Musings of a Japanese Translator in New York
Minoru Mochizuki presents a personal account of what it’s like to be a Japanese translator making a living in the United States. He then explains why he is grateful to Japanese inventors and TRADOS.
I am a semi-retired person. I spent about 40 years working basically as a mechanical engineer; first for a couple of machine manufacturers in Japan and in the United States, and then in a consulting position with another firm in New York City. Since then (May 1995), I have been a self-employed translator, i.e., the sole proprietor of my own business working primarily from home. I have to admit that the consulting role I undertook in the last chapter of my life as an employee included many hours of in-house translation and interpreting work, so I was sort of paving my way to becoming a full-time translator.
What I mean by “translation” is translation of documents written in one language (source language) to another (target language). What I mean by “documents” includes all kinds of printed matter, of public or proprietary nature, typed or handwritten, in hard copies, electronic files of various formats, or faxed. The results of my translation are normally delivered to my clients primarily as electronic files. In my case, both the source and target languages are limited to English and Japanese. In other words, I translate between English and Japanese in both directions. When I say English, I mean American English.
As to the size of the documents I translate, they are normally somewhere between 1,000 and 10,000 English words, but I recently translated a document of approximately 80,000 words. At the other end of the scale, I often do business cards that contain less than 50 words.
My field of translation is what you might call “industrial translation,” as distinct from literary translation. I translate mostly patents, contracts and agreements, operator's manuals, financial, scientific, and technical papers, sales brochures, and things of that nature.
There are a number of reasons why I concentrate on these kinds of documents:
As to the reason for (1) above, let’s look at one source of statistics available from the United States Patent and Trademark Office (USPTO): “Patent Counts – States and Countries of Origin, Calendar Year 2001,” the latest figures provided on its official website. According to the USPTO, the total number of utility patents*) granted during the calendar year 2001 was 166,045, of which 87,610 (52.8%) were granted to U.S. inventors and 78,435 (47.2%) to foreign inventors. (Patent origin is determined by the residence of the first-named inventor, as displayed on the face of the patent.)
(* “Utility patents” include the regular industrial inventions, excluding design patents - it is slightly misleading to a reader from a country in which there is a category called “utility model,” in addition to the regular patents, but the U.S. utility patents include both.)
Now, the interesting fact is that, of the total of 78,435 awarded to foreign inventors, 33,224 (42%) were to Japanese inventors! The most popular country of origin after Japan was Germany, at a very distant second with only 11,276 patents granted. Now, you may wish to ponder about the quality of these patents if you are concerned about the trend of inventions. However, my point is elsewhere, i.e., the required quantity of translation indicated by this statistic.
This abundance of Japanese patents among the total number of the U.S. patents granted (20%) has been reasonably consistent in the last few decades, based on other statistics available from USPTO. This means that 30,000 or more patent application documents need to be translated each year from Japanese into English. If there are only 1,000 qualified translators for this type of work, each translator will have to translate approximately thirty patents per year, or two to three patents a month. From my own experience, this estimate is not too far off the mark. We have plenty of work to be done for patents alone, let alone other types of translation work. At present, approximately 50% of my work is patent translation. I currently use computer-aided manual translation for most of my patent translations, which I will explain in detail later.
Regarding the reason for (2) above, the relative ease of translating industrial documents as compared to literary texts, suffice it to say that industrial documents are free of the ethnic factors that are often deeply ingrained in fictional text that deals with particular cultures and customs of the country of origin. Many terms and concepts are often unique to certain countries or societies, and they form an essential part of such texts, which cause insoluble problems for translators. This is the basic reason why the application of translation technology should be, and is, limited to industrial translations.
I have no doubt about the reason for (3) above, the higher earnings per unit time of work as compared to fictional texts. Being a literary translator, you may translate only a few novels per year at the most. I have done some calculations on how much I could earn based on available data, and the result is very disappointing. I have concluded that literary translations are done for love, not for money. I basically translate for money, so that my approach to translation is not too different from that of a person who owns and operates a machine shop. In other words, productivity, business volume, payment, and job quality coupled with client patronage, are the key factors that describe my business.
In this context, now comes an important question: is there a smart way to improve those factors?
It was during the IJET (International Japanese-English Translation) Conference in Sheffield, England in 1997 that I first learned about the possibility of improving those factors. A panel discussion entitled, “The Future of Machine and Computer-aided Translation” was presented. I don't recall now what sorts of opinions were presented, but I vaguely remember that the reactions from the floor, namely rank and file translators, were of suspicion and denial.
However, I was fascinated by a demonstration given by TRADOS, and I immediately decided to investigate further. After my initial research, I concluded (without touching the product) that it must be useful for the translation of documents with a lot of repetition of the same or similar sentences, such as computer user manuals. Since I was avoiding so-called “localization projects" on the assumption that they must be awfully boring (short, non-descriptive sentences), I thought I would have no use for tools such as translation memory. Nevertheless, I decided to purchase the product from TRADOS simply to experiment.
After many months of frustration, I am now a different person, i.e., a happy user of TRADOS. There are a few reasons for this about-face. One is that TRADOS finally became user-friendly and fully operational for double-byte languages such as Japanese. In fact, the last meaningful update in that regard came as recently as early 2003. Also, I found it quite useful and practical for many patent translations. I am now using it for more than half of the patents I translate.
TRADOS is not a machine translation tool per se, but it is definitely a clever application of computer technology to language translation work. The product does not come cheaply, but I would say it's worth it if you are involved in certain types of translation work. The essential concept is storing and building up sentences with identical syntax and terms in order to improve efficiency of translation, as well as to achieve a better quality (I mean, uniformity!). It is useless for documents that are completely free of repetitions, such as novels.
The next goal for me is to promote (in my microscopic way) the concept of controlled language, which I picked up during the EAMT/CLAW (European Association for Machine Translation/Controlled Language Applications Workshop) Conference held in May in Dublin, if I can. I think we can improve the quality of business, science and industrial Japanese texts by the introduction of controlled language. In its struggle for survival in the wave of globalization, I believe that Japan will benefit from controlled language, as we need smoother translations between Japanese and English. Most Japanese do not realize that if they exchange agreements with U.S. firms they are, in fact, putting their lives under the jurisdiction of the laws of Singapore.
It’s time for me to sign off… I must return to TRADOS and the next patent translation that is due by the end of the week.
graduated from the Tokyo Institute of Technology in 1956 in mechanical engineering and received a Fulbright scholarship in 1962 to do graduate studies in the U.S. He then worked close to forty years as a mechanical engineer for machine manufacturing companies, as well as for a trading company in Japan and the U.S. He is currently a full-time, self-employed industrial translator in Port Washington, New York in the U.S. where he lives with his wife. The couple has one son who now works as a New York lawyer in Tokyo.
Reprinted by permission from the Globalization Insider,
29 July 2003, Volume XII, Issue 3.3.
Copyright the Localization Industry Standards Association
(Globalization Insider: www.localization.org, LISA: www.lisa.org)
and S.M.P. Marketing Sarl (SMP) 2004