Although regarded as a branch of translation, subtitling differs greatly in some respects from translation of written texts. The title of a Japanese book 映画字幕は翻訳ではない [Film Subtitling is Not Translation] (Shimizu, 1992), aptly makes the point. Translating for subtitles involves taking spoken dialogue as input and producing written text as output. The discrepancy between the speed of actors' speech and the audience's ability to read written words, plus the restriction on the available space on the screen, all impose a severe limitation on the amount of text allowed for subtitles. This can be likened to the word limitation sometimes imposed on the localizer due to pre-defined string lengths when localizing software. However, while a localizer can, if necessary, make adjustments to multimedia objects, the subtitler is unable to modify the images or audio of a film. The subtitles must be meaningful to the target language audience in relation to the particular scene being shown. All these constraints and requirements make subtitling a specialized task that demands creative linguistic solutions.
Opening a can of worms
The opening of The Fellowship of the Ring, the first episode of The Lord of the Rings (LOTR), in March 2002 in Japan caused outrage among some LOTR fans over the quality of its Japanese subtitles. Numerous Web sites appeared detailing the unfaithfulness and inadequacy of the subtitles, often alleging how little the subtitler had understood the whole nature of Tolkien's mythical epic story. In May 2002, the controversy led some 1300 fans to sign a petition of complaint directed to the film's Japanese distributor, Nippon Herald, against a well-known subtitle translator who had undertaken the assignment. Having received an unsatisfactory initial response with no apologies from the distributor, the same group then sent their complaints directly to Peter Jackson, the director of the film. Subsequently, in a TV interview in December 2002, Jackson reportedly said that he would replace the existing Japanese translator for LOTR's second installment, The Two Towers (Botting, 2003).
Specific LOTR Japanese subtitle errors are discussed in detail on various Web sites. The following often quoted examples from the Fellowship of the Ring are extracted from such sites (see http://herbs.tsukaeru.jp/english_top.html and http://sa.sakura.ne.jp/~straydog/bard/toda.html) in order to indicate the nature of the complaints. Interestingly, it cannot be argued that any of the following instances was caused by the imposed limitation on the number of words. Note: the Japanese subtitles are re-translated into English and compared against the original English lines.
- Boromir's line "And the tower guard shall take up the call 'The Lords of Gondor have returned'" is rendered as "And the tower guard will cry 'The King of Gondor has returned.'" Boromir only calls Aragorn his king at his death, so this line spoils the climax of the film.
- When Boromir comes under the dominion of the ring, Frodo utters: "You are not yourself." This line is translated as "You are a liar!", thus creating the impression that Boromir is a dishonest character, rather than stressing the power of the ring.
- Knowing the ring's power, Frodo asks Aragorn, "Can you protect me from yourself?" This is followed by the ring whispering to Aragorn in order to tempt him to come under its dominion. The above line is translated as "Don't you need this?" making it sound as if the tempter is Frodo, not the ring.
- "Long has my father, the Steward of Gondor, kept the forces of Mordor at bay" is changed to "My father who was the Steward of Gondor had kept the forces of Mordor at bay," thus causing the audience to think that the Steward of Gondor is no longer - yet he appears later in The Two Towers.
In addition, the translation of the title The Lord of the Rings has also been controversial. For the translation of Tolkien's book, the title is rendered in Japanese words as 「指輪物語」 [The Story of Ring(s)] whereby dropping the word "Lord", but conveying the message that the story is weaved around ring(s) [in this case, the Japanese language makes ambiguous the distinction between the singular and the plural]. By comparison, the film title 「ロード・オブ・ザ・リング」 [Lord of the Ring] is a straight transliteration (i.e. the English words written phonetically in Japanese script), but "Rings" is rendered unambiguously in the singular. The contention is that it could wrongly imply that there is only one ring involved in the whole saga. The reason behind the change of the title from the original translation used for the book seems to be largely motivated by commercial rather than literary reasons to attract moviegoers.
Under any circumstances it is a true nightmare for any translator or localizer when his or her work is brought under the microscope and criticized in a very public manner. But, the reality is that commercial feature films are public media and that all subtitlers for commercially released films have to be prepared to face such scrutiny. While most localization work is unlikely to have the name of a single individual as the localizer on the packaging, a film's credits often identify the subtitler. The Japanese fans of LOTR maintain that errors in the subtitles affected the film adversely by not providing viewers with accurate information about the complex story and failing in many places to provide the right context for the audience.
The scale and breadth to which viewers' complaints concerning the subtitles for this film have escalated seem unprecedented, at least in Japan. This is even more surprising, given that the subtitler in question is a widely known expert with many blockbuster films to her credit during the last twenty years. In Japan's conservative film industry where she reigns as the "Subtitle Queen," this has caused something of a shock. A number of factors seem to have led to this rather dramatic development.
- Unlike most typical Hollywood productions, LOTR had very avid fans who were thoroughly familiar with the story and naturally critical of any liberty taken by the subtitler. This audience factor seems to have been underestimated.
- Subtitlers generally seem to able to handle their task without ever reading the original story on which a film may be based. However, in this case, the fans insist that the apparent lack of familiarity with the very particular Tolkien universe was fatal. Their claim is that the understanding of specific types of vocabulary and the unique atmosphere of the whole epic were central to "getting it right."
- Ridiculously short deadlines for the subtitling of big commercial films seem to have become the industry norm (with LOTR, the subtitler allegedly had only one week). Furthermore, the conservative nature of the Japanese film industry, along with a certain degree of monopolization of high-profile subtitling work by a few top subtitlers, seem to ensure that the work is invariably handled by the same individuals.
Under these conditions, something obviously had to give. However, the problem was not without a solution, namely to involve as a general reviewer/checker the translator of the original book on which the film is based. This practice is not uncommon. A recent example is the Japanese subtitling for Harry Potter and the Philosopher's Stone for which the subtitler (interestingly, the same subtitler as the LOTR) worked closely with the Japanese translator of the book. In addition to the quality issue, this process avoided any possible confusion for the many young Harry Potter fans familiar with the translations used in the popular books. Unlike the case of LOTR, the film and the book shared also the same title in Japanese.
Literary translators are generally likely to have spent more time than subtitlers can afford, pondering over appropriate translations for titles, certain names, expressions, etc. Therefore, reuse of the translator's work, where appropriate, has obvious potential to improve subtitling quality. Having said that, however, this solution does not solve all subtitling problems, even when a film is based closely on an existing book that has been translated. This is because the written words intended for books are not always suitable as direct substitutes in subtitles.
A case in point is the handling of Aragorn's nickname, "Strider," in LOTR. The translation of the book had used a rather archaic but poetic Japanese term "Haseo" (馳夫) to suggest a man who "strides" along. However, the subtitler avoided the term for the reason that it would be too difficult for young viewers to understand the meaning and because it would not look good on screen. Although this decision turned out to be unpopular, particularly among those viewers who were familiar with the original translation, it does point to the difference between subtitling for films and translating for books in print.
Subtitling and localization
Some differences between the overall processes involved in subtitling and localization are apparent. For localization tasks, the correct and consistent use of technical terminology in translation is of paramount importance both for quality assurance and for efficient handling of repetitive texts. This aspect is increasingly supported by computer-based tools. For localization, the use of specialized tools is often not an option, but a necessity, to achieve efficiency and quality.
By comparison, while today's subtitles are produced using a computer-based subtitling system, the role of technology in the subtitling process is clearly different from that in localization. For example, it is the human subtitler's task to grasp the essence of the given dialogue and to express it succinctly and appropriately. This dimension of the subtitling process is largely unaided by any tools. Today's subtitling systems are there to facilitate purely mechanical functions, such as cueing the subtitles to the right scene in the program.
On the other hand, we can draw some similarities between subtitling and localization. First, the two tasks both aim to adapt an original product in order to recreate an impact on the target language audience similar to the one experienced by the source language audience.
Second, both are used for widely-distributed commercial products. This normally means a tight production schedule imposed by a client who stands to suffer huge financial consequences if the release date is missed. Although the deadline issue is applicable to all commercial translation work, the stakes seem to be particularly high with commercial mainstream films and popular software products with pre-announced release dates. Furthermore, a wide distribution of the products means that the consequences of translation errors can be far-reaching.
Third, both are affected directly by the introduction of new technologies. For example, localization work has been extended with the advent of the Internet and the Web to cater for new demands such as Web site localization. Similarly, subtitling has been required to cater to new technologies such as DVDs.
In view of the similarities, are there any lessons that the localization sector can offer to subtitlers?
Collaborative screen translation
The localization industry has been striving to both improve the quality of localized products and to work in an increasingly cost-efficient manner. These efforts have resulted in the development of quality control measures, as well as dedicated tools for localizers. Similarly, efforts to establish good practice guidelines for subtitling have been made by groups such as the European Association for Studies in Screen Translation (ESIST). However, anecdotal evidence in both Europe and Asia points to concerns over variable standards that are becoming particularly noticeable with the increasing demand from satellite TV channels, and occasionally from mainstream feature films.
In comparison with localization, subtitling is primarily not considered to be a task based on teamwork. While a subtitler usually works alone for translation, localization typically involves several translators and/or reviewers. With increasing project size and requirements for different kinds of expertise such as a localization engineer, several translators, a project manager, etc, localization is handled by a team of people. Accordingly new tools are geared for such modus operandi, e.g., workflow programs to integrate input from dispersed project members, translation memory systems designed for networked translators, etc. This may be an area where subtitling could emulate the methodology of localization by more frequently adopting a team approach.
Building on the existing solution of drawing on the literary translator's work to formulate the subtitles where the film is closely based on the book, a collaborative subtitling environment could enhance the resulting quality. This would allow the team members to discuss and choose the best solution from suggested translations by the collaborators. This model would be the exact opposite to the current approach to rely on an individual subtitler selected from an elite group, which seems common at least in Japan. For the reason stated earlier, this approach is problematic.
The fact that some of the Japanese viewers who protested against the LOTR subtitles even offered their own translations as alternatives comes close to the phenomenon called "fan-subs". Fan-subs are subtitles produced unofficially by fans for non-Japanese viewers of various Japanese animation (anime) productions. Despite their dodgy legal status, fan-subs have been in existence since the late 1980s due to the increasing popularity of anime outside of Japan. These enthusiastic amateur subtitlers are nevertheless known to sometimes arrive at innovative solutions quite by instinct (Nornes, 1999), pointing to the importance of the subtitler's genre knowledge. The fact that LOTR has a cult-like following with fans who are thoroughly acquainted with the Middle Earth lingo and its nuances makes the audience factor very similar to the anime following. The fan-sub phenomenon perhaps provides food for thought for film subtitling, particularly in relation to the problems raised by LOTR.
Given the likelihood of further growth in demand for subtitling in terms of volume and of variety by the entertainment industry, the centralized model reliant on one high-profile subtitler seems increasingly out of step with market requirements. It is simply impossible for a single individual to be expert at all film genres and to produce high quality work in huge volume and under tight deadlines. With the maturing of teletranslation and virtual work environments, remote collaborative subtitling by a virtual team, which may include a devoted fan or expert in the genre, could well offer the best outcome for certain films. Following the fan-sub development, it is not entirely unthinkable that frustrated moviegoers, disappointed by official subtitles, are driven to produce their own, aided by digital technology which provides subtitling facilities on standard PCs relatively easily.
Watch the space
In anticipation of the second installment of LOTR, The Two Towers, about to open in Japan, New Line Cinema, the film's global distributor, released a special statement regarding its handling of subtitles in reference to the earlier "comments" (the distributor has never acknowledged them as complaints) made by some Japanese fans on The Fellowship of the Ring. It explains the complete auditing process for The Two Towers that extensively involved the translator and the publisher of the Japanese version of the LOTR book as well as New Line Cinema stepping in to approve the Japanese subtitles which were especially translated back into English for that purpose. In any case, the fans will, no doubt, report back on whether the new quality control measures prove satisfactory.
The availability of platforms such as the Internet has clearly contributed to the ability of film viewers to bring their critical comments to the public arena and to stimulate a public debate using specific examples for contentious points. This type of reaction could equally arise in the localization industry where a user group, unhappy with a certain localized product, could launch such a campaign. Electronic word of mouth over the Internet travels a vast distance extremely rapidly and can have considerable impact. The LOTR subtitle campaign seems to have succeeded to some extent in raising public awareness of what is involved in subtitling and how it can go wrong, even in the hands of a professional. Using the same platforms, the film industry could move into a new collaborative mode of subtitling whereby ad hoc teams could be set up on the network appropriate to the type of film. With the rapidly diversifying market for subtitling propelled by digital technology, the drive to achieve quality within time and economic constraints is growing in importance.
One hopes that the LOTR experience will prompt the film industry, along with its subtitle suppliers, to reflect critically on the current process and methodology. The entertainment industry looks set to grow further in the global marketplace, and language support will be a critical factor in this growth. Now seems to be a good time to think creatively about ways to augment the unique human talents of subtitlers.
Acknowledgement: I would like to thank Dr Eithne O'Connell, subtitle expert, of SALIS, Dublin City University for her valuable comments on my earlier draft.
By Dr. Minako O'Hagan
School of Applied Language and Intercultural Studies (SALIS),
Dublin City University
Botting, G. (2003, Jan. 5). "Rings fans put 'subtitles queen' to sword." Sunday Mainichi.
Ivarsson, J. (n.d). "A short technical history of subtitles in Europe."
Nornes, A.M. (1999). For an Abusive Subtitling. Film Quarterly, 52,3:17-34.
Shimizu, S. (1992). 映画字幕は翻訳ではない [Film Subtitling is not Translation], Tokyo: Hayakawa Shobo.