No. 11 — January 14, 1984

Editing and publishing the newsletter require expenditures of much time, energy and money, and it is no longer practical to offer the newsletter free of charge. Even the donations sent in by readers have not alone been sufficient to cover the editor's expenditures of time and money. However, the newsletter was never intended to be a money-making venture, and the editor, convinced that it fills a definite need, is willing to continue publishing it without outside assistance for a while longer. Therefore, the editor plans to continue publishing the newsletter regularly as a one-man venture until June, 1984 (one year after publication of the first issue), and then to reassess plans and decide whether or not to continue to put it out single-handedly. Do readers have any comments about whether they wish the newsletter to be continued and how this should be done, or suggestions about any changes in policy? Could the newsletter be carried on after June, 1984 as a collective venture? Are there any volunteers who would like to collaborate in or take over the work of editing and publishing it for one more year, from June, 1984 until June 1985?

Editorial policy is to promote the exchange of both technical and cultural information which might be relevant to professional Japanese-English (or English-Japanese) translators. The newsletter is meant for active readers, and you are encouraged to write for publication in it. Undoubtedly your schedule is busy, but if you think the newsletter is valuable for you and you want to stay on the mailing list, you must either take the time to write something for publication, or else send in money, etc. to help pay for the expenses, or both. (Suggested minimum amount of donation for a non-contributor is $20 for the year May 1983–May 1984. Subscribers in Japan and Europe should add extra for postage to their donations.) Thanks to those who have sent in items for publication and/or money for the expenses, and dire warnings to those who send neither: your names will be removed from the mailing list. If there is an asterisk before your name on the envelope, that means that I am not sure whether you wish to continue to receive the newsletter. You must write to me indicating that you want to receive it, or your name will be dropped from the mailing list. (See dire warning above.) Those who need specific back issues should write me (enclosing a check sufficient to cover the postage) and specify which issues they want.

The distributor in Japan no longer has access to a copying machine, and unless other arrangements can be made readers in Japan will be receiving their issues (I hope temporarily) by sea mail from the editor. European readers will also be receiving their issues by sea mail. If there are any readers in Japan or in the U.K. who have access to a copying machine and who would be willing to act as volunteer distributors for their respective countries, please contact me.

Thanks to readers who have pointed out that it is possible to use a copying machine to print on both sides of the page. I know that it is physically possible to run the pages through twice, but whenever I do this on my machine (a Xerox 2600) there is a build-up of toner on the inside of the machinery. This causes smudging of all copies, and a service call is necessary. The maintenance personnel have warned me not to use the copying machine for this purpose, at least for extended runs. I am upgrading my copier from a 2600 to a 2830, which I am told can print sheets on both sides. Maybe I can do the next issue that way. DLP

* * *


In our issue No. 8, Frederik L. Schodt reviewed Japan's Modern Myth: The Language and Beyond, the book by the mythoclastic American professor Roy A. Miller. The ideas expressed by Miller struck me as being very controversial and relevant to the work of Japanese-English translators, and I was interested in what others had to say about the book and about the ideas. I was especially interested in knowing whether Miller's book has been published in Japanese translation and, if so, what has been the reaction to his ideas in Japan. We all know how interested the Japanese are in foreigners' views of them, and here at last, one would think, they have a very plain-spoken view of Japanese "myths" by a prominent and controversial "foreigner."

John D. Lamb wrote to the book's publisher, John Weatherhill, Inc., and received the following reply (dated November 30, 1983) from Pamela Pasti, assistant managing editor:

We certainly appreciate the amount of space and attention you and your contributors are giving to Japan's Modern Myth. It is especially encouraging since, to answer one of your questions, the response by Japanese publishers has been less than enthusiastic. To date, two have expressed great interest and then pulled out when they actually read the book. (In a similar vein, I was personally asked to review the book for a Japanese journal that retracted its request when someone finally checked out the contents.) We are still negotiating with interested firms, but I for one will not be surprised if they choose not to forgive Mr. Miller's "excesses," despite Frederik L. Schodt's willingness (and of course ours) to do so. In short, the need for serious critical books on Japan (not the aoi me no mita drivel) has yet to be recognized by Japanese publishers. I might add that sales in Japan have been surprisingly good, an indication that Japanese are purchasing the English edition — the public, at least, has a healthy curiosity about other views, even those that "cut to the bone."

If your readers know of any publishers who might be interested in taking on Mr. Miller's book, I invite them to contact us.

Wanting to obtain the reaction of a Japanese, I asked Hiroaki Sato of New York to read Miller's book and comment on it for me. Sato's reaction to the book was one of extreme antipathy, and he wasn't able to overcome his revulsion long enough to read through the entire book, but this is what be writes (December 7, 1983) about the first few chapters which he read:

About Roy Andrew Miller's JAPAN'S MODERN MYTH, I must say the book is far worse than his previous one, THE JAPANESE LANGUAGE IN CONTEMPORARY JAPAN (Hoover Institution, 1977). The only advantage I found in this book over the same author's hour-long lecture I once had to endure at a conference is that I was free to put it away after the initial chapters. At the conference with solemn-looking panelists, I didn't have the luxury of getting up and leaving the room.

Since I have read only the first few chapters, I'd like to mention just one aspect of the book: the author's utter laziness. Miller makes little effort to document his case. The bone of the book's contention appears to be that the Japanese give too much attention to their own language. But in making this assertion Miller gives, in Chapter One: The Making of the Myth, for example, only two concrete examples: JAPAN ECHO, an "English-language public-relations journal ...Widely distributed by Japanese embassies and consulates throughout the world," and "a direct-advertising brochure from a Tokyo film company" intended for "potential American customers."

Miller has an excuse. "It is, of course, impossible," he says, "to cite specific statistical data." To be sure, "specific" or "precise" (as he puts it elsewhere) data may be hard to come by, but figures for giving a general idea should not. Take publications. In Japan today, annual directories of publications are put out, in addition to traditional catalogues from publishers and dealers. Miller could have checked them and given rough estimates so obtained against comparable figures in the United States. The latest edition of the NIHON TOKEI NENPO (JAPAN STATISTICAL YEARBOOK) shows that in 1980 a total of 27,891 "new editions" (new books, revised editions, and reprints) were published in Japan, of which 565 titles were devoted to "philology." The yearbook also shows that of the 3,325 magazines published in the same year 9 were devoted to the "Japanese language." The percentage points of 2% and less than 0.3% do not indicate to me the kind of overwhelming interest that Miller says the Japanese show in their language.

What happened appears to be this: Miller sat around with a couple of Japanese and heard them say that Japanese is a "rich" language because of the existence of several words for something for which there may be only one word in English. He then decided to condemn the whole Japanese by writing books. How else can one explain Miller's contempt for assembling substantiating evidence to build his case? As Bernard Saint-Jacques has shown in reference to Miller's Hoover book, Miller has done the same thing before. (See "Language Attitudes in Contemporary Japan," THE JAPAN FOUNDATION NEWSLETTER, Vol. XI/Nos. 1-2, July 1983) If anything, JAPAN'S MODERN MYTH shows Miller has gotten lazier.

Because he has so little material to go on, Miller has to fill the space with verbiage. And because he is a professor, he declaims. The only other book filled with so much pompous verbosity that I have read is Miller's Hoover book, but it was shorter.

In the above-mentioned article "Language Attitudes in Contemporary Japan," Bernard Saint-Jacques, a professor in the Department of Linguistics at the University of British Columbia, examines the language attitudes directed in Japan towards non-Japanese speakers of Japanese. Dr. Saint-Jacques discusses the descriptions by Roy Andrew Miller (in his book The Japanese Language in Contemporary Japan) and by Eleanor Jorden (in an article entitled "Linguistic Fraternization: A Guide for the Gaizin") of the language attitudes of contemporary Japanese. Miller states his law of inverse returns (a foreigner is given less credit for his accomplishments the better he gets at the language), and Jorden argues that "there is a general reluctance among Japanese to talk with strangers." Dr. Saint-Jacques argues that these interpretations are based on insufficient evidence are are often mistaken.

Examining the "evidence" adduced by Miller for reaching his conclusions about the language attitudes of the contemporary Japanese, Dr. Saint-Jacques remarks: "The evidence, I suggest, is not what one would expect in a monograph written by a specialist. In point of fact, anybody who would make such absolute judgements about the language attitudes and the culture of another nation based on this type of evidence would either not be taken seriously or be suspected of biased views."

Concerning Miller's assertions that the Japanese have a 'mystical' attitude about their own language, Dr. Saint-Jacques remarks: "I see nothing wrong with 'linguistic mysticism.' Every language has its own unique and evocative power, formed by its cumulative cultural history. The Japanese are not unique in feeling that their language has a spirit of its own. Hirakawa Sukehiro, in his Defense of the Spirit of the Japanese Language (1981), has made it very clear that the notion of kotodama ("the spirit or soul of language") as set forth in Watanabe's 1974 article — the focus of the first chapter of Miller's monograph — has little to do with the prewar nationalistic myth."

Dr. Saint-Jacques quotes passages written by Arabs, Finns and Catalans which speak of a similar linkage between their language and their national spirit. I am sure that similar statements could be amassed from countries all over the world. The Welsh, the Bretons, the Estonians, the Flemish, the Poles, the Koreans, and so on.

In order to clarify the question of Japanese attitudes towards Japanese-speaking foreigners, Dr. Saint-Jacques conducted two sociolonguistic surveys in Japan. The first, conducted in Tokyo during October and November, 1981, was designed for foreigners who know Japanese and who live in Japan. A total of 150 subjects responded. Omitting the details, I will give merely an abstract of the results of the survey.

Question I. When you ask for information or services in correct Japanese, do Japanese answer you in Japanese? 96% answered YES.

Question II. Do you have the impression that the Japanese like to communicate with you in Japanese? 98% answered YES.

Question III. Do you think that the fact that you can speak Japanese creates suspicion and mistrust with the Japanese? 97% answered NO.

Question IV. Do you agree with the statement that the more progress you make in the language, the less it will help you to make friends and favorably impress the Japanese? 97% answered NO.

The second survey, conducted during the same period, was of 500 Japanese, randomly selected. They were interviewed in Japanese by a Japanese university student. Here is an abstract of the results:

Question I. Have you ever spoken Japanese with a foreigner? 79% answered YES.

Question II. When addressed in Japanese by a foreigner, do you answer in Japanese? 94% answered "USUALLY, YES."

Question III. Do you like to speak Japanese with foreigners? 96% answered YES.

The results appear to be an incontrovertible refutation, in facts and figures, of some of the basic arguments put forward by Miller. The following is the conclusion reached by Dr. Saint-Jacques:

At a time when there were very few foreigners who spoke Japanese, and when most foreigners in Japan were perceived as members of a military occupation, one could perhaps hypothesize that the language attitudes of Japanese could have been peculiar. Even then, this proposition would have to be corroborated by solid and substantial evidence. What seems to be very clear, however, is that the language attitudes of contemporary Japanese are not those described by Miller and Jorden. At the end of his short review of Miller's monograph, Higa has expressed the following desire: "This reviewer sincerely hopes that American policymakers will not be led by this monograph into believing that the Japanese are a bunch of irrational and xenophobic people." (236) I would like to extend this desire to all those who are friends and students of Japan.

Have other readers read Miller's writings? I would appreciate hearing your views about them.

* * *


A Tokyo company called Chūkintō Insatsu, which specializes in translations and printing, chiefly in Arabic and Persian, has developed a word processor capable of handling inputting and composing in these two languages and in English. Word processors for Arabic and Persian do not exist in the Middle East, and this is believed to be a world "first". Since there is much interest in the product in Middle East countries, the company intends to export it to them through trading companies.

The word processor developed by Chūkintō Insatsu (called Aracom System) is based on an NEC 16-bit personal computer (PC9801) and has a number of fonts, such as bold-face, in addition to the ordinary typeface for text. The Arabic type fonts were designed independently by the company.

Arabic uses 29 basic letters, but they assume different forms depending on their position in a word. Fonts for Arabic, Persian, and English are each contained in floppy disks, and choice of language fonts and editing and composing of input texts can be done simply by pressing buttons.

Two printers are supported. One is a plain-paper printer, and the other is an IBM electronic composer. The new equipment is expected to increase greatly the productivity of composing in the two languages, which in the past was done manually in Japan. Most Arabic texts printed in Japan are manuals for exported products, and the new word processor will make it possible to create data bases of frequently repeated texts. The price of the word processor was not mentioned. (Nikkei sangyō shimbun, December 7, 1983)

Nichimen Co., Ltd. also has developed an Arabic word processor jointly with a Tokyo software house called Nihon Kagaku Gijutsu Kenkyūsho. The word processor will be exported to the Middle East and North Africa and also to Europe and America beginning in April.

This word processor uses a 24-dot matrix printer. It can also be used for Persian and Urdu, as well as English words. In addition to word processing, the equipment also is equipped with personal computer functions. The sales price will be $10,000 per unit. Nichimen expects to ship 300 units during the first year. (Nikkei sangyō shimbun, Jan. 4, 1984)

* * *


Judy F. Wakabayashi of Kawagoe-shi, Saitama-ken, writes the following letter:

I have only been receiving your newsletter for the past two months, but am now enjoying going through the back issues that Mr. Lamb has sent me. I have found the definitions of "in" words such as "nekura" extremely useful, and hope this feature continues in the future. Even though I do more general translations than technical translations, one never knows what one will be called upon to translate, and so the glossaries are also bound to come in handy some day.

Past issues have taken up the question of the role of women in translation and whether or not translation is a rewarding career. I have a small child to look after, and so I have found that translation is ideal in that it allows me to work at home at hours to suit myself and my son. I have never felt any handicap from being a woman (possibly because most of my work comes through agencies) — on the contrary, I have found that most companies are so glad to have a native speaker of English translate their material that they are not even fussy about whether or not I have any background in that subject. This is obviously not an ideal situation, but it does mean that I have more work than I can handle. This seems to be quite a contrast to the situation of translators living outside of Japan. For me, the challenges and freedom of a translation career are certainly rewarding.

I have found that the newsletter rather over-emphasizes articles on computers and word processors. I myself use a word processor for translating, and obviously this tool is a real boon for translators, but it is only a tool, and I would prefer to see more space devoted to articles that deal more directly with translating Japanese, or with the practical uses of computers in translating (e.g. creation of lists of terms not in normal dictionaries). Despite this criticism, I am very much looking forward to the next issue.

[Editor's note: The newsletter is edited outside of Japan and serves readers on three continents. One way that it can help is by informing translators in one country about the problems encountered by translators in another country. Translators living in Japan are, of course, not inconvenienced, but the fact that Japanese-language word processors made by Japanese manufacturers are not available outside of Japan has been causing considerable trouble for individuals and institutions in the U.S. I have received many inquiries about how Japanese-language word processors can he purchased here, which equipment is recommended, and how to have it serviced. If this were not such a serious problem for translators outside of Japan, we would not have to emphasize it so much in the newsletter. Readers in Japan could help by trying to convince Japanese manufacturers that there is a need for their products outside of Japan. - DLP]

* * *


Frequent inquiries have been received about whether Japanese-language word processors can be purchased in Japan and used in the U.S. A possible solution to the problem might be to use Japanese-made personal computers with word-processing software. There is a full list of this software on pages 283–294 of the book Waado purosessa no subete, Saishinpan (Newton bessatsu, Kyōikusha, Oct. 1983). Some readers have written in about this question. Bernard Susser of Kyoto, Japan writes (December 28, 1983):

Word processors are mentioned frequently in issues 9 and 10 but your readers should be aware of another option: the use of word processing software with a personal computer. I have an NEC 9801 with an optional kanji ROM; I can use American-made word processing software (I have WordStar) and also can run Japanese word processing software. There is a wide variety of the latter now available, from ¥20,000 to over ¥100,000; most allow romaji input — you can type in Japanese using the standard QWERTY keyboard, and do not have to relearn how to type. One also needs a dot matrix printer with kanji printing capability.

This is what Terry Kleeman of Albany, California writes (December 10, 1983) about this:

There are by now, it seems, any number of good, reasonably priced word processors on the market in Japan. I have seen advertised an entire system, including printer, for ¥480,000. The problem is that they are being sold in Japan. They require a different voltage power source, but this can be rectified by purchasing a transformer or modifying the power supply. More important, they are not warranted, and probably not serviceable over here. Who could afford to send the whole thing back to Japan every time something goes wrong? Further, many of the dedicated word processors, like word processors over here, use unique operating systems and data formats that make them incompatible with other systems. I suggest that one approach is to see what Japanese computers are available here, and ask whether they can be modified to use Japanese script. I think at least one model, the NEC 8800, can. By purchasing the kanji ROM card for this model (from Japan), and a word processing program that allows romaji input, you should have access to hiragana, katakana, and JIS Level I kanji (approximately 3500). Unfortunately, NEC has never developed a Level II ROM card for this computer, nor, to the best of my knowledge, has any private vendor developed such a character set for the 8800, either in hard or software. Because this computer is among the most popular in Japan, a wide variety of software is available for it. Further, it can use standard CP/M or, with a 16-bit expansion card, MS-DOS (i.e. IBM-PC) software. Graphics resolution is 640x400, permitting a 40x25 display of characters in a 16x16 matrix. The Epson QC-10 is another model for which this may be possible. In Japan Epson markets a word processor built around this microcomputer, the Exword 10. It has an optional JIS Level II font, and 24x24 dot kanji. I do not know how difficult it would be to convert the American version to something like the Japanese, or if American dealers would service it after the conversion was made. Sanyo and Olivetli might also be worth looking into in this regard. Finally, there is talk that the IBM 5550 may be introduced in the U.S. in the future. In the most recent issue of Byte (November 1983) there is an article on the 5550.

There remains the question of how to get the material printed out. Most systems depend on specialized kanji printers with character fonts onboard in ROM. Obviously, these are available only in Japan, and since printers tend to require frequent service, it probably would not be feasible for anyone without a lot of technical experience to buy from there. I know that for the NEC 8800 there are word processing programs that permit printout through a normal Epson FX-80, but this limits print quality to the level of the 16x16 screen display set. I do not think that 16x16 gives an acceptable level of clarity or elegance for anything but the most informal communication. Most important, portions of complex, multi-stroke characters turn into blobs of ink. A 24x24 font gives legible, well-proportioned characters, but they are still noticeably "dotty." State-of-the-art systems now use 40x40 or finer with laser printers. If I might just interject my own personal comment, I think that all of these systems are fundamentally wrongheaded in their approach to this question. With continuing advances in the resolution of printers and monitors each font is destined to be outdated relatively soon after it is laboriously created. Further, as matrices become finer, bit-by-bit definition becomes inefficient. I suspect the system of the future will consist of graphs defined by vector plotting, perhaps using the elliptical equations of the recently introduced Conographics system for the IBM-PC.

In your last issue of the newsletter you mentioned a new catalogue of Japanese word processing equipment. A friend in Japan recently sent me a copy of the Autumn 1983 issue of OA Joho (¥1,800). This issue has a listing of all the major dedicated word processors and minicomputer-word processors available in Japan. The product descriptions are short and unsatisfying, giving almost no information on the hardware doing all the work, but it does reproduce samples of the printed output of each system (except, for some reason, the IBM 5550). I was surprised by the variations in quality of output among machines all using the same 24x24 matrix.

Dan Kanagy, a computer programmer in San Francisco, also writes (December 14, 1983) the following on the same subject:

There appears to be some interest among translators in the United States for a word processor which can handle both Japanese and English text. We know that such things exist in Japan, but what to do across the Pacific?

I believe the Epson QX-10 is the best compromise for the near future. This is a stripped-down version of the Epson QC-l0 marketed in Japan as a Japanese word processor. By adding a kanji ROM, replacing the English-only keyboard with a Japanese/English keyboard, and replacing the printer with a 24 by 24 dot Epson printer, you will have the same machine as the QC-10. Of course, you will also need Japanese word-processing software, for a fully functioning Japanese word processor.

The Epson QX-10 is usually sold with an Epson RX-80F/T printer, a monitor, two disk drives, and bundled software for around $3,000. My understanding is that the printer already has the capability of printing kanji text, given appropriate software. The drawback is that the print head is not dense enough to distinctly print high stroke-count kanji. You really need the 24 by 24 dot printer for professional work.

The Epson QC-10 is sold in Japan for around 1,000,000 yen. This includes the high-density printer, kanji ROM, a monitor, two disk drives, and bundled software. The printer itself makes for a good portion of the price.

The advantage of going this route is that the Epson QX-10 is widely marketed in the United States, so that if your computer needs servicing, more likely than not, it can be done here.

Now how to get the components for the conversion is another matter. Talk to your local Epson QX-l0 dealer or write to Epson. Perhaps you know someone returning from Japan who can help you out. All of the above should be considered at your own risk. Theoretically this should all work, and there is a translator in the midst of it, but no converted machine cranking out Japanese text before our eyes. I am not certain if the Japanese text processing software that comes with the QC-10 handles Japanese text exclusively, or will process mixed Japanese and English text. Nonetheless, even without a conversion, I believe that you will find the QX-10 to be more than satisfactory for your English word-processing needs, with a potential to handle Japanese text, if you choose to go that route in the future.

In closing, I would be interested in hearing what computers are being used for your word processing, and praise or criticism about the word-processing software that you use. Also if any of you have another solution for getting a Japanese word processor in the United States, I would like to hear about it.

From the Editor: So would I. Those who are using computers with English-language word-processing software, please write in and give us your evaluations of the hardware and software you are using.

About the Epson QX-l0, Richard Willis writes (December 24, 1983) to suggest that "anyone contemplating buying an Epson QX-l0 read Jerry Pournelle's review of the machine in the August '83 Byte ... he found the system very easy to use, but very slow. He was specifically focussing on the VALDOCS version of the machine; the CF/N version is probably no slower than any other Z-80 based system. But potential buyers should spend a few hours using any computer system before actually purchasing it, especially for 'esoteric' applications."

Incidentally, prospective purchasers of Japanese-made personal computers should be reminded that the market share of Japanese-made products in the U.S. is not a large one. According to an article in the Nikkan kōgyō shimbun (December 28, 1983), Epson's sales of personal computers in the U.S. have recently been running at only about 5,000 units a month. This appears to be the largest sales in the U.S. of any Japanese-made personal computer (NEC's 5200-05 APC sold only 2,500 units during the first six months of 1983, and its PC series — including the 8800, 8200, 8000, and 6000 — has sold 30,000 units during the first six months of 1983 in the entire world). On the other hand, the entire U.S. market for personal computers this year is expected to be enormous, reaching a total of some 8,000,000 units, and next year IBM will enter the lower end of the market with its inexpensive PC Junior. Epson is now studying strategies for increasing its sales of personal computers from 5,000 to 10,000 units per month, but I would expect it to be an uphill battle for any Japanese exports. Epson is renowned as a manufacturer of small-size printers. It now has a 30% market share in Japan and is aiming at increasing its market share in the U.S., which has gone down lately but was at one time 70%. (Nikkan kōgyō shimbun Jan. 4, 1984)

I have a suggestion. There are many Japanese-language word processors available in Japan, but none are being sold in this country because the Japanese manufacturers do not believe there is a market for them and are afraid to assume responsibilities for furnishing maintenance for them. We know, on the other hand, that there is a market here, although it is small and dispersed geographically. Therefore, if we want to buy the products here, we must make a collective effort to convince the manufacturers that someone wants to buy them. My suggestion is to approach the problem collectively rather than on an individual basis. A manufacturer might ignore one individual but would be less likely to do so if confronted by an organized pressure group of purchasers. Form a pool of five or six prospective purchasers, preferably all in one area. Call yourselves something like the "Northern California Japanese Word-Processor Prospective Buyers Group." Choose Japanese manufacturers such as Epson or NEC or Toshiba which already are known to have sales and maintenance organizations in the area. They would be more sensitive to pressure. Approach them with an offer to buy four or five units from whichever manufacturer will supply them and promise to service them. Let them know that the group will boycott other products of the manufacturer if the manufacturer refuses to meet the needs of purchasers in the area.

I personally feel that a public-spirited manufacturer ought to be willing to meet the needs of customers and to sell them whatever products they need even it the number of purchasers issmall. Why should manufacturers be permitted to sell only products which they believe will be profitable and to deny customers the right to buy other products simply because the manufacturer is "afraid" of furnishing service? In fact, why do not the manufacturers try to put out products which can be used with a minimum of hardware and software changes for a large variety of different languages? Why should Japanese manufacturers export Arabic word processors to the U.S. and Europe and refuse to export Japanese-language word processors to the same countries? See other articles in this issue. - DLP

* * *


In his letter (December 10, 1983), Terry Kleeman writes the following about choosing translation as a profession. I decided to reproduce it here because I know that other readers are similarly situated and might have some useful advice to offer.

The discussions of technical translation as a profession have been very interesting. I am in a position similar to that of the hypothetical person you advise, one contemplating entering this field full time. I am a graduate student in the Ph.D. program in Oriental Languages at Berkeley. To supplement my income I do Japanese-English translation and interpreting in my spare time, and am also a teaching assistant for first-year Japanese. If I continue on my present course, my future will include: two years until taking my orals, an indeterminate period while writing my thesis (four or five years?) several years of "gypsy scholar" life, taking one-year non-tenure track appointments, and with luck, eventually a tenure track position at a university. Even after this I will be poorly paid, will have to live in constant fear of offending a senior faculty member until I receive tenure, and will be under constant pressure to publish while coping with a heavy teaching load. I think there is a good possibility that following this course I will eventually achieve what for me is a desirable lifestyle, including a decent income, time for both leisure and for the pursuit of my scholarly interests, and a certain degree of respect from the community. However, the tribulations that lie along that road make me wonder if this is in fact my best course of action. The prospect of a decade or more of uncertainty and financial insecurity is anything but salutary. For this reason, I have thought of devoting myself fulltime to translation and interpreting. What I would want from a life of translation includes: financial security, founded on a steady supply of work at good wages, reasonable leisure time, and the opportunity to live in Japan or, better yet, visit it frequently. Translating is hard, taxing work, and I would hope that I could make a comfortable living with an average of no more than 30-35 hours per week in front of the CRT, although I would not be adverse to working considerably more than this for short periods of time if a particular job required it. Your example, in continuing to publish your own research, is particularly inspiring. I have invested over ten years in learning classical Chinese and religious Taoism, and would never want to totally abandon this field. As a translator I would concentrate on the computer field, and to do this well I would have to retrain myself to a certain degree, but this would not be unpleasant. The financial logistics of it would pose a greater problem. I think that I should also consider two other options, teaching Japanese full-time, and teaching English in Japan. The first would be very difficult outside of a university environment, just not enough demand. The second is more appealing. Ideally, I would find a well-paying job at a private junior college. This would afford an adequate income and a good deal of leisure, to be spent on my own research or free-lance translation and interpreting as I choose. I am not sure if I would need specific training is ESL to do this. I hope I could get by with just my M.A., but I understand the market is tightening in Japan. At any rate, I have lain my case before you. Any comments or advice concerning my assessment of my position would be most appreciated.

Finally, let me address a couple questions you have brought up in the course of previous newsletters. Japanese is indeed a more difficult language to learn than common European languages, and I think that translators to or from Japanese deserve a premium over and above that paid to people translating between Western languages. It takes much longer to learn Japanese well, and having learned it, it is much more difficult to translate between totally unrelated languages. As to the question of whether knowledge of Japanese or of technical matters is of greater importance, I must confess that my own weak technical background makes me somewhat less than impartial. Still, it seems to me that even in the most technical material I have translated there have been very colloquial expressions that I cannot imagine someone understanding without having lived in Japan and spoken Japanese for some time. Perhaps this would be less important when doing abstracts, but I wonder if someone without a strong background in Japanese could do a competent, faithful translation. Also. I should note that when I have consulted experts on certain questions, they have often been as puzzled as I about the specific reference of certain technical terms. It is a difficult question, but I also can't help but wonder why someone with that degree of technical knowledge would not himself be working in industry, rather than hiding himself among us word-manglers.

[From the editor: Not all translators are full-time career translators. There are many scientists and specialists working in industry who like to translate in their spare time in order to earn extra money. And (haven't you heard?) there are also some unemployed chemists and physicists, some of whom turn to translating as a temporary stopgap until they find steady jobs in industry or research again. A beginning translator should be prepared to compete, and, as is so often the case in human life, no one is ever going to promise you security. - DLP]

* * *


Nippon Electric Co., Ltd. and Bravice International Inc. have joined together to develop an epoch-making machine interpreting system. The target date for completion of the system is 1990. According to a front-page article in the Nikkei sangyō shimbun (December 14, 1983), the companies' executives say that their final goal is to use the system for telephone conversations, enabling persons not knowing each other's language to converse with each other over the telephone.

The basis of this machine interpreting system is to be the machine translation system which Bravice is to put on sale in January, 1984 (see No. 9, p. 2–3). This system will be combined with voice recognition and voice synthesis devices in which NEC excels. A voice recognition system will be used as the input device, and a voice synthesis system will serve as the output device. When speech is input, the machine translation system will translate it automatically, and the translation in another language will be supplied to the other person from the output device.

The following three problems must be solved before developing the system: (1) How to raise the ability to recognize voices of specific speakers; (2) How to raise the precision of the machine translation system; (3) How to develop a conversion technology for use when outputting the translated contents into the voice synthesis device. Both companies say that they will mobilize their specialists in these fields to solve these pending problems.

The companies think that existing computer hardware will be quite sufficient for use with the new system, and software is what mainly requires to be developed. In fact, if artificial intelligence technology becomes available soon, it may be possible to take advantage of it, and in that case it is possible that the NEC-Bravice interpreting system may be realized even sooner than 1990, says the Nikkei sangyō shimbun

- - - - - - - - - - -


According to an article in the Nikkei sangyō shimbun (December 16, 1983), Computer Services Corporation (CSK, capitalized at ¥15,646,000,000, 3,421 employees; see no. 8, p. 11–12; no. 9, p. 2–3) recently established a research institution, a company called "CSK Sogo Kenkyujo," aiming at system development utilizing artificial intelligence. This indicates that CSK is moving towards metamorphosizing itself into an all-round information-processing enterprise. Interestingly, the new company will employ Mitsuharu Yata, formerly Senior Researcher at the Electrotechnical Laboratory (under MITI's Agency of Industrial Science and Technology). Yata is considered to be a foremost authority in the artificial intelligence field.

The new company CSK Sogo Kenkyujo will be capitalized at ¥300,000,000, and the company's president will be Isao Ohkawa, president of the parent company CSK. It will aim at developing various types of application systems, including research and development in the application of artificial intelligence to machine translation.

CSK and Bravice International recently became affiliated through stock exchanges and decided to embark on joint research in artificial intelligence. The CSK research will be done chiefly by the newly established CSK Sogo Kenkyujo.

Mr. Yata's employment by the new company has attracted industry-wide attention. When he resigned the Electrotechnical Laboratory as of October 15, he was Senior Researcher in the Program Research Office, Computer Science Department, and played an important role at the Electrotechnical Laboratory in the Fifth Generation Computer project being implemented under the leadership of MITI. Mr. Yata move into the private sector is considered to be a blow to the MITI project but an advantage to CSK. Having obtained the services of a first-rank authority on artificial intelligence, CSK expects to make rapid progress into this new field.

- - - - - - - - - - -

Something significant is definitely brewing here. As reported in our issues No. 8 and 9, Bravice in 1982 purchased a controlling interest in Weidner Communications Corporation, an American company engaged in developing computerized translation systems. Then Computer Services Corporation in October, 1983 acquired a 30% interest in Bravice, and now the news of the NEC-Bravice tie-up comes on top of that. The NEC-Bravice collaboration is especially surprising because of the disparity in size between the two companies. NEC, one of Japan's largest corporations, has some 27,596 employees and is capitalized at ¥58,351,000,000. Bravice is a small corporation capitalized at ¥220,000,000 and has only some 80 employees.

CSK's establishment of its own research entity with the collaboration of Mr. Yata seems to indicate that the CSK-Bravice-Weidner axis, probably with the behind-the-scenes backing of NEC, may in the future try to play a leading role in the development of machine translation and other artificial intelligence systems in Japan, perhaps even attempting to rival the Fifth Generation Project. Concerning disappointment with the Fifth Generation Project, see an article elsewhere in this issue.

* * *


I found the following entry in the 1984 edition of Gendai yoogono kiso chishiki. It describes the same phenomenon mentioned by Dan Kanagy (no. 10, p. 8):


英語が日本人の生活の中に定着しつつある。かつて中国語が日本に渡来し、定着し、やまと ことばに漢字をあてはめた時代を第一次平安時代というなら今や第二次平安時代である。たと えば「私はピンク色のワイシャツを着ているが、今日はちょっと暑い。そこでエアコンを入れ る。ネクタイはしていないが、スラックスをはきうしろのポケットにはハンカチが入っている 」という具合。戸は「ドア」に、新装開店は「オープン」にそして英語の動詞は「ダブる」「 ネグる」などのように、日本語化しただけでなく語尾変化まではじめている。アメリカのコロ ンビア大学のハーバード・バッシン教授は、この日本語の変化から日本の文化や社会の変り方 を鋭く指摘している。たとえば、いま流行の初歩的なパソコン用文法書の一節はこうである。 「RUNコマンドは、オープン状態にあるすべてのファイルのクローズおよぴメモリ内容の初 期化を行ってから、プログラムをメモリにロードし、その後プログラムを実行します。」

* * *


Don Cyril Gorham of Silver Spring, Maryland, writes the following (December 17, 1983):

As regards Japanese academic as well as scientific and technical organizations, I have found the two publications identified hereunder quite useful. Apart from the data on the organizations themselves, it sometimes happens that this type of publication is the only place where you can locate the sometimes rather arbitrary English names for some of these organizations:

Zenkoku Gakukyōkai Sōran (Directory of the Learned Societies in Japan) compiled by Nippon Gakujutsu Kaigi Jimu Kyoku (Executive Office of the Science Council of Japan). Published Feb. 25, 1981; price ¥5,200. Organizations are grouped into two major categories: cultural and natural sciences, each of which is, of course, further subdivided. Major sub-groupings of the cultural category are: literature, philosophy and history, law and political sciences and economic sciences. For the second group (natural science), the sub-groupings are: pure sciences, engineering, agricultural sciences, medical sciences, and university-affiliated organizations. There is an index by name of organization as well as by the subject. 643 pages.

Zenkoku Kenkyū Kikan Sōran (Directory of the Research Institutes and Laboratories in Japan) edited by Nippon Gakujutsu Kaigi Jimu Kyoku (Secretariat [sic] of the Science Council of Japan). Published March 30, 1974; price ¥4,600. Again an authoritative and quite comprehensive listing including: governmental (232 organizations), public (665), special juridical organization affiliated (40), non-profit organization affiliated (184), and others (16). Also listed are research institutes and laboratories connected with national and private universities and with private firms. There is a comprehensive set of indexes, rather rare for Japanese publications. A total of ten different indexes in English and Japanese languages. 1112 pages.

As a matter of fact, there are two other publications of interest in this area, both put out by the Gyōsei Kanrichō: the Shingikai Sōran (General Survey of Deliberation Councils) and the Tokushu Hōjin Sōran (General Survey of Special Juridical Organizations). If anybody is interested, I can supply additional bibliographic information.

* * *


Some readers have written to comment about ideas advanced in no. 10, p. 13–16. Don Cyril Gorham writes (December 17, 1983):

Re Ted Ohtani's comments on the extent to which a translator should or should not go in restructuring or in creating sentences, I feel as follows. Restructuring to the minimum extended needed to produce smooth and understandable target-language renditions: yes, I feel as a professional this is part of the job. However, when it comes to "creating sentences," if that means that you are adding some thought that is not in the original but that you may feel helps clarify matters, this is quite risky and is to be avoided. At the same time, "creating sentences" in the sense of breaking up a long original sentences in the source document I feel is quite legitimate.

Charles Ridley of Palo Alto, California, writes (December 14, 1983):

Incidentally, I find that my propensity to divide up Japanese sentences into shorter elements varies with my degree of mental fatigue. In the morning when I am sleepy and in the evening when I am worn out, I tend to run to shorter sentences in English. During the day when I am most alert, I am more apt to work out longer more complex productions.

* * *


The Institute for New Generation Computer Technology (ICOT), the central agency of the Fifth-Generation Computer Project being promoted by MITI, has introduced a VAX11 Model 710 minicomputer manufactured by the American company Digital Equipment Corporation (DEC) for use in its research in artificial intelligence. This was reported in a front-page article in the Nikkei sangyō shimbun (December 22, 1983).

ICOT, headed by Takuma Yamamoto (president of Fujitsu), was founded in April, 1982 and is an ambitious, high-risk project. The Fifth-Generation Project is scheduled to complete its fundamental research this year, and full-scale applied research will begin next year. At its founding the project had the participation of the government (MITI's Electrotechnical Laboratory), the computer industry (eight firms: Fujitsu, Hitachi, NEC, Mitsubishi, Matsushita, Oki, Sharp and Toshiba) and academics. It has been reported that the computer industry is becoming dissatisfied with the way the government bureaucrats and academics are running the project (see no. 4, p. 1–3), and the Nikkei sangyō shimbun article reports that there are currently "seven computer and electrical manufacturers" including Fujitsu and NEC participating in it. (One company has evidently dropped out in the meantime.) As reported elsewhere in this issue, Mitsuharu Yata, formerly Senior Researcher at the Electrotechnical Laboratory (under MITI's Agency of Industrial Science and Technology), has recently gone to work for a CSK-funded research institute devoted to artificial intelligence. Yata is considered to be a foremost authority in the artificial intelligence field, and his departure is considered to be a blow to the ICOT project.

In addition to the VAX11 minicomputer, ICOT is also introducing a programming language called C-PROLOG for artificial intelligence use which was developed jointly by DEC and researchers at Edinburgh in Scotland. It is said to be a version of PROLOG which DEC has improved so that it can be used with minicomputers of medium or smaller sizes.

Late in 1982 also, ICOT introduced two sets of large-size DEC minocomputers, the DEC System 20. This, reports the Nikkei sangyō shimbun, indicated that Japan's Fifth-Generation Program started out in a manner completely dependent on American-made computers. ICOT has already put into operation a system utilizing PROLOG with the two sets of DEC System 20 which it introduced last year. However, the System 20 proved to be "too big" in its scale and its performance properties, and ICOT was unable to make full use of it. For this reason, it decided to introduce the VAX11 and C-PROLOG at this time.

The newspaper also reports that budgetary considerations played a role in the decision. Using large-size equipment to develop a system proved to be too expensive, and it is expected that the medium-size system introduced at this time will make it possible to cut down the development expenses. In order to promote the joint research, VAX11 minicomputers and CPROLOG will also be delivered shortly to the seven Japanese manufacturers who are participating in the Fifth-Generation Project.

The aim of the ICOT project was to develop domestic Japanese technology in the artificial intelligence field, and it was widely assumed that the Japanese would become pioneers in world-wide research in the field. Hopes had been entertained that the Fifth-Generation Project would make significant contributions to machine translation, among other things. Now disappointment is being expressed because the ICOT project is to rely totally on "American-made" products both for the minicomputer hardware and for the software to be used in the artificial intelligence research, which is the nucleus of the whole project. The Nikkei sangyō shimbun concludes that this may cast "dark shadows"on the future of ICOT, the aim of which was to develop domestic Japanese technology.

However, an article in the Nihon keizai shimbun on December 28, 1983, said that ICOT has announced that it had completed part of the hardware for a "reasoning machine" (suiron mashin), which it will use as a tool for developing new software to be used in the Fifth Generation Computer. According to ICOT, this is the first time anywhere in the world that a prototype has been successfully completed. The "reasoning machine" executes efficiently the syllogisms which are the basis for reasoning by means of "pattern-matching' functions and "back-track" functions.

* * *

言葉の研究 = ニコヨンとクロヨン

構造も発音も似通っているこの二つの俗語は一見同じ部類に属しそうなものですが、見かけにもよらず時代的背景も使用分野も違っていて、混同されてはまずい言葉です。エコヨンの方は古く、旧版の「広辞苑」にも出ています。定蓑は、日傭(ひやとい)労働者のこと。「広辞 苑」によると、もと職業安定所が出していた定額日給が二百四十円だったところからいわれる。 つまり、百円を「一個」として、二個四(にこよん)というのが語源だそうです。終戦直後の時代を想わわせる、ちょっと古めかしい響きのある言葉です。 

眼を転じてクロヨンという俗語は、労働者をいうのではなく、税金関係の用語です。再近手に入った1984年版の「現代用語の基礎知識」によると、タロヨンとは、国が納税者から税金を徴収するにあたり、三つの異った種類の納税者から徴収する税金の捕捉割合が違うということを言い表している言葉だそうです。つまり、最も高い捕捉率はサラリーマン(給与所得者 )の場合でぺ所得の9.0%が補足されている。営業所得者は60%、農業所得者は40%の所得しか捕捉されていないことをいい、このことから「九・六・四(くろよん)」というようになったそうです。


* * *

XEROX Office Systems Division, 3450 Hillview Avenue, Palo Alto,
California 94304. OSD-T8301 August, 1983

This is a paper describing the general user-interface context of the Fuji Xerox 8012-J Star system. The author, Joseph D. Becker, has the title of Manager, International Advanced Development, XEROX Office Systems Division. The Star (8010/8012) workstation is a largescreen personal computer system announced by Xerox in April, 1981. The Japanese typing method for the JStar was developed in close co-operation between Xerox and Fuji Xerox from 1976 to 1982.

According to the paper, the Fuji Xerox "80l2-J Star" system was demonstrated in Tokyo in October 1981, and came on the market in October 1982. Prototype Chinese typing capability with the JStar was demonstrated in Washington, D.C. in September 1982, although Chinese will not appear in the product until 1984.

JStar software for Japanese typing uses a "virtual keyboard" image with hiragana symbols in place of the normal English keyboard. The "virtual keyboard" for Japanese appears as a rectangular window on the large screen, which has a resolution of about 1/2 mm (1 point or 1/72 inch) and a total of more than 825,000 individually-addressable pixels. The large screen measures 17" diagonal, and a clearly legible virtual keyboard occupies less than 11% of the screen space, leaving plenty of space for documents, etc.

The typist enters the desired word phonetically in hiragana. To convert the spelling of a word to kanji, the typist presses the SPACE bar, which functions as a special "lookup" key in the virtual keyboard. The system locates the kanji spelling of the word in a dictionary stored on its hard disk. If there are homophones, the system presents the alternatives to the typist, who chooses the desired one.

The choice between homophones is presented to the typist by displaying a virtual keyboard a display set on which the choice is indicated on a picture of a keyboard. To choose a homophone, the typist presses one of the keys on the keyboard corresponding to the desired characters displayed on the screen.

JStar supports all 6,349 JIS Level-I plus Level-II kanji as well as all the hiragana and katakana, plus the European and punctuation characters. These characters are all available for display on the workstation screen in six different size/style combinations: 8-point size Mincho style, 18-point size Gothic style, and 10- and 12-point sizes in both Mincho and Gothic styles. The total storage for display character images alone is more than 1 Megabyte at the workstation. The JStar typing dictionary contains (currently) over 110,000 Japanese words, occupying somewhat more than 2 Megabytes of storage at the workstation. Thus, more than 3 Megabytes of the workstation hard disk are devoted to Japanese typing-related data. Unfortunately, the JStar does not currently have a bitmap editing facility allowing the typist to draw a character (a kanji, a logo, or an arbitrary symbol) which is not provided in the dictionary. Such "gaiji processing capabilities" are among the most attractive features provided in some Japanese-language word processors made in Japan.

The JStar has a grammatical processing capacity which can analyze long inflected forms of Japanese words. The unit to be analyzed is the bunsetsu, defined as a unit starting with 1 or more kanji and ending with 0 or more hiragana or other attached characters.

According to the paper, the Chinese-language typing capability is based on the Mandarin phonetic alphabet known as bopomofo. Tone distinctions are omitted in inputting. There is also an automatic romanization-to-bopomofo conversion input system for those who do not want to learn the bopomofo keyboard layout.

The paper does not describe the printer used to output documents in the JStar system, but I understand it is a print server and laser printer costing about $30,000.

Richard Willis, who kindly sent me a copy of the paper by Joseph D. Becker, writes that the latest AOC Office Automation Systems MarketGuide lists the Xerox 8012 12M Star Workstation at $19,088 (1 to 4 units). The price drops to $12,380 in "quantity." The 8044 10M Print Server & Laser Printer is listed at $29,655. A one-year maintenance contract on these two systems, he adds, runs $6240! According to information which I printed in a previous issue (see no. 2, p. 13), the multiple-language software for the Star costs $1200 without Japanese, and the Japanese software costs an additional $1200; screen display fonts and a dictionary cost an additional $800; and Japanese printer fonts cost $1000. A Japanese-variant keyboard is also available and costs an extra $500. Thus, a fully configured Star system for Japanese typing would cost over $53,000, plus annual maintenance charges of over $6,000. Although the equipment sounds fabulous, writes Willis, "none of us 'peasants' (read 'small businessmen') can afford them." "I think what struck me most about the IBM 5550 was that, in contrast to IBM's American PC and small office systems, it was very aggressively priced." If it weren't, he adds, "it wouldn't last a week in Japan."

It would be interesting to know if Xerox has sold any of these expensive 8012-J Star systems, and if so who has bought them. I doubt that any translators will be able to afford them, which is unfortunate because they more than anyone else would benefit from such a system. But read on:

* * *


Digital Research Japan, the Japanese branch of the American software company Digital Research (Pacific Grove, California), announced in December that it has developed an operating system for personal computers which is capable of converting Japanese kana texts into kanji on a bunsetsu basis, rather than on the individual kanji or jukugo level. The software will be supplied to nine companies (including Mitsubishi Electric, Tokyo Sanyo and Sord) from February to May, 1984. Japanese texts are first input in kana or romaji in bunsetsu units. Then a conversion key is pressed, and the texts are analyzed grammatically and subjected to the necessary kana-kanji conversion to produce mixed texts including both kana and kanji. About 30,000 lexemes (jiritsugo) can be recorded in a dictionary on a floppy disk, and the user can update or change the dictionary contents by means of a dictionary file maintenance utility. The advantage of the system is that the dictionary need contain only the uninflected forms of each lexeme, and this makes it possible to increase the dictionary contents to double or more.

Conversion on the bunsetsu level has been provided in dedicated word-processors, but it was hitherto impossible in word-processing software with 8-bit computers. The software is incorporated in the CP/M system developed by Digital Research. It can be used in either 16-bit or 8-bit personal computers. The price has not yet been decided. (Nikkan kōgyō shimbun December 20, 1983)

* * *


U.S. IBM's deliveries of its personal computers during 1984 are expected to grow to some 2,500,000 units — about three times the number delivered during 1983, which are expected to total about 850,000. This is due to the appearance of the PCJr (code-named "Peanut"), which will cost less than $700 and is expected to come on the market early in 1984. A new highend model personal computer, based on the Motorola 8600 microcomputer, is also expected to be announced some time during the year. Thus, IBM is expected to win over an ever higher percentage of the sales of personal computers, according to an article I saw in the December 24, 1983 issue of the Nikkan kōgyō shimbun.

IBM is said to expect to sell more than 1,000,000 PCJr's annually, and many American consumers are deferring purchases of home computers until after the PCJr appears on the market in 1984. This has caused severe drops in sales of competing products marketed by Atari and Texas Instruments. IBM is expected to completely surpass Apple during 1984 and establish an unshakable position as the top manufacturer in the personal computer field.

An indicator of IBM's plans may be the number of personal-computer keyboards manufactured for it on an OEM basis by the Japanese company Minebea. Minebea has been manufacturing large numbers of keyboards for IBM personal computers at its Singapore factory (NMB Singapore) since 1982. According to an article in the Nikkei sangyō shimbun (December 23, 1983), the Minebea Singapore factory's capacity for manufacturing personal-computer keyboards will be increased by more than 3 times the 1983 level, to about 500,000 units per year. At the beginning of 1983 the plant's output was only about 10,000 a month, but as of December 1983 this had been increased to 35,000 a month. In January 1984 the monthly output is expected to reach more than 40,000, reflecting strong demands from IBM. Various types of automated production equipment will be introduced into the Singapore factory, quality control will be improved greatly, and the number of production lines there will be increased from 2 to 3.

IBM watchers already have been speculating widely about the collaboration between IBM and Matsushita Electric Industrial Co. in manufacturing the PCJr home computer. It has been reported that IBM has contracted with Matsushita to produce its PCJr, but that discussions about setting up a joint venture have been bogged down. (See no. 6, p. 14) Now IBM appears to be engaged in discussions with Sony Corp. about collaboration in the manufacture of the PCJr. A lead article in the January 1, 1984 issue of the Nikkan kōgyō shimbun reports that it was learned on December 31 that IBM U.S. and Sony are engaged in wide-ranging negotiations about either the establishment of a joint IBM-Sony company or the possibility of IBM's consigning manufacturing of the home computer to Sony. Sony, which is interested in advancing from the home-electronics field into the information equipment area, is reported to be displaying a lively interest in the discussions, which may result in an agreement early in 1984.

When the IBM top executives visited Japan late in October, 1983, they began discussions with Sony representatives while at the same time continuing their meetings with Matsushita. The IBM-Matsushita discussions on a joint venture have been dragging on fruitlessly for three years with no agreement in sight. This led IBM to step up its approaches towards Sony. In an effort to further the negotiations between both companies, IBM decided to adopt a highprecision television receiver for use in its high-quality graphic system, the 5080, which went on sale in November, and also to adopt 3.5" type floppy disk drives for use in its personal computers - another move of strategic advantage to Sony, although IBM and Sony have yet to reach a definitive OEM agreement for supplying Sony products in these areas.

IBM's plan appears to be to adopt Sony-made floppy disk drives, provided that IBM and Sony reach an agreement to collaborate in the personalcomputer field. IBM is believed to be insisting on very stiff conditions in terms of costs in its personal-computer discussions. On the other hand, Sony is reported to be responding positively for the following reasons:

(1) If Sony collaborates with IBM, this will enhance Sony's "brand image" when it advances into the field of information equipment.

(2) Even if there is little advantage for Sony in collaborating with IBM in the personal-computer field, the arrangement will be highly advantageous for Sony at least in obtaining contracts for related products.

(3) Sony has been planning to reverse its policy of selling its products exclusively under its own brand names. It is already planning to mount a vigorous campaign to sell its semiconductors to other outside manufacturers.

According to the Nikkan kōgyō shimbun president Norio Ohga of Sony confirmed that discussions are taking place with IBM but refused to comment on the contents.

* * *


In an effort to improve its competitive power against other rival manufacturers such as Fujitsu and NEC, IBM Japan will begin in 1984 to export its Multistation 5550 workstation to Taiwan and Southeast Asia. IBM Japan's Fujisawa Plant completed at the end of 1983 a model change converting the Japanese-language processing functions to Chinese kanji functions. IBM Japan plans to export at least 1,000 units of the workstations to Taiwan during 1984. Exports to other areas in Southeast Asia such as Singapore, Malaysia and Hongkong are also planned.

IBM Japan is also preparing units for export to mainland China, which uses kanji differing somewhat from those used in Taiwan. IBM Japan, Kanematsu-Gosho Ltd., and International Business Machines Corp. have an agreement on collaborating in exporting OA equipment to China, and it has been reported that the trio will make an all-out effort to sell the Multistation 5550 personal computer to China.

Korea is also regarded as a promising market for the 5550, and IBM Japan plans to develop a 5550 with Korean-language specifications. (Nikkan kōgyō shimbun, January 3, 1984)

* * *


An article in the Japan Times (December 30, 1983) listed a number of new Japanese words coined during 1983. For example, hard-working businessmen are called "worproholics" (word-processor workaholics), and OL (office ladies) are now called CL (computer ladies). Some women are called anmarizoku ("unmarried set") because they give priority to their jobs over marriage.

3D-zoku refers to children who always talk back to their parents, beginning their sentences with words like datte, demo or doose.

The biggest concerns among college boys are PPMM (pro wrestling, plastic models, microcomputers and manga).

There is a new style of writing Japanese called the "ABC buntai" in which letters of the English alphabet are substituted for similar-sounding Japanese phonemes. For example, tanoshii is written "tanoC" (with tano written in hiragana, and a capital C standing for shii). The words ooini waraeru are written "Oi ni waraL" (with wara written with the kanji for "laugh"). (This is reported also on page 1000 of the 1984 edition of Gendai yoogo no kiso chishiki.

* * *


We are all aware of the crucial importance of borrowings in Japanese from English. In this connection, Alexander Shkolnik remarks (see no. 9, p. 8) that some Russian translators working from Japanese to Russian who do not know English are seriously inconvenienced by English words written in katakana. One translator he knew found even simple words like ootorooda ("autoloader") or beroozu ("bellows") to be insuperable problems.

We English speakers are also sometimes confronted with analogous problems when we find Japanese borrowings from languages other than English. Apart from classic loan-words like the monstrous noie-zahharihikaito which has been in dictionaries for decades, there are some other German and French loan-words used in narrowly specialized fields which might present problems to translators. For example, Charles Ridley of Palo Alto recently came across a term anzattsu in a patent relating to steel manufacturing. It turned out to be the Japanese version of the German word Ansatz which, he found, in addition to meaning "beginning" also means "deposit, crust or encrustation." The latter was the meaning intended for it on this occasion. Speaking of steelmaking, I am reminded of the Japanese reading anbaa or anba for "invar," which evidently is a word of French origin. The Japanese phonetic reading would have been quite different if the word had been a borrowing from English.

I was once asked for help in deciphering a Japanese loan-word written in katakana muuton. The only possibility which occurred to me was the French word mouton which seems to have dozens of technical meanings other than the usual significance of "sheep."

Have other readers ever come across technical loan words from languages other than English which have presented problems to them? How have you solved these problems? (Let's not, for the moment, get into a discussion of non-technical words like maronie or kasutera which we can already find in ordinary dictionaries such as the fourth edition of Kenkyusha's New Japanese-English Dictionary.

* * *


The 1984 edition of Honyaku jiten (¥980, published as a bessatsu of The English Journal by a company called Aruku) has a number of articles which might be of interest to us.

First mention must go to the article about translating comics by Frederik L. Schodt (Manga no honyaku wa doko made kanoo ka) pp. 122–125. There is a photograph of Fred on p. 122.

On pages 151–154 there is an article about machine translation. The article mentions the Bravice-Weidner system (see elsewhere in this issue) and the SYSTRAN system. At the end of the publication (pages 187–191) there is a list of translation companies in Tokyo and other areas.

On page 38 is an article by Moriyuki Okochi of Tom giving eight do's and don'ts for Japanese technical translators (1. Avoid direct, word-forword translation. 2. Use a dictionary. 3. "Clear, Concise, to the Point" 4. Use expressions which have only one possible meaning. 5. Do not neglect study of specialized fields. 6. Be humble about your English. 7. Read English written by natives. 8. Meet your deadlines)

On pages 132–134 there is a general article describing the current situation in technical translation in Japan. Here is an excerpt from the article describing the fees paid to translators in Japan and a graph showing some of the fees paid by clients to translation agencies for J-E and E-J translations.








[Scanned Image]

On page 9 there is an article describing an "Electronics Translators Bank" operated since 1982 in Tokyo by an advertising company called Ad Bureau. 70% of the work is from English to Japanese, with special emphasis on preparing technical manuals about NC machine tools. Some of the clients are manufacturers in the U.S. Fifteen translators are now registered with the company in the two top ranks and are supplied with work. The company intends to increase the number to 50. The fees paid by the client to the agency are from 3,000 and 5,000. The agency takes a 12% commission and pays the remainder to the translator.

* * *


An article published in the Sept./Oct. '83 issue of the Society of Federal Linguists' newsletter reported a talk given by Richard M. Bender, former chief of the Joint Publications Research Service (JPRS) (see no. 10, p. 18–19). The article is reprinted in the latest issue of the ATA Chronicle.

According to Mr. Bender, about 850 freelance translators and 150 typists are currently under contract to JPRS. Many of them have worked for JPRS for many years. The demand and volume appear to be increasing all the time. JPRS handles between 50 and 60 languages on a regular basis. About 85 million words per year are translated at the current level. This includes about 25-28 million words from Russian, 8–10 million from French, and 6 million each from Chinese, German and Spanish. The JPRS has its own inhouse editorial staff who review and edit the translations before they are published. The editors are also available for discussion of problems encountered by translators.

Translators are paid for their work on a "sliding scale" which takes into account the quality of their work and the demand for their skills. The highest rate now paid by JPRS is $75 per thousand words for translations of scientific and technical material from certain "exotic" languages, but they must be "camera-ready" (i.e., requiring no editing or retyping). Because of the current economic situation, the greatest demand from JPRS's customers right now is for Japanese science and technology.

A fairly new translator of non-technical Romance languages may be paid as low as the mid-$20's per thousand words (counted from the original, not from the product). A good French or Spanish generalist is paid about $30 per thousand, while Russian S&T translators average in the mid-to-upper $30's. These rates are for camera-ready copy. The rates are reviewed every year, and can be raised if a translator has exhibited consistently high quality. In addition to the base rate, a "research" rate may also be added to compensate translators for "non-translating" time spent in library research, inserting illustrations, etc.

According to Mr. Bender, the best way for a freelance JPRS translator to increase his or her income is to increase production without affecting quality. Many translators are now using word processors to speed up their production; some have realized increases of 30 to 50%. In the future, he said, translators may even be able to send their work electronically "on-line" to word processors in JPRS officers.

Good translators can make a good living from JPRS and reside just about anywhere they choose, Mr. Bender said. One Russian translator makes $50,000 a year. One lives in Tucson, another in Hawaii. JPRS tries to start new translators as high as possible, and to help others improve. Some have "graduated" to running their own translation services, or have taken jobs as full-time translators in government and industry. JPRS is especially looking for Japanese and Russian science and technology translators, but Chinese and East European languages are also in demand, and West European languages are always useful.

* * *


Back in issue no. 6, I wrote an article on the "crisis in the Japanese language," in which I spoke about the proliferation of kana words and also mentioned a "kind of nostalgic reawakening of interest in kanji." "Some say," I wrote, "that this vogue is caused by the dissemination of Japanese-language word processors, on which even an inexperienced operator can, by inputting something phonetically, call up a whole list of fresh, unfamiliar-looking kanji combinations on the screen." I said I thought it was "rather nice that, the Japanese themselves find their own native script so novel and fresh." (no. 6, p. 3)

Now an article in the Nihon keizai shimbun (December 11, 1983) points out that word processors have not only changed Japanese offices, but may also become a powerful force which will influence the Japanese language to such an extent as to shake Japan's culture itself. The effect of word processors has been felt especially in the kanji area, or rather in the attitudes of the Japanese people towards the use of kanji in their writing system.

Professor Toshihiko Kato of Gakushuin University has been using a word processor for the past three years. He notes that the number of kanji he uses with a word processor is about 20% more than the number he uses when he writes by hand. One needs only to press the kana-kanji conversion key to call up the kanji on the screen, even if one does not know the kanji. "I make a conscious effort to confine myself to the tooyoo kanji but most people begin to use kanji indiscriminately. The manufacturers are also competing with each other to increase the number of kanji they input. It is an extremely convenient machine, but one also senses a dangerous omen that it may change the language in an undesirable way."

The advocates of using kana for writing the Japanese language have been seriously affected by the new trend. The Kanamoji Kai, an organization which has been advocating the use of kana since 1920, is facing the biggest crisis in its entire history as a result of the rapid spread of Japanese-language word processors. The main argument in favor of abolishing kanji has been the fact that they are cumbersome and difficult to write. But by using a word processor, a person can overcome the difficulties of the Japanese ideographic script by merely pressing a button.

The kanamoji movement's support seems to be crumbling away even in its home territory. The late Chubei Ito, who was one of the leaders of the Kanamoji Kai during the prewar period, introduced the use of kanamoji in in-house documents in his company, C. Itoh & Co., Ltd. The company used katakana for its in-house documents until July, 1983, when it changed over to hiragana-kanji documents. The reason again: the availability of word processors.

The managing director of the Kanamoji Kai, Yoshio Tsukahara, says that when word processors first came on sale, some the members of his organization were quite upset. The group is discussing steps to be taken, but it is undoubtedly true that the pro-kanji forces have gained strength. Some members now feel that, when word processors come into common use in the home, there will no longer be any need for people to memorize kanji at all, and the result will be, not a greater familiarity with kanji, but rather a decrease in people's ability to use them. The kanamoji group is beginning to apply pressure on word-processor manufacturers in an effort to have them provide automatic spacing between words so that word processors could be used conveniently for inputting documents in kana.

Shigeru Watanabe, the dean of a Tokyo junior college, is an authority on systems engineering and also on kanji. He advocates returning to the point of origin of the kanji as pictograms in the Chinese Oracle Bone inscriptions. He emphasizes their pictorial character and encourages the creation of new kanji. He points out that word processors can be used for this purpose. He also proposes that more specialized kanji be added to word processors. (Nihon keizai shimbun, November 4, 1983)

Another result of the dissemination of word processors has been a critical attitude towards the tooyoo kanji, the restricted list of 1945 kanji which has been in use since the early postwar period. Some persons feel that limiting the use of kanji in the interest of simplification has had the effect of impoverishing the written language, making it insipid and lacking in expressive depth. The increased use of word processors could exert a powerful impetus favoring abolition of kanji restrictions. Is perhaps a rebirth of Japanese kanji writing in the cards?

* * *


Bernard Susser of Kyoto, Japan has kindly sent us the following information about five dictionaries he feels might be useful:

1. NIHON-TEKI KEIEI O SETSUMEI SURU TAME NO JISHO (Hikino Takeshi and Nagano Akira, eds., Tokyo: Daiyamondo-sha, 1982).

引野剛司・長野晃編「日本的経営を説明するための辞書」。東京:ダイヤモンド社、1982 。

Useful for economic terms. A nice feature is the liberal use of quotations from magazines like Time, Fortune etc. and from American- books on Japanese management.

2. DENTŌ-TEKI KŌGEIHIN GIJUTSU JITEN (Dentō-teki kōgeihin sangyō shinkō kyōkai, eds., Tokyo: Gurafikku-sha, 1980).

伝統的工芸品産業振興協会編「伝統的工芸品技術辞典」。東京:グラフィック社、1980 。

Although unfortunately arranged by prefecture (!), it does have a good index. If you want to know the histories of the various folkcrafts or how these objects are made, this is an excellent source. The flowcharts describing the processes of manufacture are quite good.

3. EIWA KAGAKU YŌGO JITEN (Tokyo: Kōdansha, 1975; Buruu bakkusu B-268).

「英和科学用語辞典」。東京:講談社、1976、ブルーバックス B-268 。

Technical translators won't need this but for people who occasionally run into scientific writing this is very handy. Has a Japanese-English index.

4. JITEN NO JITEN (Tsukuda Jitsuo and Inamura Tetsugen, eds., Tokyo: Bunwa shobo, 1975).

佃実夫・稲村徹元編「辞典の辞典」。東京:文和書房、1975 。

A handy introduction to dictionaries and other reference works in all fields. Unfortunately now out of date; other defects include the arrangement, which practically guarantees that you have to look in two places to find what you want, and the lack of full bibliographical information. Still, did you know that there was a dictionary of motorcycle terminology and a cement and concrete yearbook?


「別冊 The English Journal エレクトロニックス時代の英語」1982年9月。

Has a nice list of Japanese-English technical dictionaries (pp. 161–163). It also has a long list of Japanese manuals for writing various kinds of technical English (pp. 158–161).

* * *


This is the third part of a glossary of terms on chemistry and allied fields contributed by D.A. Fraser, 22 Gresley Road, London, Nl9 33Z. (See No. 9, p. 16–17 for my introductory comments.) Mr. Fraser will be glad to have corrections or confirmations.

keikōtai luminophore
keikkei fluorometer
keikō zōhaku senryō fluorescent whitening agent
keikō hyōhakuzai fluorescent whitening agent
keiryō konkurito light concrete
keikabutu silicide
keitai anteisei dimensional stability
keiso gomu silicon rubber
kekkyoku tuyosa strength limit
kendaku bunri elutriation, clarification
kenma polishing, buffing
kenkyō microscopic examination
kensyuku crimp, crispation
kessyōkakuzai nucleation reagent
kenryōsen calibration curve
kizi sekken solid soap ?
kihakuka rarefaction
kihatusei enki tisso volatile basic nitrogen (VBN)
kikō gas cavity
kihōzai foaming agent
kikasei evaporability, volatility
kinkei mould
kire-tozai chelating agent
kire-to tekitei chelatometry, complexometric titration
kire-toka chelation
kisui bunriki steam separator
kōbunkai photolysis
kōdo luminous intensity
kōgaku kasseitai optically active substance
kōdo tekitei photometric titration
kōgakutai enantiomer
kōenso-shori chlorination of mains water after filtration
kōka seni vulcanized fibres
kōki bussitu aromatic principle
kokuenka graphitization
kokyōka photo-induced crosslinking
komazi coomassie (stain)
kondakukei nephelometer
konkō amalgam
kongō blending (of rubber)
kōonsō thermostat
konren milling or kneading (plastics man., with roller)
kōon sokutei pyrometry
kosō zyūgōhō solid phase polymerisation method
kōsoku ekitai kuromatoguiahu-hō high performance liquid chromatography (HPLC)
kōsosei photoelasticity
kōtakuzai brightener (in electroplating)
kottan animal charcoal
kōzyō osui industrial (or factory) sewage
kōtyōryoku limit tensile strength
kōwa amalgam
ko-ziraito cordierite
kōzyō kōbutu colloidal minerals
kōzyūgō photopolymerisation
kūkan sokudo space velocity (SV)
kūki nukikō loophole
kūki syōka sekkai air-slaked lime
kumorido haze
kunzyōzai fumigant
kuraiogen freezing mixture, cryogen
kurasureisyon clathiation
kureizen teni Claisen rearrangement
kurezoru sekkeneki liquid cresol oil
kusuriyaki cyaniding
kyōka crosslinkage
kyaputan Captan (ag.chem.)
kyodai bunsi macromolecule
kyōdo tenacity (of fibres)
kyōkazai reinforcing agent
kyōmaku laminar interface
kyōmizai taste corrector
kyyakusan conjugate acid
kyōsen (tomosen ?) ground stopper
kyōryō ketugō crosslinkage
kyōryokuzai synergist
kyōzatubutu impurity
kyūro filtered off at the pump
kyūkōdo light absorption coefficient
kyūkakuteki nucleophilic
kyūdensiteki electrophilic
kyūsyō spherulite
kyūsyūban filter
kyūsyūkan cuvette
kyūsyō zyutenzai spherical filler
makururon Macrolon
maikuro bizu micro beads
maikuro suhueya microsphere
mannebu Maneb (fungicide)
maration malathion (ag.chemical)
marutitto maltitol
masuta battchi master batch
marutoru maltol
mekarubamu mecarbam (insecticide)
mengai henkaku shindo perpendicular (out-of-plane) deformation vibration
mesu-appu seems to mean "to fill up to measure" (in a measuring flask)[from German messen + English up ???]
merubinaito merwinite
metarotionein metallothionein
metarikon-ho metallizing by spraying
miru be-su mill base (in paint manufacture)
metoru metol
mikaku kensa organoleptic test
miru potto pot mill (ball mill)
miseru Can mean either micelle (micell) or "miscella" (different meaning)
miri tōryō milliequivalent
mitudo kōbaikan density gradient tube
momento moment (in fabrics testing)
monogen monogenetic (solution)
mozō kami vellum paper
moru nitsu molar fraction
Moru-en Mohr's salt
muhennōsei inertness
naiōryoku internal stress
naisō suru to charge
nasu-kei hurasuko three-necked flask
nankazai plasticizer
nannensei noninflammable property
nendansei viscoelasticity
nekkuin "neck-in" (plastics man.)
nendo gurauto kō argillization
nenekisitu mucin
nenketusei caking value
nenritu coefficient of viscosity
nensyō gasu exhaust gas
nensyō seisan gasu combustion gas
nensyōsei flammability
neriko seikei dough moulding (plastics man.)
netuzyūryō bunseki thermogravimetric analysis
niga-yu niggerseed oil ?? (from seeds of Guizotia oleifera)
ni ringu knee ring (in fibres man.)
niziku ensin biaxial stretching
niziku osidasiki twin screw extruder
nizyūsen (sanzyūsen .... ) doublet, triplet, quadruplet (peaks in NNR spectroscopy)
nobi kakōsei elongation characteristic
nobi ritu relative extension
nokori kasu anode slime
nyūkayō emulsifiable (polyethylene etc.)
nyurando syokubai Nieuland catalyst
nyūzai emulsifiable liquid (in ag. chem. jargon)
omori gravity
orefin jugoyu polyolefins
origoma- oligomer
origotō oligosaccharide
orusin orcin, orcinol
pairen Pylene (trade mark)
oya-ion parent ion (in mass spectrometry)
pa-karaizingu parkerizing (rustproofing)
parapurekkusu Paraplex (trade mark)
parako-to paraquat (herbicide)
pereteiza pelletizer
peruron Perlon (polyamide fibre)
peberu miru pebble mill (= bell mill)
piresuroido pyrethroid
ponti punch (plastics man.)
purasutozoru Plastisol
pozi positive (in photography)
purasutikku haihin no kaisyō sairiyō collection and regeneration of plastic waste
purkatto precut (column) in chromatography
purometorin prometryne (herbicide)
raito uo-ta light water (firefighting foam)
reban levan (polysaccharide)
ramikon Lomicon (trade mark for polypropylene)
ramine-shyon seikei extrusion coating, laminating
ratora ne Rattler test value (powder metall.)
Raue-hō Laue method
reibai cooling agent, coolant
reikan kakō cold processing
reikakusen frost line
reinetu siken heat test
reitansu cement grout
rentan-hō bricuetting
renzoku kihō intercommunicating cells (plastics)
reorozi- rheology
reopurekkusu Reoplex (trade mark)
ribingu jūgō living polymerization
reppe-hō Reppe process
Ressi ondo memori Reaumur scale
resurin rethrin

Let me know if any of you turn up information which might be of interest to professionals working in the field, and I will incorporate it in another issue. Keep those tidbits coming in. January 14, 1984

Donald L. Philippi
715 Tenth Ave.
San Francisco, CA 94118
(415) 752-7735

* * *