Multilingualism on the Internet In recent years, American culture has increased its worldwide influence through internationa

admin2013-07-02  35

问题                     Multilingualism on the Internet
    In recent years, American culture has increased its worldwide influence through international trade and Hollywood productions. As the Internet reaches into ever more remote corners of the globe, an obvious question arises: Will it amplify this trend, so that English is used everywhere? Or will a diversity of languages enrich the online universe? Some observers predict that local languages will not survive online: English will rule.

    Such a sweeping dominance will have drawbacks. Most people use English as a second language, and their grasp of it may be quite elementary — sufficient only for understanding basic information such as the weather report, and sometimes not even that. For more in-depth discussions, almost everyone tends to fall back on his or her native language. If the Internet does not allow multilingual conversations, its role as a facilitator of international communication will be severely limited. Mistakes and mis-understandings will become rampant, and many users will be cut but of the tremendous opportunities that international communication has to offer.
    Several forces will affect the diversity of languages most likely to be found on the network in the future. At present, about 60 percent of the Internet’s host computers are located in the U. S. Almost in every corner of the globe, the world’s connections to the Internet are very popular. Furthermore computers everywhere are becoming increasingly linked. As the cost of installing communications networks continues to fall, the distribution of Internet users will come to resemble that of computers.
    With its low cost and theoretically easy-to-use technology, the Internet allows some writers — particularly those using Latin alphabets to publish or exchange messages in their own tongue. Some promoters of native languages have already used the medium to their advantage. For instance, roughly 30 percent of all World Wide Web pages published in French come from Quebec, even though French Canadians represent only 5 percent of all French speakers. But the worldwide reach of the Internet also favors a language that can be, at least superficially, understood by the largest number of people. As a result, I believe the Internet will support many languages for local communications and English for limited international discourse.
    Of course, the technical difficulties of communicating in the majority of the world’s languages are not trifling. Hardware and software were first designed to process English text. But difficulties linger even with standard Latin characters. In the early days of the Arpanet — the predecessor of the Internet — only electronic-mail messages coded with seven-bit ASCII text could be sent.(In this code, each of 128 characters is specified by a string of seven binary digits.)Nowadays the Extended Simple Mail Transport Protocol permits the processing of the eight bits required for communicating in ISO-Latin, prescribed by the International Organization for Standardization. ISO-Latin allows for 256 characters, so that the diacritical signs(such as acute and grave accent marks)of all western European languages can be displayed. But because many interlinked computers on the network have outdated software, the eighth bit sometimes gets dropped, rendering the message almost incomprehensible. Out of 12,000 users who received the daily French news that I sent out at a time via the Internet, 8,500 asked to receive a version coded in seven-bit ASCII rather than the crippled ISO-Latin version.
    Although some recent programs can express their output in many different scripts, most are essentially bilingual: the software can deal with only one local language, such as Japanese, and English. One could argue that the technical obstacles to displaying other alphabets are temporary. Unicode(ISO 10646), a coding scheme for characters of most of the world’s scripts, is being progressively implemented. The code allows a user to receive almost any language(although it may not always be properly displayed).
    We nonetheless have a long way to go before we get to a truly multilingual Internet, in which an author can include a Greek quotation in a Russian text that will be properly displayed on the reader’s computer in South America. Software standards with this kind of capability are emerging. But the primary software producers, in their race to dominate the market, keep producing new versions, giving little chance to the usually small enterprises that develop multilingual products to keep up.
    In real life, interpreters help to overcome language barriers. Human translators can also be employed on the Internet, but given the volume and variety of exchanges, they will play a limited role. Only machine-aided translation can bring us closer to a world, perhaps a utopia, where all the attendees at a virtual conference of the United Nations can each use his or her native language, which will be simultaneously translated into all other languages.
    Research on machine-aided translation has been pursued over the past 50 years with somewhat mixed results. The systems actually in use are small in number and located mostly in Japan, Canada and Europe — the last of which faces the largest multilingual translation load. Electronic interpreters are usually just bilingual and need to be heavily specialized if they are to produce raw translations good enough to be revisable by human editors.
    The first system available for general public use was Systran, which could translate 14 pairs of languages and was accessible as early as 1983 on the French Minitel network. Used by the European Commission, Systran now converts hundreds of thousands of pages a year. Another success story is the Meteo system, which translates Canadian meteorological bulletins between English and French. It handles 80,000 words(about 400 bulletins)every day, with only three to five human editing operations for every 100 words.
    Multilingual translation will benefit from a two-step process now being developed by several groups. The text is first thoroughly analyzed into component parts(title, paragraph and sentence), clarified when possible by a dialogue with the author, then translated into an intermediate, abstract representation — which is used to generate translations in different languages. The effort is worth the expense when the text needs to be translated into more than 10 languages. The United Nations University in Tokyo has recently announced a 10-year collaborative project for implementing this two-stage scheme.
    But a truly multilingual Internet will come to pass only with concerted international effort. Will we give it enough priority? The answer is not clear. It is so easy to let ourselves drift toward English as a unique common language.
In the second step of the two-step process, the text is translated into______.

选项

答案an intermediate,abstract representation

解析 处理过程的第二步即原文中first…,then后的内容。
转载请注明原文地址:https://jikaoti.com/ti/o1bFFFFM
0

最新回复(0)