3. TRANSLATION TECHNOLOGY: PAST AND PRESENT
Translation technology is defined as “the branch of translation studies that specializes in the issues and skills related to the computerization of translation” (Chan, 2004:258). In application, it can be further classified as MT (Machine Translation), HAMT (Human-Aided Machine Translation), and CAT (Computer-Aided Translation), while in terms of its technique components, it includes Translation Memory (TM), Terminology, Terminology Database (TDB), etc.
3.1 Historical Sketch
In ancient Greek, some intellectuals explored the idea of machine translation by suggesting using mechanical methods to analyze human language (Feng, 2004:12). Modern translation technology has been developed for about 70 years after it came into being in the 1930s. This may be divided into four stages, the beginning phase from 1930s to 1966, the wane from 1966 to 1970, the revival from 1970 to the end of the decade, and its prosperity from the end of 1970s onwards (Figure 1).
Figure 1: Development of Machine/Machine-aided Translation (c.f. Feng, 2004: 20)
Many translation systems have been developed, among which TAUM-METEO in Canada, EUROTRA in EEC (or EU), etc., are successful stories of this technology.
3.2 State of the Art
The speed of machine translating is quite surprisingly fast, but the effect is not always satisfactory in every field. Hence, translation memory, a new translation technology, is now widely used.
Translation memory, defined as “a multilingual text archive containing segmented, aligned, parsed and classified multilingual texts, allowing storage and retrieval of aligned multilingual text segments against various search conditions” (Chan, 2004: 251), abbreviated as TM, is a program in computer-aided translation. The program stores paralleled segment of source text and target text in a database, the “memory”. When a new source text is to be translated by the translation system, the program will search the database for similar source segment for each one to be translated, and give the target segment as a suggested translation. The translator can judge and decide to accept, modify, or reject the translation provided by the system (Chan, 2004: 251). In other words, translation memory is like the learning ability of a translation system. It can memorize what has been translated. When coming across a similar text, it will try to translate automatically based on its learning. The more the system translates and memorizes, the better and more effective it will be.
Another technology is Terminology. When many translators work on a big translation project, such as the translation within EU, there must be established translations for certain terms. Translation system can put the terms in a term bank or a database, from which translators can search the translation of a certain term. This technology is quite helpful in guaranteeing terminological consistency of team work and achieving terminology standardization in translation.
Today, there are various types of software based on translation memory and terminology. SDL TRADOS can be cited as an example, which is the leader in the industry and has its translation business based on the concept of “translation memory”. It can support translating among 65 languages, and authorizes users to manipulate memory database and term bank. Moreover, it supports many file formats, including .doc, .ppt, .html, .xls, etc., and can maintain the typeset of the translation same as the original file. Now, SDL TRADOS has over 40,000 corporation customers and has taken up 70% of the share in computer-aided translation software market worldwide.
Obviously, localization and translation technology cannot go separately. The accuracy and efficiency of localization depends on, to a large extent, on the evolution of translation technology.
3.3 Interaction of Localization, Translation Technology, and Terminology Standardization
Localization is far more than translation, which is a very important component. While translation technology will undoubtedly accelerate the process of localization, localization provides tremendous opportunities for the development of translation technology. In other words, translation technology would not have developed so much if it had not been for the need from localization. Take translation memory as an example. If different translation systems can exchange their translation memories, the efficiency of them will be improved. So, there should be standards to regulate translation memory in different systems. Thereupon, Translation Memory Exchange (TMX) has come into being, which is “the vendor-neutral open XML standard for the exchange of Translation Memory (TM) data created by Computer Aided Translation (CAT) and localization tools” (http://www.lisa.org/standards/tmx, retrieved on 15 March 2008). It will also make the interchange of translation memory easier and safer.
Another example is that html tags in a webpage will be recognized as images in SDL TRADOS so that translators don’t have to handle them. As a matter of fact, these tags function as defining the typeset, hyperlinks, etc. If they were translated, the webpage would not be what has been expected. This can be illustrated by the example below:
Here are some source codes written in HTML (Hypertext Markup Language), because computer technology originates from USA, an English-speaking country, code language is usually English. In Code 1, a one-row-and-one-column table is defined, and in it there is a text: SDL TRADOS. Code 1 will be parsed by a web browser and displayed as Figure 2. In website localization, content in the brackets shouldn’t be translated because they are tags regulating the typeset of the webpage, and some are even program sentences. When these tags are translated into, say, Chinese, as Code 2 shows, the web browser can not parse the code correctly, so it will just take it as pure text, and there will be something missing, as Figure 3 displays. However, what we want to see is Figure 4, i.e., just the information in the webpage rather than the tags in the source code.
As a matter of fact, the source code of a web page is far more complicated than the example shows. If every tag is translated, doubtlessly to say, the web page will be messier than that in Figure 3.
In addition to the demands on the development of translation technology, localization also calls for terminology standardization. In different language backgrounds, terms are not unified, which may cause problems in exchange and understanding. It is necessary to build up standards for terms in different cultures, so that communication will be more convenient. This can partly help to explain the emerging businesses related to TBX (Termbase Exchange). With powerful function in terminology management, modern translation technology can make this possible.
4. LOCALIZATION AND TRANSLATION TECHNOLOGY IN THE CHINESE CONTEXT
Because of the distinction between Chinese and foreign languages, localization and translation technology in the Chinese context have their distinctive features.
4.1 Localization in the Chinese Context
Localization in China started in around 1993 when some software localization service providers came into being in Beijing and it has made considerable progress in the last decade (http://www.globalization.com.cn/news/view.asp?id=903, retrieved on 15 March 2008).
The number of people who use Chinese today has reached 1.5 billion, a quarter of the world population and, for a long time, China has been regarded as a huge world manufacturer of low-cost labor force and a large market for products and services. More and more foreign companies come to invest in China. A problem that they have to face is to adapt their products and services to the Chinese context, i.e., to carry out localization for them. For instance, Nokia, the world leading mobile phone manufacturer, began providing manuals in simplified Chinese and launched the first GSM mobile phone in China in 1995 (Chari & Wu, 2007). In 1998, it realized that Chinese inputting method should be provided in its products. So in 1999, its Production Creation Centre was set up in Beijing, developing functions like user interface and inputting method for Chinese users. Later, manuals in traditional Chinese were also supplied for users in Hong Kong and Taiwan. Gradually, the need for Chinese user interface and manuals rose in areas outside China, such as Malaysia, Singapore, Vietnam, and other parts of Asia. In addition, some functions, 3G for example, are not useful in China’s Mainland, so Nokia has to remove them in the process of localization.
Nokia’s localization model for China market shows that in the Chinese context, localization is not merely translation of text. In fact, it also includes adaptation of products in some other aspects, such as Chinese user interface, Chinese inputting function, special functions like lunar calendar in China, etc., owing to differences in language, culture, and technical environments between China and other countries, Chinese and English in particular.
4.2 Translation Technology in the Chinese Context
In China, translation technology development began in the 1950s. Researchers in China’s Mainland, Taiwan and Hong Kong all have contributed greatly to it (Qian, 2005).
There are now many translation systems and software in China, such as Dr. eye, Huajian, Kingsoft Express, Oriental Express, Tongyi English-Chinese Translation System, Yaxin CAT, Yiba, etc., all of which can help to deal with translation between English and Chinese.
With China’s entry into WTO, more and more foreign companies come to China, along with which is the need for localization. One point of note is that different Chinese languages, written form in particular (traditional and simplified Chinese), are used in China’s Mainland, Taiwan, Hong Kong and Macao, which implies that the translation necessitated by localization in the Chinese context is not merely translating from foreign languages into Chinese, but into different “Chineses”.
This is a big challenge or rather an opportunity for the development of translation technology in the Chinese context. In reality, the scenario is much more complicated, since translation is not identical with localization, the latter being more inclusive in its subject areas as it was mentioned above.
4.3 Interaction of Localization and Translation Technology in the Chinese Context
In the Chinese context, localization is particularly necessary, because Chinese culture, language, and social background are quite different from those of foreign countries. And localization in the Chinese context has challenged translation technology, owing to the big gap between Chinese and foreign languages, such as differences in syntax, character de/en-coding, etc.
Against the localization background, standardization is very important in that it is playing a significant part in improving the product quality and qualification of an enterprise. If there were no standards, quality would not be guaranteed. In the course of localization, it is necessary to adapt a lot of elements to local standards. Companies have to draw up standards suitable in the Chinese context in order to promote their productivity and efficiency and cut down the cost.
Take E-Business as an example. There are tens of thousands of terms of products in the market, and different companies have their own terms. Therefore, it is necessary to clarify them in accordance with certain standard in the first step so that it will be convenient to manage so many terms in the cyberspace and transmit them between different organizations or businesses. The second step is to translate them into the local language, in which translation technology is very helpful at least in the following three aspects.
In a translation project, semantic accuracy is of the utmost significance, for it must convey precise information to the readers. For example, if a segment in a contract is translated in two different ways in the target language, misunderstanding may be caused with the consequent financial losses. Translation memory systems will surely enable translators to avoid this issue. Take SDL TRADOS for example, in its translator’s workbench interface, translators can take advantage of “concordance” function so as to look up segment equivalents. Also, the system will show users to what extent the new segment is matching with what is stored in the memory. With the help of translation technology, translators can keep semantic accuracy more conveniently and effectively without compromise in quality.
A huge translation project these days usually involves many translators working via the network on different parts of the project in different places, a major point of which is to identify equivalents for specialized terms (Bowker, 2002:77). When coming across a certain term like “mobile phone” that has more than one equivalent in Chinese, a translator from China’s Mainland may translate it into “手机”, while another from Hong Kong may use “流动电话” or “手提电话”. It will be troublesome that two literally different terms referring to the same object exist in the same technical documentation if it is just for users in either China’s Mainland or Hong Kong. Fortunately, a terminology management system with contextual information on a term used in different regions can solve this problem, because it can “help to ensure greater consistency in the use of terminology, which not only makes documentation easier to read and understand, but also prevents miscommunications.” (Bowker, 2002:77)
With its advantages in speed and quality, translation technology can cut down the cost in translation. It enables a translator to work more efficiently and effectively. Moreover, some reusable sources in the database make it possible to avoid doing similar work on materials that are highly repeated, such as routine documents (monthly report, quarterly report, etc.). Further more, it will be easier to keep the typeset of the target text same to that of the source text, like SDL TRADOS and Yaxin CAT. All these advantages definitely save time in translation.
As Covos (2005) argues, translation can reduce “the tax burden”, “warehousing costs and cash flow optimization”. In this sense, translation technology will benefit localization by lowering much cost.
On the other hand, demands in the course of localization will bring about more opportunities for translation technology development. For instance, in computer science, English letters are single-byte characters, while Chinese characters are double-byte ones. Problems will pop up in character encoding. Researchers’ efforts on this issue will ultimately lead to a big promotion in translation technology.
5. FUTURE OF LOCALIZATION AND TRANSLATION TECHNOLOGY IN THE CHINESE CONTEXT
On the basis of the analysis in the previous sections and with the development of economic globalization and the increasing influence of China on the global stage, it is evident that localization and translation technology will have a bright future in the Chinese context. Meanwhile, there are also some barriers on the way, especially for translation technology.
5.1 Future of Localization in the Chinese Context
There is no doubt that China is a huge market. In addition, it is rich in affordable labor force and hard-working engineers. However, less than 1% of the population can use English properly. In the past decade, China’s economy has kept growing by nearly 10% each year, with its GDP taking up 11% of that of the world (Zhao, 2003). Now, foreign companies are swarming into the Chinese-speaking countries and regions, which include mainly China’s Mainland, Taiwan, Hong Kong and Macao. Since the scenarios are quite similar, the term “Greater China” is used to refer to all four for commercial strategies.
Localization in the Chinese context is not within the Greater China only, but also in other Asian countries such as Singapore, where Chinese is widely used as an official language. It can be estimated that localization industry in the Chinese context will be quite promising and profit in this area will be very considerable.
5.2 Future of Translation Technology in the Chinese Context
With the rapid growth of Internet Technology, research on translation technology is still an important field and industry in the 21st century, which is closely related to computational linguistics, artificial intelligence, and many other disciplines. For example, SDL TRADOS received an annual revenue of over 94 million pounds in 2006 (http://www.sdl.com/company/press-releases-sdl/press-release-sdl.htm?id=1122, retrieved on 15 March 2008). Also, there will be many job positions for translation technology engineers, and translators who are proficient in translation technology and able to translate “software applications, multimedia products, web pages, and even on-line chat sessions” (Bowker, 2002: 131) will be more competitive in the job market. So it is no longer adequate for a translator to be both bilingual and bicultural. In order to be qualified, he/she should also be technologically proficient, which has called for a technological turn in translation studies (Chan, 2004), translator training in particular.
When we look forward, there is still a long way to go before translation technology can truly meet the demands of the users. Generally speaking, things to be done for the research and development of translation technology in the Chinese context should gear towards the following aspects.
As it is pointed out by Hutchins (1999) and applicable to MT and MAT in the Chinese context (Qian, 2005), translation software now available are still expensive. How to develop an efficient system that is of low cost, high reliability and requires less work on constructing the translation memory for individual translators is another emerging problem. Besides, translation systems for minor languages should also be further and proactively explored.
It is necessary for scholars in the Chinese context to learn from and exchange with others (Qian, 2005) in localization and translation technology. Successful experiences of the Localization Research Center at the University of Limerick, Ireland (http://www.localisation.ie/index.htm, retrieved on 15 March 2008), the leader in localization/translator education and research, could be a resources pool for interested people to do the same in the Chinese context. Future translators who are proficient in language, culture, as well as technology, will be more competitive in the job market.
A SinoTermbank should be proposed and established. In Europe, there is EuroTermBank, which is “an online terminology portal publicly available since January 2007” (Vasiljevs, 2006). It is very useful for translators and translation service providers. According to him,
a large number of potential EuroTermBank users are freelance and/or agency employed translators who are increasingly moving from usage of paper to electronic dictionaries and Web-based term collections and reference sources. The EuroTermBank service provides individual translators and translation companies with authentic target-language equivalents for terms, thus contributing to the efficiency and quality assurance of translation. Translation companies may potentially also become EuroTermBank content providers and develop CAT tools to access this content (ibid).
However, the EuroTermBank does not provide service in Chinese. So, it is urgent to build up a Sino termbank in order that it will be useful for translators and translation agencies related to Chinese.
Attention should be paid to “spoken language translation”, which still eludes us and could be a very ambitious project (Somers, 2003: 7).
Attention should also be paid to network teamwork, from stand-alone systems, so that multiple users can share the same resources.
It is quite effective and profitable for companies to carry out localization in the course of exploring market overseas for products and services. In this process, translation technology has played an irreplaceable part, in that it solves the language problem in localization and facilitates the process. Meanwhile, the need for localization provides opportunity and market for the development of translation technology. In other words, localization and translation technology are two processes that are mutually beneficial. At present, the need for localization is becoming more and more imminent for all companies that are planning to sell their products and services in foreign markets. China is becoming a big market in the world, which provides many opportunities for foreign companies. Localizing their products and services in the Chinese market will bring them more opportunities vital to their development. Hence, the development of translation technology in the Chinese context is quite significant, for it will affect the speed of localization to a large extent. This will undoubtedly supply many opportunities for translation technology and for the translation industry, especially companies and translators (professional and free-lance) concerned with research on translation technology and translation service for localization.
 Bowker, Lynn (2002): Computer-Aided Translation Technology: A Practical Introduction. Ottawa: University of Ottawa Press.
 Chan, Sin-wai (2004): A Dictionary of Translation Technology. Hong Kong: The Chinese University Press.
Chari, Deepak and Eileen Wu (2007): Nokia’s Localization Model for China market. Available at http://www.lisa.org/archive/forums/2007beijing/presentations.html, retrieved on Apr. 10, 2007.
Covos, Wagner (2005): Using Translation to Cut Costs in the Energy Sector. Available at http://www.gala-global.org/en/resources/CcapsCovos_EN.pdf, retrieved on Apr. 13, 2007.
 Feng, Zhiwei (2004): Studies of Sci-Tech Translation. Beijing: China Translation and Publishing Corporation, 12-34.
 Hutchins, John (1999): The Development and Use of Machine Translation Systems and Computer-aided Translation Tools. Paper Presented at the International Symposium on Machine Translation and Computer Language Information Processing, 26-28, June, 1999, Beijing.
 Lommel, Arle and Rebecca Ray (2007): The Globalization Industry Primer. Switzerland: The Localization Industry Standards Association.
 Qian, Duoxiu (2005): “Prospects of Machine Translation in the Chinese Context”. Meta, 50 (4): 234-242.
 Somers, Harold (2003): “Localization and Translation”. In Harold Somers (ed.) Computers and Translation: A Translator’s Guide. Amsterdam/Philadelphia, John Benjamins Publishing Company, 67-86.
 Vasiljevs, Andrejs (2006): “EuroTermBank: Exchange of Multilingual Terminology Resources”. China Standardization, (12): 52-53.
 Zhao, Jie (2003): “Localization in China”. Available at http://www.globalization.com.cn/news/view.asp?id=671, retrieved on Feb. 13, 2008.