piece) has chosen to change the construction of a sentence that was initially in the active voice with an implied subject.

In the Japanese, it is not explicitly stated (or important) who did the videotaping, simply that it was done and the subjects were the objects. The English rendering is expressed in the passive construction 'Subjects were videotaped...' The translator is able to avoid inserting a subject by using a grammatically acceptable English construction that also leaves out the actor in the sentence without creating any question about who or what has done the videotaping. No 'I' or 'we' is needed and the translator continues to use the passive construction throughout the abstract.

The meaning of the two sentences is generally kept intact, and the translator has avoided adding any new words that do not appear in the Japanese. Unfortunately, there remains a problem in that there is a very rich and more complex passive construction in Japanese (Kinsui, 1997), which plays a key role in Japanese sentence construction, and it is not only a grammatical device, but also a key part of Japanese aesthetics (Sasaki, 2002:). This sentence is clearly not in the passive voice. Although no additional words have been added to the English translation, the grammar of the sentence has been seriously altered and some baggage that might be associated with a passive construction has been added. Although it is less tangible, the voice of the sentence is much different.

Here the translator has chosen to solve the inferred subject problem by simply including the implied subject in the English translation. The subject '私達は' (or 'we') is not found in the Japanese text and both of the verbs '行い' and '検討した' lack subjects. Again, as with the earlier example, since it is obvious to the Japanese reader who the actor is, the subject has been dropped.

In this case, the inclusion of the word 'we' does not sound especially awkward in the English translation. Moreover, a back-translation of the sentence with a subject '私たちは・・・検討した' is also completely acceptable. The meaning has not been changed at all. Unfortunately, the inclusion of the new word does change the voice of the abstract and the translator continues to use the 'we' construction throughout the text. The result is a translation that seems to emphasize the word 'we' as an actor, something that is absent in the Japanese text.

In both cases, the translator chose only one construction to deal with the inferred subject problem, creating voices that both do not seem altogether very natural in English: one being especially agent-less and one being especially agent-oriented. In both cases, the translator has not been able to effectively translate the meaning and voice of the text in a natural way.

All of these translations took an 'all-or-nothing' approach to subject-implied Japanese constructions and nuanced language that did not have a one-to-one equivalent in English. This paper will argue for a more integrated methodology that takes into account the consequences of voice in the English translation, rather than constructions that are either purely active or passive. If the translator is not tied to simply solving the same problem the same way in every instance, the result should be a much more fluent text.

4.0 Text Analysis

As noted above, the translated text evaluated in this report (see Appendix) is neither completely academic nor narrative. Consideration was taken to fully understand the purpose of the text and the intended readership before decisions were made about how constructions should or should not be rendered. Given that the text was to be read primarily by non-native speakers of English, care was taken to render sentences in a way that would be simple to understand and easy to compare to the Japanese text. The report theorizes about English education, but the main narrative of the text is of a teacher teaching his own students. Given the nature of the text, efforts were made to blend both active and passive constructions to create a text that was the most accurate rendering of the Japanese grammatical constructions, while, at the same time, being careful to create a style that could be read as naturally as possible. Additionally, since the translation was done in co-operation with the writer, more liberty was taken in issues of gain, since the author desired a text that was most natural for Western readers.

In this translation, active constructions were almost exclusively used to render active constructions and vise-versa, as they rendered the Japanese voice (and grammar) in the most accurate way. This required several important choices, namely, the decision to use the personal pronoun 'I' in describing the author's research. The personal pronoun '私' never occurs in the Japanese. The inclusion of the 'I' pronoun allowed for very simple rendering of phrases like:

Constructed as a passive sentence, it might be rendered, "English that is used for real, everyday communication was studied and a school life that overflowed with this 'Living English' was sought after". But overuse of this type of construction (which would have been necessary as the construction appeared again and again) was deemed unnecessarily awkward.

The resulting translation with 'I' ("Therefore, I studied English that was used for real, everyday communication, and sought to construct a school life that overflowed with this 'Living English'") clearly has a more conversational and personal tone. Whether or not the personal pronoun should be used in academic writing is certainly a valid question, but when presented with both options, both the translator and author considered the 'I' rendering to be the most natural and truest rendering of the Japanese that seemed to fit best with the register of the essay: a teacher speaking to other teachers about English pedagogy.

Cultural and genre considerations were the most important factor in making this decision. With no precedent and no guidelines for what might be appropriate or inappropriate in this particular journal, the author felt freedom to express his findings with the personal pronoun. The translator and author considered whether or not this would negatively affect the perception of the paper in English as being less academic. Certainly, this concern might be especially pressing if the text was to appear with other texts in a journal that discouraged the use of the pronoun 'I'. As the report was unlikely to be judged by these standards, it seemed most natural to render the sentences as they were originally constructed.

Ultimately, the decision to include the personal pronoun led to a gratuitous use of the word 'I' which, although it is an accurate rendering of the Japanese grammar, might have negatively affected the register of the text. Displaying the text by itself, without the Japanese, and having rendered some sentences in the passive voice might very well have helped the text flow more naturally. But as the English was intended to work in concert with the Japanese and was not intended to be read exclusively, it seemed most beneficial to leave the constructions the way they were, allowing ease of reading and consultation for non-native readers.

Several Japanese active constructions were rendered as passive. One example, the initial sentence of the text:

The passive form 'were taught' was used instead of the Japanese verb 'する' meaning 'to do' as the thrust of the paragraph focused on the children and the teacher had not yet been introduced in the text. In the Japanese, given the exclusion of the 'I' pronoun, the problem of agency does not appear. In English, however, it seemed odd rhetorically to first mention the 'I', when no relationship had yet been established in the text. The use of 'I' at this point would be unnecessarily confusing. In this case, it seemed more prudent to keep the focus on the children until the situation was completely introduced and the 'I' could be clearly understood from the context. In this case, rendering the sentence directly (in a grammatical sense) seemed less important to the content and accuracy of the translation than to create a paragraph that flowed well in the TL.

In the translation, care was taken to balance the voice of the TL and the grammar of the SL, both key considerations for Japanese translation. Simply choosing to create a rule for the translation (i.e. to render all implicit subjects with explicit ones or with passive structures) did not seem prudent to effectively solve the problems posed by this translation. By choosing to use more than one construction, SL grammar and TL voice were maintained.

In this translation, the key issue was thinking carefully about the cultural expectations of the author writing in Japanese and how these expectations interacted with the expectations of the imagined reader. The translation required a more careful consideration of the cultural expectations both as an academic paper and as a supplement to the Japanese essay. As this text shows, cultural expectations in translation are not simply negotiating source language culture and target language culture. Especially in this text, the translation was expected to be appropriate and understandable to both the Japanese and Western readers. Several key steps played important roles in making this translation as culturally relevant and accurate as possible.

First, the translation was as collaborative as possible. Because the author had a solid knowledge of English, questions of voice and interpretation were directly placed in the hands of the speaker of the sentence. Of course, whenever this is possible, collaboration between author and translator is extremely helpful in creating a text that is accurate. What might or might not be consequential should be (when possible) negotiated between author and translator. The decision to include the personal pronoun was key to creating a register and genre in English. This solved the problem of grammar constructions and implied subject. The resulting register was deemed acceptable to both the author and the translator, although might have been reconsidered if the translation had been published alone.

Ultimately, as a supplement and following these protocols, this translation is adequate for the purpose it served. Because the expectations for the English translation were considerably lower than they might have been had the paper been published solely as a translation, there was some leeway for taking on a distinct voice that was more narrative than academic. Stricter genre expectations may very well be a blessing in disguise for translators who are seeking out the most appropriate, relevant ways to translate a given structure as these expectations can provide a guide for how a text should or should not be rendered in the SL. For the limited expectation and uses of this paper, the register was appropriate and accurate, and the translation fulfilled the purposes which had been set out.

Baker, M. (1991) In Other Words. London: Routledge.

Coulthard, M. (1992) Linguistic constraints in Translation. Studies in Translation. Ilha do Desterro. 28 (Special Issue): 9-23.

Ishii, H and N. Kamakura. (2006) Actions and behaviors found in clients with Alzheimer's disease while feeding: Investigation into cognitive relevance. The Journal of Japanese Association of Occupational Therapists. 25/6: 497-511.

Kinsui, S. (1997) The influence of translation on the historical development of the Japanese passive construction. Journal of Pragmatics 28: 759-779.

Nakamura, Y. et. al. (2006) Study of the abilities about habituation-dishabituation in persons with severe motor and intellectual disabilities. The Journal of Japanese Association of Occupational Therapists. 25/6: 512-522.

Reiss, K. (1971) Type, Kind and Individuality of Text: Decision Making in Translation. In Venuti, Lawerence (ed.) (2000) The Translation Studies Reader: Second Edition. London: Routledge. pp. 168-179

Sasaki, K. (2002) Poetics of Intransitivity. In Marra, M. Japanese Hermeneutics : Current Debates on Aesthetics and Interpretation. Honolulu: University of Hawaii Press. pps. 17-24.