American-Owned Japanese Patent Translation & Japanese Deposition Interpreting Boutique Decades of Experience In-depth Subject Matter Expertise Based in Japan, Serving the World

Resources FAQ About Us Site Map

Japanese Patent Translation Deposition Interpreting Inquiries Clients Japanese

HomeResourcesJapanese Patent Translation BulletinNo. 14: The Difficulty of Finding Translators with the Three Core Competencies Required of a Japanese Patent Translator Japanese Patent Translation Bulletin No. 14 (May 6, 2010) The Difficulty of Securing Translators with the Three Core Competencies Required of a Japanese Patent Translator: Cutting Through the Hype and the Misconceptions Executive Summary:  Acquiring the three core competencies required of a Japanese patent translator is not a trivial problem, and is made more difficult by recent trends in the translation industry. In our last issue, we looked at the three core competencies required of Japanese patent translators. This issue, we will look in more detail at just what is required, and take a look at some industry trends that are making it harder, if anything, to be assured of securing the services of translators having the required skills. Source Text Comprehension Since the one of the most important assumed readers of a Japanese patent specification in its original Japanese-language form is a Japan Patent Office examiner, the translator ideally should have the same level of reading comprehension of such documents as would be expected of a JPO examiner. This is not an easy condition to meet, particularly in view of the demographics of Japanese-to-English translators, regardless of whether their native language is Japanese or English. The Humanities Barrier.  Long experience shows that a good portion of people translating patents from Japanese to English, be they translators having English or having Japanese as their native or "dominant language," have a humanities background and would seldom, if ever, encounter anything similar to patent style texts in their everyday life. In the case of NES (native English speaker) translators, this background is often a background in the study of the Japanese language or culture, most often at a university outside Japan, with no relevant technical education or experience. NJS (native Japanese speaker) translators aiming at translating patent from Japanese-to-English often have a background in English literature or a similar subject acquired in Japan. Without a considerable amount of study, it is unlikely that such translators will be able to reach the level of comprehension necessary to comprehend Japanese patent documents at the required level. The Barrier of Distance from Japan.  For many NES translators, Japan was a one-time home, and provided an opportunity to live and use Japanese in Japan. Once they return to their home countries, however, it is easy for such translators to lose their Japanese ability, for several reasons. One, obviously, is that everyday life outside of Japan is not usually conducted in Japanese. Another is that many NES translators do not read Japanese texts other than the ones they are asked to translate. Amazing, but true. Translation School Hype.  Providing translation training to a willing potential customer base is a fairly large business in Japan. Several magazines aimed at potential customers for such for-profit schools openly tout patent translation as an exciting, high-paid career. Surprisingly for what would appear to have a very narrow readership, such magazines can be found in most major bookstores. Numerous ads in such magazines (including ads masquerading as articles) hint--and some even flatly state--that you need not have a technical background to do patent translation. Naturally, the magazines are looking out for their advertisers, which are largely translation schools. This has surely lured people into patent translation who have a very low probability of success. Subject Matter Knowledge Translating the Words is Not Enough.  If the language in a patent must disclose an invention (and it must) and must enable a person skilled in the art to which the invention belongs to practice the invention (and it must), it is natural to assume that patent specifications are most often written for a "person skilled in the art." That person is reasonably assumed to have subject matter knowledge. For example, if the invention disclosed is a new method of fabricating a mask for manufacturing integrated circuits, it is reasonable to assume that the assumed reader of the patent specification has knowledge of such technology. Without an extreme effort and or specific study, it is highly unlikely that a patent translator with a humanities background is going to be able to do more than "translate the words" in the patent specification, and clearly more than merely superficially translating the words is required. Ability to Write the Target Text in the Style and at the Level Expected by the Reader Target Language Writing Skill.  Before we even get into the issue of patent style, there is the issue of writing ability in the target language (English in the case of Japanese-to-English translation). People in the Japanese-to-English translation industry will know from experience that very few translators are capable of writing in their non-dominant (non-native) language at a level that does not require extensive editing, which is most often better characterized as being rewriting. Even when an editor works on a text written in the non-dominant language of the writer, even if a very large amount of time is devoted to rewriting, it is often not possible to bring the document up to the level of quality that would have been possible if it were to have been translated by someone writing in their dominant language from the beginning. The Myth of Machine Translation The Expensive Gamble of Prior Art Translation.  Translation of prior art patent documents could be a very expensive gamble, and one in which the gambler (translation customer) might not even have a clue as to the payoff until a considerable amount of money is spent. Small bets are usually not sufficient, and upon seeing a prior art document translated at what could be a painfully high cost, the hindsight reaction can very well be "we should not have had this document translated." Grasping at Machine Translation Straws.  In an effort to avoid huge cash burns in search of relevant prior art, some translation consumers turn to translations created by machine translation systems. Some of the MT translation providers send the product of their MT systems to low-income venues for rewriting. Results, not surprisingly, do not approach what is possible with human translators used throughout the translation process, although users faced with a huge stack of hay in which to find a needle might see value in purchasing low-cost poorly done translations as a first step in the process, followed by proper translation of a smaller subset of documents executed by professional human translators after passing the full stack of hay through the MT needle-searching process. Machine Translation Systems Are Poorly Equipped to Compete with Humans.  Whereas human skills and diligence can result in result in high-quality translations, the machine translation system is seriously handicapped in all three competencies. Machine translation systems cannot "comprehend" the source text. Machine translation systems don't comprehend language in the way that is possible by humans. Machine translation systems cannot have "knowledge" of the subject matter. Unlike a human translator, a machine translation system, possessing information but not true knowledge, cannot do a reality check of its translation output based on real-life knowledge of the subject matter being translated. Here again, the human translator excels. Machine translation systems have no "dominant" or native language in common with any human. Machines have neither English nor Japanese as their "native" language. Asking a machine translation system to read the source text as would a native reader and then write a target (translated) text as a native writer is the height of folly. As we note below, however, machine translation systems are not the only approach to translation in which the translator is native to neither the source language nor the target language. The Shift of the Translation Process Offshore to Translators Who Compete with Machine Translation Systems by Not Having a Relevant Native Language Magic Done in the Dark: The Subcontracting Spiral and Where it Leads You and Your Documents.  Perhaps unbeknownst to most users of Japanese-to-English translation services, when a user order translations from a translation provider, the ensuing subcontracting and sub-subcontracting can take the documents to be translated to venues in which the translators--just like machine translation systems--have native competency in neither Japanese nor English. What is more amazing is that the translation providers might not even care or know (in the case of sub-subcontractor) about this machine-translation approach using humans with skill deficiencies to replace machine translation systems. The Internet has made this approach extremely easy. Reverse Auctions Enable Translators with Marginal Skills to Bid the Job Down until Document Reach Venues Unknown.  Again, generally unknowable to translation users, Japanese-to-English translation jobs from agencies are finding their way onto web-based reverse auctions, in which jobs most often go to the lowest bidder. A look at the people bidding on these Japanese-to-English jobs reveals that some are in places like India (putatively an English-speaking venue), but that many are in places such as China, the Middle East, and Eastern Europe. In many cases, these jobs are going to translators in venues in which translators, for political or economic reasons, will almost never be able to experience Japanese language directly in Japan, nor do they have English as their dominant language. In this sense, these translators are competing with machine translation systems. With the intervention of a translation broker, however, translation clients are generally shielded from this rather alarming situation. Essentially, the broker acts to perform "venue laundering." It does not appear that there are any statutes that require a translation seller to mark a translation "Made in China," for example. Internet Help for the Helpless Self-Help Mailing Lists Say it All.  Japanese-to-English translators use a number of tools to help them get through their translations. One of these tools is the mailing list for such translators. But some of the questions posed by members on such mailing lists--presumably with the hope that the questions will not be seen by their clients--clearly demonstrate that some of the posers of the questions should not be attempting the translations being asked about. Questions posed on such lists are often so basic as to reveal serious deficiencies in the most basic knowledge required to execute a proper translation. Imagine, if you will, a mailing list for US patent litigators, on which an attorney member asks what rule 30(b)6 of the Federal Rules for Civil Procedure is. Would you trust patent litigation to such an attorney? Questions about translation are just that basic on the mailing lists for translators. Strategies for Success Just what can be done? We have several suggestions, focusing on the process of vetting translation service providers before you latch onto one that sends your documents to someone even they cannot evaluate as a patent translator. More in our next issue.

www.lise.jp/ptb_014.html © Lise & Partners, Inc. Last modified: Thursday, 08-Jul-2010 08:10:19 GMT