Corps de l’article

Introduction

For a long time, translation has played an important role in constructing a culture within a given country. It sometimes influences the national literature, even becoming the basis for the introduction of new styles of writing. In contemporary culture, however, translation seems to be increasingly dominated and controlled by a commercial power, namely the publishing industry.

Itamar Even-Zohar’s polysystem theory defines a culture within a country as a hierarchical, multi‑level structure. According to this theory, the highest position is occupied by either innovative or conservative literature (Even‑Zohar, 2000, p. 193). Though Even‑Zohar does not specify any active players in the polysystem, clearly publishers, editors, copywriters and literary agents play major roles in the construction of culture. In particular, literary agents undoubtedly play a key role in the book production process (Munday, 2001, p. 153). In Anglo‑American culture, there is an increasing tendency for editors to rely on literary agents and events such as the annual Frankfurt Book Fair, where literary agents represent the national literature to be exported (Schulte, 1990, p. 2). Although some Translation Studies scholars have recognized the importance of literary agents, few studies have reported on their role in the book industry of a specific country.

This paper is an attempt to describe the position of translation in the contemporary Japanese book industry by focusing on the intervention and the role of Japanese literary agents. It also aims to reveal current publishing strategies regulating the techniques employed when translating books of a particular category (other than literature). This study begins by considering the polysystem theory, which provides a more precise explanation of the internal organization of a culture and its relationship to translation. Thus, it looks at translation as a “product” rather than focusing on its “process” or its “function.” The research follows the first two phases of Gideon Toury’s methodology: (a) situate the text within the target culture system, looking at its significance or acceptability, and (b) compare source text (ST) and target text (TT) for shifts in order to identify relationships between “coupled pairs” of ST and TT segments, and suggest generalizations about the underlying concept of translation (Toury, 1995, pp. 36‑39, 102). To carry out the first phase, statistics from UNESCO will be presented describing contemporary Japanese culture as a polysystem. (These statistics represent cumulative bibliographical information on books translated and published in UNESCO member states, including Japan.) For the second phase, the research will focus on the actual production of a translation. At this stage, the study will present the results of an interview with the CEO of TranNet, a literary agency, and the data[1] obtained from the agency, which describe the activities behind the production of a translation in Japan, revealing the publisher’s influence on the translation process.

It is hoped that this article will provide useful insights into the production of book translations in contemporary Japan and the important role played by literary agents. This fresh look at the position of translation within a contemporary culture is sure to stimulate further research on the translation process and the function of each translation within a context of cultural creation.

1. Book Production in Japan and the Polysystem Theory

1.1 Book Production and Translation in Japan

Books are major components of culture. A country’s book production, despite its cultural significance, is generally controlled by a commercial power (Hale, 1998, p. 190). Publishers exist to produce books, translated or not, and to make a profit, not to construct a culture. As such, the publishing industry can be seen as a commercial power. In the actual publishing process, the central players are editors, whose projects are usually subject to financial constraints. Editors, therefore, seek books that will provide the publisher with both prestige and profit.

When it comes to the production of translated books, editors prefer authors with established reputations in their own countries, or authors whose books have already been translated into English (Hale, 1998, p. 190). Publishers, not translators, initiate most translations, despite the fact that translators, with their profound knowledge of their target language, are ideally suited to function as “the key figure in the establishment of cross‑cultural communication” (Schulte, 1990, p. 2). Publishers, however, have other preoccupations altogether. They tend to offer contracts for a modest flat rate, seeking to minimize translation costs (Venuti, 1998, pp. 31‑66). As a result, the final product (the translated book) is typically very much shaped by editors and copy‑editors, who take into account the financial constraints of publishers, along with market trends in the publishing industry, conceived by the complex network which Fawcett calls “Power Play” (Fawcett, 1995, p. 189).

Book production in a culture is typically understood from the standpoint of these actors: editors, publishers and literary agents. Polysystem theory provides a different perspective. It helps us to understand culture as a system and to examine book production and the position of translation within a given cultural system.

1.2 Polysystem Theory

As previously stated, Even-Zohar’s polysystem theory holds that a culture has multiple strata, and the highest position in the literary polysystem is occupied by either innovative or conservative literature. The relationship between these two is in a constant state of flux since they are in competition for readers’ attention. Since culture in a polysystem possessing various levels of literature is constantly changing, the position of translated literature is also an unstable, evolving one (Even‑Zohar, 2000, p. 196).

However, Even-Zohar found that translated literature is usually displaced from its central position and tends to occupy a secondary position in the polysystem (Even‑Zohar, 2000, p. 196). The peripheral position of translation can be revealed by the percentage of book translations published. In the United States, translated books constitute a small percentage of overall book production. Moreover, American publishers tend to be more interested in selling their own literature and exporting it to other countries than importing foreign literature, a phenomenon dubbed “trade imbalance” by Venuti (1995, p. 14). Although we must bear in mind that the situation in each country varies over time, it seems safe to say that translated literature rarely occupies the primary position in the polysystem.

Even-Zohar shows that, when translation occupies a peripheral position, it has little influence on the central system and can even become a conservative element,[2] preserving conventional forms and conforming to the literary norms of the target culture system. Translators, then, tend to use existing target culture models to render the TT.[3] Thus, the peripheral position of translated literature in the polysystem may condition and shape publishing strategies (Even‑Zohar, 1978/2000, p. 196‑197).

Although the polysystem, as conceived, is in constant flux, the position of translated literature usually falls into a secondary position, where the tendency is to conform to target cultural norms. It would therefore be effective to apply Toury’s methodology, originally introduced in an attempt to make generalizations about underlying translation norms (Toury, 1995, pp. 36‑39, 102), to describe the strategies at work in the translating and publishing stages under the target language norms.

However, recent studies on publishing strategies (Alvstad, 2003) seldom focus on the role of the literary agent despite the fact that its significance has already been recognized (Munday, 2001; Schulte, 1990). It is, therefore, interesting to examine the activities of a literary agent who acts as a mediator between translator and publisher in a country where translation occupies a peripheral position. It is here that a discussion of the Japanese literary agent may well provide interesting insights into contemporary publishing strategies, as well as into related translation techniques. This paper will present the results of an interview with a literary agent from TranNet, a literary translation agency, about its electronic translation audition systems, as in the first phase of Toury’s methodology (situating the ST and TT). We will then go on to discuss some model translations provided by the agent, as in Toury’s second phase (comparing the ST and TT), to see whether publishing strategies are shaped by the intervention of a literary agent.

2. Situating and Comparing the Text within the Target Culture System

2.1 Translation in Contemporary Japan

Before focusing on the first phase of Toury’s methodology, i.e., situating the ST and the TT within the target cultural system, it is useful to know the position and the percentage of translated books within different cultures. Appendix 1 shows the number of new titles along with the number of new translations published in Japan from 1991 to 2001. Newly published titles (in grey, leftmost columns) show a slight increase, while new translations (in black, centre columns) are little changed. It is clear that the total number of translations is far less than that of new titles, representing no more than 8% of overall book production over a 10 year period. Moreover, a comparison with major European countries (Portugal 44%, Italy 26%, Spain 24%, France 18%, Germany 14%, Britain 3%) (Hale, 1998, p. 190) reveals that the percentage of translated books in Japan is at the low end. This figure indicates that the position of Japanese translation is peripheral and that the literary polysystem in Japan is dominated by national, untranslated book production. From this, we can predict that Japanese publishing strategies are shaped by strong target‑culture norms, preserving conventional forms of Japanese literature.

The statistics also include the number of translations from English‑language originals (in white), showing a similar tendency to newly‑published translations overall, with a slight increase over the last 10 years. It is also important to recognize the ratio of translations from English among all translations. Indeed, English accounted for over 75% of translations in Japan between 1991 and 2001. Citing a European average of 60% (Hale, 1998, p. 190), this figure shows that there is a significant preference in Japan for books from English‑speaking cultures: translations in Japan seem rather homogeneous in terms of source language, despite translation’s potential for introducing unfamiliar cultures. When it comes to techniques of translating from English, the variety of models and styles available of translations from English would seem to make it easier for translators to produce texts that conform to target cultural norms.

Similarly, an examination of the figures of books translated from English by category (Appendix 2), shows that, between 1991 and 2001, Literature was the most translated, followed by Science, especially Applied and Social Sciences. Arts/Games/Sports was the fourth‑largest category of published translations from English, particularly practical guides and how‑to books. This category has been growing since 1991, showing that translations of Arts/Games/Sports texts are more and more accepted and popular in Japan. This reveals a certain trend in translations from English: after Literature and Science, the number of practical guides in Arts/Games/Sport subjects has been increasing over the last decade.

In summary, an examination of these statistics provides a clear understanding of the position of translation in Japan, which, from 1991 to 2001, represented no more than 8% of national book production, placing it in an extremely peripheral position in the Japanese polysystem. More than 75% of translations have English as a source language, making it the dominant source‑language during those 10 years. There is also a clear preference for Literature, followed by Applied Sciences, Law/ Social Sciences/Education, and Arts/Games/Sport.

2.2 Comparing ST and TT Paratextual/Textual Features

2.2.1 TranNet, A Newly Established Literary Agency in Japan

As key players in book production (Munday, 2001, p. 153), literary agents are a useful focus as a research object. The results below are drawn from an interview with TranNet CEO Masahiro Takano in December 2005.

TranNet was established in 2001 and has become a major Japanese literary agency. Its business includes electronic calls for translators, systems for publishing translations and distance‑learning courses in literary translation (English and Japanese). There are almost 3000 TranNet members (as of December 2005) who pay an annual membership fee, making them eligible to participate in electronic translation auditions. Some are already established literary translators who wish to expand their working field. Others are translators of business documents, or aspiring translators still learning the techniques of literary translation who dream of becoming freelance literary translators. TranNet has been selected here as a case study for two reasons: (a) it is officially supported by many leading Japanese publishers such as Kodansha, Shogakukan and Nikkei BP, and it has also created a new electronic publishers’ network by issuing periodical book reviews to 700 Japanese editors from 200 Japanese publishing houses,[4] which should help uncover certain trends among publishers; and (b) since December 2005, TranNet has posted 269 electronic calls for translators, and its electronic audition system, an original and unique project in the publishing industry, has been instrumental in the publication of 192 translations. As it is the only literary agency to hold calls for translators, this should help uncover some of the norms at play in the contemporary Japanese polysystem.

3.2.2 Data on Translations and Paratextual Features

An electronic call for translators proceeds as follows. A publisher proposes a foreign‑language book to TranNet for which they would like to find a translator. The publisher’s book choice is based either on the publisher’s own interest or on TranNet’s book reviews. The publisher and TranNet work closely together to create a “call‑for‑translators program,” including such factors as deadline, number of pages and financial remuneration. TranNet extracts the part of the book to be translated (usually around 1,000 words), creates a summary of the book and provides important notes for the translation, based on the publisher’s objectives (target readers, desired writing style, etc.).

A call for translators containing all this information is then sent out by email to all 3000 TranNet members. Members are even able to peruse a few pages of the book in PDF format to see the text in its original context (in some cases, with illustrations and font styles). TranNet puts out a call for translators almost every two weeks and also provides feedback—out of 10 possible grades—for each participant with a model translation of the extract from the book to be translated. A model translation is a version that is considered to be ideal by the publisher and TranNet.

In the call-for-translators system, it is the publisher who initiates the translation project and maintains a leadership role with TranNet, whose role seems to be to carry out the publisher’s wishes, passing them on to the eventual translator. Although the literary agent offers book reviews to publishers, its role in the choice of books to translate appears rather limited. This indicates that prospective translators are supported by the powerful network of the literary agent, who acts according to the publishers’ intentions.

3.2.3 Paratextual Features Discovered in the Call-for-Translators System

The interview with Masahiro Takano and data from all 269 calls for translators presented by the literary agent provide insight into some interesting paratextual features which constitute factors behind publishing strategies. The data, in addition to the original text and model translation, consists of the same information in the “call-for-translators program”—such as a deadline, the title of the original text, the source language, the provisional Japanese title (if any), the subject, a book summary, the volume of translation, the translation period, (whether preliminary or final), specifications about royalties, and notes for the translator. Each of these eight paratextual features is described in greater detail below.

Source Language

Nearly all of the calls for translators were translations from English. Specifically, according to the Masahiro Takano, 90% of auditions were for translations from American English. TranNet does arrange translations from languages other than English (though a few times a year at best) outside of the normal call‑for‑translators system, and invites translators of other foreign languages (such as Chinese, Spanish, Italian, German, Korean and French) to do a special test translation. The dominance of English can be seen in results from Index Translationum[5] and shows that there is more demand for books written in English than in other languages. As there is also an abundance of Japanese translators with an excellent knowledge of English, the electronic call‑for‑translators system is in place instead of asking translators individually to do a test translation. It also indicates that publishers prefer books from American English, implying that these books share certain cultural features with Japan, and so are more easily accepted by Japanese readers.

Categories

TranNet began to categorize calls for translators after the 70th call. Business/Economy is the largest category (70), followed by Subculture/Lifestyle/Beauty (45), and Narratives/Arts/Films/Music (26). Other categories include Hobbies/Sports/Games (22), Humanities/Politics/Law (16), Medical Sciences (10) and Natural Sciences/Applied Sciences (6). Although these categories do not correspond to those in the Index Translationum, it is interesting to note that Literature, usually the most often translated category, has not been included in the call‑for‑translators system. This is probably because books categorized as Literature (especially bestsellers such as The Da Vinci Code or those written by well‑known authors such as Stephen King) already have an experienced translator or academic to translate them. Most frequently translated categories in the call‑for‑translators system roughly follow the order presented in the Index Translationum statistics, where both Business and Economy are placed under the Social Sciences, and Subculture/Lifestyle/Beauty is in Arts/Games/Sports. It is safe to say that the electronic audition system fairly reflects the actual translation categories in Japan. Considering that it is the publisher who initiates translation projects, calls for translators seem to conform to target‑culture norms for books on Social Sciences and practical guides (Arts/Games/Sports).

Titles in Japanese

A provisional title of the translation is not usually presented in the call‑for‑translators announcement. According to Masahiro Takano, it is usually the publisher who decides on the title of the translation during an editor’s meeting at the final stage of publication (after the final proofreading). The editor usually uses a literal translation of the original title and rarely listens to the translator’s suggestions, although the translator often knows the context better than the editor. In some cases, the publisher changes the title completely, based on advice from the sales department, which is relatively familiar with book market trends and sees the translation as a commercial product to be sold in great numbers. When determining a title for the translation, even editors must sometimes bow to the sales department, which reveals the complex “Power Play” of the publishing industry (Fawcett, 1995, p. 189).

Preliminary and Final Translations

Eighteen percent of the electronic calls for translators are designed to find a translator to submit a preliminary translation (Shitayaku in Japanese), which is then checked by an editorial supervisor selected by the publisher. The publisher then prints and sells the final translation under the supervisor’s name. An editorial supervisor is chosen either because s/he is an expert in that field, or because by including the name of a well‑known writer as supervisor of the translation, the book sells better than if it bore the name of an unknown translator. Although calls for translators are usually organized to find a translator for the final translation (Uwayaku in Japanese) and the percentage of preliminary translations represents no more than one‑fifth of the entire number of calls for translators, this statistic helps us understand the part commerce plays in the production of translations.

Royalties and Contracts (Kaikiri)

Each translator is given a contract for the final translation. The translator does not receive royalties or copyright for the book. This is rather similar to the situation of American literary translators, who tend to work from contract to contract, usually for a modest flat rate (Venuti, 1998, p. 31‑66), and to whom publishers are often reluctant to grant copyright or to pay royalties, in an effort to minimize translation costs (Venuti, 1995, p. 189).

Call-for-Translators Judges and Model Translations

Each call for translators is supervised by a few judges selected by TranNet from among professional literary translators in the field or academics specialized in English. TranNet, working with the judges, also produces a “model translation” for the call for translators and collaborates with the editor, but it is the editor who finally proofreads and corrects the preliminary translation, which then becomes a “model translation.” It seems obvious that editors also intervene at the textual level, giving advice on style and tone to make the text more acceptable to target readers.

Criteria for Selecting a Book

Based on the data provided by Masahiro Takano, publishers tend to select books calling for translators according to the following criteria: either the original book has attracted wide attention, including both readers and critics, in the foreign country, or the book addresses a trend of interest to Japanese readers or deals with current social issues in Japan. Publishers attach more importance to the first criterion. A book is rarely selected because it might stimulate or educate Japanese readers. Market forces in both the source and target language cultures are one of the most vital criteria, as suggested by Munday (2001, p. 153); translation production is predominantly dependent on a larger social trend, with little potential to inspire readers by importing new ideas from foreign countries. As for literary agents, although they provide publishers with in‑house book reviews, they exercise little influence over publishers’ decisions despite the agent’s role as a cultural mediator with profound knowledge of foreign book markets.

Criteria for Evaluating a Good Translation for the Call for Translators

TranNet establishes its own criteria for each call for translators. The audition is defined not as an exam to evaluate translation techniques, but as a selection tool to find a translator who can best meet the publisher’s terms for the book in question. As shown above, the call for translators itself is largely under the control of the publisher, whose main objective is to sell as many copies as possible and burnish its reputation (Hale, 1998, p. 190).

When the call for translators is sent out, each participant is given, along with the source text, a copy of the publisher’s notes about the translation, which include the most important instructions. A participant will lose many points during the audition if s/he chooses not to follow these instructions and produces a translation in his/her own particular style. In the call‑for‑translators system, this is seen as a serious mistake, and is as ill‑regarded as failing to translate an entire paragraph or failing to render an appropriate translation for a keyword. Publishers sometimes attach more importance to the tone and rhythm of the entire text than to the correctness of the translation.

Moreover, TranNet defines readability as one of the crucial points to be considered, which is understandable since readability is stressed in many translation guides (Kono, 1975; Ando, 1995). In the call‑for‑translators system translations of general‑interest material should be as easy to read as any Japanese book on the market and should maintain a certain level of attractiveness in terms of textual rhythm, nuance, tone, etc. The definition of readability is usually broad, and there is generally a consensus that it is a reader‑oriented concept (Robinson, 1997). According to the audition system, a translation, being a commercial product, should be an enjoyable read rather than an obvious translation of a book from a foreign language. This implies that the translation should be as fluent as if it had been originally written in the target language (Venuti, 1998).

To summarize, the role of the literary agent appears to be to support the publisher’s translation projects as actively as possible. Results from source language and subjects provided by the UNESCO data show that the literary agent usually follows the general trend of the publishing industry and does not influence publishers. The literary agent in the call‑for‑translators system works closely with publishers and never challenges their primary objectives.

Furthermore, the examination of several paratextual features has shown that literary translators have little financial power, and sometimes submit a preliminary translation, which has to be checked, and is often sold under the name of an editorial supervisor. The phenomena apparent in the electronic call‑for‑translators system appear to reinforce the publisher’s power in the production of translations. Publishers hold almost complete control over their translation projects, from book selection, to the call for translators, to finalizing the translation with appropriate corrections, turning the translation into a commercial product that can be successfully sold.

2.2.4 Textual Features of the Audition System[6]

The previous section described the paratextual features of the call‑for‑translators system, which clearly indicate that there are various external constraints behind the audition that become visible through the publisher’s intervention. A closer look at the translations will be helpful at this stage to describe the publisher’s interaction with the text. In the second part of this paper, we will apply Toury’s methodology (comparison of ST and TT as a model translation) to the call‑for‑translators system and discuss it further.

The research begins with a selection of calls for translators for books from the Subculture/Lifestyle/Beauty category from 2001 to 2005. This category was selected for the following reasons: firstly, Subculture/Lifestyle/Beauty has never held a primary position either in the Index Translationum or in the TranNet calls for translators, it rather holds a secondary position, implying that translations from this category are more conservative than others, as discussed by Even‑Zohar (2000 [1978], p. 196), and consequently exhibit a higher level of usage of conventional forms that conform to the literary norms of the target language culture. Secondly, it is assumed that the target readers of Subculture/Lifestyle/Beauty books are from a more varied audience than that for books in the Business/Economy category, which might lead to some interesting points of discussion regarding the role of target readers in the process of producing translated books. The ST and TT (the model translation) were then compiled into an electronic format with translation notes for the call‑for‑translators announcement.

A closer look at the 45 book translations in TranNet’s Subculture/Lifestyle/Beauty category shows that there are 5 global translation trends that publishers and the literary agency opt for: (1) character usage (hiragana, katakana and kanji), (2) addition and supplementation, (3) translating ST keywords, (4) consideration of the ST author, and (5) consideration of the TT reader. The examples below are taken from recent calls for translators.

Character Usage

As written Japanese uses three types of writing systems, the model translation also makes comments on the usage of all three. According to the model translation and the notes to translators, there are three methods for denoting an animal or a material to choose from, but they must be chosen with care. The model translation and the notes to translators establish that it is always important to consider which is the easiest way to read a text. Reading the translation aloud is a recommended method for seeing which text reads most easily, and it is also helpful to consult the list of some two thousand kanji characters designated for everyday use.

This example shows three possibilities for certain words: “frog” can be written using three different Japanese systems かえる／カエル／蛙, “heron” さぎ／サギ／鷺 and “excrement” ふん／フン／糞. In this case, the model translation chose katakana for all three to increase readability. This usage of katakana, which is certainly different from the one of beautification, gives the words a more modern, educated sound, as suggested by Baker (1992, p. 36). The example shows that readability is also produced by using different characters in different contexts, according to the textural complexity. It is assumed that Chinese characters were avoided because they look heavy and complicated in this context: the author is not describing research results but rather speaking informally from his imagination.

Addition or Supplementation

This technique involves adding a word or a series of words in the target language to clarify the meaning of the source language.

The example above demonstrates that, in order for the Japanese translation to have a logical meaning, the expression “putting cancer” is rendered metaphorically with “putting powder of cancer.” In the model translation, versions such as “mixing a carcinogenic substance” or “burying cancer” were rejected, since the meaning and usage in Japanese is wrong. This example indicates that the translation should correctly reflect the exact meaning of the ST expression, even if this involves resorting to metaphorical addition or supplementation.

In this example, the ST “this” is rendered by “our mind’s” based on the context. This is one of the explicitation techniques recognized as a constituent of “the universals of translation,” and is also considered a standard technique, especially in the Japanese language (Ando, 1995, p. 72). The model translation and the notes to translators say translating “this” literally is not wrong given the context but is rather “unkind to the target‑reader.” The word “mind” is added to make the meaning easier to understand. This implies that the publisher’s understanding of the TT reader influences the translator’s semantic choices. There are several similar examples involving subject or demonstrative pronouns, and supplementation is recommended so that the Japanese text can be easily understood without referring to the original. In short, readability is an important factor when producing a translation.

Translating ST Keywords

Translating ST repetitive keywords is a particularly important task. Since the keywords determine the entire style or tone of a text, the translator should continue to use the same words throughout the text. For example, in The Drive of Your Life: How to Navigate Your Way through Life’s Journey Using Simple Self‑Help Strategies (b), “mind” and “spirit” are keywords; these two concepts establish the structure of the text; the book, in fact, has two parts divided by these two keywords. In the model translation, TranNet admitted that, as many participants suggested, there were several possibilities, and that without a thorough reading of the entire text or a good understanding of the author’s aim, it would have been rather difficult to pick out these two words from a series of others (“sense” and “soul,” “brain” and “heart,” etc.). After discussing several possibilities, it was suggested that other books in the same category be consulted to find appropriate words, i.e., look for examples (on the Internet and in bookshops) from books on the same subject, in this case, “Self‑Help” books, that would suit this similar context. They opted for “heart” and “soul” as Japanese keywords (心と魂) because they were used with the most frequency, the conclusion being that translators need to examine similar texts to identify current trends and usage. Translators should choose those keywords most often used and “distributed” in a given category because these words have already proved to be suitable in the publishing industry. This technique shows that producing an expression or a keyword in the TT is sometimes dependent upon other texts in the same market category (both in other translated books and first‑language Japanese books). It also shows that the translation of keywords in a book in the Subculture/Lifestyle/Beauty category tends to conform to the norms of the existing target language culture.

Consideration of ST Author

The three features discussed above can be considered textual manipulations: adding or supplementing Japanese words, limiting the use of kanji and determining keywords according to the norm reflected in other books on the same subject. Consideration of the ST author introduces another level of manipulation. It does not involve a specific translation technique, but seems to suggest how a book will be accepted by the target‑readership. Considering the ST author at the translating stage goes far beyond the textual level, even involving the target reader’s expectations of the book.

For example, The Voice of Tobacco: A Dedicated Smoker’s Diary of Not Smoking (a) is presented in the model translation and the notes to translators as being written by a middle‑aged Englishman with a 40‑year smoking habit. After watching a BBC documentary on a parasite, he suddenly realizes that smoking is like swallowing a tobacco parasite and decides to quit smoking. The text is written in a colloquial tone in the form of a diary expressing his emotional conflicts. According to the notes to translators, attention should be paid to this information about the author (especially “English,” “middle‑aged” and “male”) and the translation should be written in colloquial Japanese, not necessarily slang or the vocabulary of young boys, but in the voice of a middle‑aged Japanese man expressing himself in normal conversation. Accordingly, the translator must imagine the author of the ST when reproducing the story in Japanese.

Similarly, considering the career of the author also appears to be important. A guide for children entitled How to Help Children Find the Champion Within Themselves (c) was written by British athlete David Hemery, Gold medal winner in the 400 m hurdles at the 1968 Olympic Games in Mexico City. This information is presented both in the call for translators and in the model translation along with the notes to translators, implying that it will be important when choosing the style (formal/informal, tone, rhythm, etc.) of the text. The model translation along with the notes to translators also mentions such elements as the year of his birth (1944) and his appearance (that of a gentleman and sportsman). The resulting translation should reflect this image of David Hemery, leading a Japanese reader to expect an honorific style of speaking. It shows that when translating a book by an author who is relatively well‑known or has achieved something special, as with David Hemery, the translator should immerse him/herself in the author’s life, and this should be reflected in the Japanese TT.

It is sometimes also necessary not simply to pretend to be the author, but to think of the real situation in which the book takes place. A guide on child‑rearing, The No‑Cry Sleep Solution: Gentle Ways to Help Your Baby Sleep through the Night (d), was written by a woman who has raised 4 children on her own. The book is based on real‑life experiences and thorough research. The model translation and the notes to translators state that the translation should sound as though the author were addressing future mothers and fathers who want to learn from her personal experience, not in a university lecture room but in a baby‑care classroom, with the participants seated in a circle. This sympathetic imagination suggested by TranNet influences the stylistic choices of the target text, e.g., colloquial language, affectionate words (preferred by women) and avoidance of heavy kanji characters.

These three examples demonstrate that consideration of the ST author’s gender, age, nationality and career is important in the translation of subculture subjects (practical guides and how‑to books). The social context of the ST author is not always explained in the ST but has become essential information for the translator.

Consideration of TT Reader

Another extra‑textual factor that goes beyond the ST is consideration of the TT reader. This reader is sometimes described in the call‑for‑translators announcement so that a participating translator may more easily determine the tone of the translation. The following examples show the target‑reader information for each book presented at the release of the call for translators.

The Drive of Your Life: How to Navigate Your Way through Life’s Journey Using Simple Self‑Help Strategies (b)

This self‑help book is designed for professional, well‑educated young women who are sometimes upset because of the conflict between what they have accomplished and what they want to do in the future. The tone and style of the translation should be helpful and supportive, that of someone giving a helping hand and making the reader feel relaxed.

The No‑Cry Sleep Solution: Gentle Ways to Help Your Baby Sleep through the Night (d)

This book is designed for first‑time mothers and fathers who go to baby‑care classes. It is recommended that some technical words concerning pregnancy be replaced by borrowed words in katakana instead of kanji, since the book is targeted at young couples who much prefer reading a text that sounds light, and who would also appreciate the novel sound of borrowed words since they already have a basic understanding of English. It is also recommended that other child‑rearing books or magazines be examined for general trends found in this kind of book.

100 Things Guys Need to Know (e)

This book is targeted at young boys in grades 5 and 6 or in junior‑high school. The translations of the comics section should match each character in the illustration. The translation should read easily and present a certain level of friendliness on the part of the author.

It is interesting to see that the target reader is indicated along with the book’s specific role. In the first example, the translation should sound affectionate, especially as the book is targeted at young women. In the second example, the recommendation is that the translation should use more borrowed words than kanji characters for the young couples who, the publisher assumes, are used to this type of borrowed English vocabulary. The third example indicates that the attitude of the ST author towards the ST reader should be preserved since the TT reader will share the same profile as the ST reader. Therefore, the tone of the translation should reflect the author’s warm‑heartedness. These examples show how TT readers are clearly defined and that the translator should take into account the gender, age and sometimes the career of the target reader and then produce a translation to suit the target‑reader’s expectations. These examples clearly show how the translations are situated in the complex network of the publishing industry to take into account the publisher’s idea of the target‑reader group.[7]

The examination of these five textual features shows that publishers expect translators to produce translations that read well and sound as if they were not translations at all. This is evident from the model translation, which specifies that particular attention should be paid, when choosing the overall style of the translation, to the social context of both ST author and TT reader. This is partly because categories such as Subculture/Lifestyle/Beauty have never been main translation categories and, therefore, according to Even‑Zohar, they naturally lead to the preservation of existing conventional target‑language forms. Readability, particularly in contemporary Japan, is a target language culture’s main concern (Furuno, 2002, p. 121) regarding these types of books, and the results shown above support this contention.

As model translations are usually created with the cooperation and participation of the publisher, it is clear that publishers tend to support translations that contain no trace of the ST and that can be understood without readers even being aware that they are reading a translation. Publishers are even prepared to show translators their ideal conception of the future Japanese version through the model translation along with the notes to translators, which obviously shapes the style and tone of the translation. The “translator’s invisibility” is already recognized in the field of literature (belles‑lettres) (Venuti, 1995, pp. 2‑5), but a secondary‑positioned category such as Subculture/Lifestyle/Beauty also seems to show a strong trend toward “concealing the translator behind the book,” the goal being to make the book profitable.

Conclusion

This paper has demonstrated that, in the call‑for‑translators system established by literary agents between publishers and translators, publishers exercise more power than literary agents (and, of course, more than translators) in the production of translated books. Situated on the periphery, the process of producing translated books is usually shaped by the publisher’s requirements. In particular, the translation of books in the Subculture/Lifestyle/Beauty category, which has never occupied a primary position, is largely dominated by a powerful network of publishers strongly supported by literary agents. In the call‑for‑translators system, literary agents have little influence over publishers, who control the production of translated books.

Additionally, this multifaceted network demonstrates how publishers control the selection of books to be translated, even indicating the desired translation style. Imposing an editorial supervisor (Shitayaku) and a lack of royalties (Kaikiri) are clear signs that publishers place their emphasis on profits and prestige. This is evident even at the textual level, with publishers sometimes deciding the style and tone of a translation to give the text a particular rhythm in order to increase acceptability among target‑readers, i.e., the “consumers” of this commercial product. Readability is stressed and achieved by adapting the social context of the ST author to the TT target reader. Elements such as gender, age and career of both the ST author and TT reader are considered in order to increase readability and determine a style appropriate to the TT reader.

This paper has looked at the role the literary agent plays in contemporary Japanese translation publishing strategies, an area of Translation Studies in which little research has so far been carried out. This study focused more on the call‑for‑translators system than on the activities of literary agents. It would be interesting to study the interaction between publishers and literary agents, and the behaviour of both to see how this affects the production of translated books. Also, it would be useful to carry out similar studies focusing on a different language (for example, translations from French), and on a different category (for example, Social Sciences), or on books published during a given period, to see whether different conditions demonstrate a similar tendency in publishing strategies. Finally, this research was carried out through a manual analysis of selected data; it might be useful to conduct computer‑assisted analyses using electronic corpora to investigate translation shifts by looking at the occurrence of the translation of a particular word in books from a single category. Revealing these shifts, which are considered a common phenomenon in translation, may help bring to light tendencies in techniques employed by translators when translating books from a particular category and so add to our understanding of how publishing strategies work in actual book production.