Subscribe Share/Bookmark

Sunday, April 4, 2010

Translation Technology

Translation Technology and the Translator
John Hutchins (University of East Anglia, Norwich)

1. Introduction
Translators are perhaps the most critical audience for presentations about the automation of translation. Many of them will agree with comments made by J.E.Holmström in a report on scientific and technical dictionaries submitted to Unesco in 1949. Having heard that some researchers were investigating the possibilities, he thought that "the resulting literary style would be atrocious and fuller of 'howlers' and false values than the worst that any human translator produces". The reason was that "translation is an art; something which at every step involves personal choice between uncodifiable alternatives; not merely direct substitutions of equated sets of symbols but choices of values dependent for their soundness on the whole antecedent education and personality of the translator." His comments preceded by five years the first tentative demonstration of a prototype system, and were based on pure speculation. Nevertheless, such comments have been repeated again and again by translators for nearly fifty years, and no doubt they shall be heard again in the next fifty.
However, we shall see that computer-based translation systems are not rivals to human translators, but they are aids to enable them to increase productivity in technical translation or they provide means of translating material which no human translator has ever attempted.

Helping Japanese Software Go Global
Anthony Wong. (Asia Pacific)
SOURCE: The Globalization Insider

While localization for the Japanese market is a common topic in the West, the challenges of localizing Japanese software for other markets are much less well known. In this article, Anthony Wong examines the general environment and the technical, cultural and linguistic issues involved in going global from the Japanese perspective.

Since the early 1980s, Japan has emerged in several respects as one of the world's major industrial powers. It has already overtaken Germany, to become the second largest global economy. In economic terms, Japan is prominent almost everywhere. It also depends for its livelihood on its economic relations with distant parts of the world, and its products are valued globally.

Japan has had unparalleled success at blending its own culture with elements from the West and elsewhere in Asia. Its economy is export oriented, and internationalizing its products and localizing them to suit other parts of the world has always been done very successfully.

Apart from the raw materials it is forced to import, software is one of the few areas in which the terms of trade for Japan are not positive. Software is still predominately an import item. This is not too much of a surprise, though, as almost 80% of the software used in the world is supplied by the US, and the rest comes mainly from various European countries.

It is well known that much research has been done and many empirical studies have been carried out on the process of localizing software to suit the Japanese market. Not so much, however, is heard about how Japanese companies internationalize and localize their software for the global market. Although Japan does not export a lot of its software, in areas like industrial automation, precision engineering and the like, there is still strong demand for Japanese software to go global.

Why, then, are these applications still relatively unheard of and undocumented?

One reason is perhaps the fact that most of the work done in this direction is of no great interest to the West. Back in the early 1970s, many tertiary education and research institutes in Japan had already started addressing these problems. A lot of papers and documents designed to help Japanese companies internationalize their products for the global market already exist. A very good example is the proliferation of Japanese support for mainframes in other parts of the world (especially Asia), which existed long before a lot of countries started looking into using computers.

Compared to the effort required to adapt Western software to the Japanese market, localizing Japanese software to suit other markets is relatively easy. However, there are still many issues (both technical and cultural) that need to be carefully addressed.

Technical Issues
As has already been mentioned above, Japanese software has so far failed to achieve a status similar to that of automobile or consumer electronics in terms of worldwide popularity. Even those firms which are producing excellent quality software have no urgent need to look beyond the Japanese coastline, since Japan itself is already a pretty big market.

In the area of software internationalization, these Japanese have to use a more advanced model right from the beginning. Typically, Japanese software has to support Japanese, English and at least one other language (usually Chinese, Korean, or other Latin-based scripts). Japanese is written in a mixture of well over 2,000 Han characters (Kanji), most of which can be read in a variety of ways, and two very different phonetic systems derived from Kanji, which are called Hiragana and Katakana and are collectively known as Kana. This multiple writing system produces not only what is possibly the world’s most complicated language in common use today, but also requires Japanese software developers to think internationally right from the initial design and analysis stage.

For example, Japanese software developers will need to look into issues such as avoiding conflicts between their code-page and other code-pages using the same code-points (this is similar to most localization issues). The fact that many different standards coexist in the Japanese market certainly makes the task even more daunting.

One possible solution is, of course, Unicode. However, this is more of a long-term solution, since the advent of Unicode is only recent and not many systems are able to support it as yet. The short-term approach is more in favor of a locale-based solution. Apart from the many proprietary solutions available in the mainframe area, most developers are quite happy with the well-known software internationalization and localization standards such as POSIX, ANSI and the X-Consortium's XPG4 (X/Open Portability Guide, System Interface Definition Issue 4). Naturally, Japan has always been one of the most important contributors of ideas in these areas.

A fact that is not so well known to the West is the relative popularity of Japanese software in other Asian countries like Korea and China (including Taiwan). Japanese engineers can easily handle the specific hardware systems in these countries, since there are even more such specific systems in Japan itself.

Received wisdom is that if a piece of software can support Japanese well, it will be relatively easy from a technical point of view to localize it into other languages.

Cultural and Linguistic Issues
For Japanese software publishers, the greater challenge when localizing into other cultures is probably not a technical one, but a cultural and linguistic one.

It is a well-known fact that getting Japanese to English translation resources is much harder that the better-known business in the other direction (i.e. from English to Japanese). Translating English into Japanese is already very costly, and it is not uncommon to have to multiply this cost by a factor of 2 or 3 for a similar amount of work being translated from Japanese into English.

The scarcity of this resource is, of course, the main reason for this. The fact is that translations from Japanese to other languages are mainly done by those few foreigners who know the social and cultural aspects of Japanese society very well, and who are at the same time competent in the Japanese language (typically, the holders of the Level 1 JLPT test). This situation is of course not too bad if one is working between different Asian languages, but it is at least several times harder to find Westerners that fit the bill.

Despite the fact that the Japanese language borrows, modifies and adopts foreign words quite extensively, this in no way detracts from the strength of Japanese culture or its fundamental homogeneity. Most Japanese are not well prepared for fluent communication with the outside world and, likewise, foreigners do not find it easy to achieve fluency in Japanese. In some ways the study of foreign languages in Japan has lost most of its significance, except to those scientists and researchers who wish to puzzle out the original foreign texts. The teaching of foreign languages is also frozen into a system that students find boring because of the heavy emphasis it places on grammar and classical texts alone.

The difficulties in translation also lie in the fact that there are radical differences in syntax between Japanese and other languages. It is a common myth that Japanese is closer to Chinese than to other, Latin-based languages. This is only partly true (in the sense that Chinese is the root of Kanji). However, in terms of the basic structure of the language, Japanese remains completely different from both English and Chinese (incidentally, Chinese and English are languages in which the word order determines the meaning, as in “the cat chases the mouse”). Japanese remains a strictly agglutinative language in which the concluding word, which is a verb or adjective, attracts subsidiary elements that indicate the meaning in terms of tense, mood, politeness and whether the sentence is active, passive, negative, or even a question. Combining the Japanese word order with either English or Chinese words would produce sentences of confusing gibberish that would defy direct translation. The only other languages which bear a resemblance to Japanese in terms of structure are Korean and Turkish. However, few Japanese choose to learn Korean (though more Koreans learn Japanese), and even less learn Turkish.

One common problem encountered by translators working from Japanese into other languages is the sharp difference in the levels of politeness and formality in the language. This is difficult to handle for at least two reasons: One is the importance of the social and age differences that are rooted in the recent Japanese past and are deeply embedded in the culture. The other is the tendency of Japanese to omit elements like the subject from a sentence, if they are already clear from the context of the level of politeness. In Japanese, there is often no need to specify “I’, “you” or “they”, since in the age- and group- conscious society, polite, humble, or neutral forms have already indicated the subject.

Due to the general lack of resources available for the translation and localization of Japanese into other languages, the standard practice in the industry is to translate from Japanese to one language first (often English for obvious reasons), and then to continue localizing from English to the other target languages. Over 80% of FIGS localization is done in this way. The main drawback of this practice is obviously the higher chance of producing unreliable translations.

The “inter Asian” approach is, of course, somewhat different. The model for direct translation of Japanese into other Asian languages is well established, especially in respect to Korean and Chinese, and the influence of Japanese in Korea, Greater China and Southeast Asia can never be overestimated.

Another new dimension is being added by the ‘Pacific’ perspective. As part of the decentralization of the production process, more and more Japanese software is being designed and developed offshore. These peripheral development centers are mainly situated on the West Coast of the USA and in other Asian countries such as Singapore, India and the Philippines. This new production model has already resulted in some interesting developments. One is the need to re-engineer and localize the software back to the Japanese market; the other is the increased need for translations from Japanese into other languages (mainly English). In contrast to the past, when only materials targeted at users had to be translated, there is now an increasing need for materials at higher, more sophisticated levels (like research papers and design specifications) to be translated from the original Japanese source.

This will certainly provide new challenges and opportunities for the localization industry.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.