Teaching MT through pre-editing: three case studies

Teaching MT Through Pre-editing: Three Case Studies Chi-Chiang Shei Department of Foreign Languages and Literature National Tsing Hua University 101, Sec.2, Kuang-Fu Road, Hsinchu 30013 Taiwan ccshei@mx.nthu.edu.tw Abstract This article reports on three cases of teaching translation or English as a foreign language using pre-editing tasks with a machine translation system. Trainee translators or English learners were asked to input a Chinese or English paragraph into an MT system, observe the irregularities in the output, and subsequently edit the source text and input it again in the hope of getting better output from the MT system. Student reports on pre-editing procedures were collected and analysed, as well as thoughts and suggestions about using MT. It is argued that pre-editing MT input can boost student learning in the cognitive and affective domains, apart from training students to use MT systems. 1. Introduction It has been explained elsewhere by the author (Shei, 2002) that translating into a second language (L2), although not a preferred direction of translation, nevertheless necessary due to the lack of is native speakers in local settings where the service is required. In this context learning an L2 is a subset of learning to translate into the L2. Also, translation is increasingly being reconsidered as a viable means for learning an L2 (Cook, 1998). Thus it can be argued that L2 learning and learning to translate become more and more intertwined. This article explores the use of MT not only in the context of translation teaching but also in the context of L2 learning, showing that MT can be “a useful engine” in both settings, probably providing a convenient link for the two intellectual activities. Based on the types of input into an MT system, three categories of pre-editing tasks are distinguished in this article. � Pre-editing text written in mother tongue: Input is a Chinese text written by the native speaker. Output is the English text produced by the MT system. Object is to modify the Chinese text until the English translation is in the best condition. � Pre-editing L2 text written by native speaker: Input is an English text written by the native speaker. Output is the Chinese text produced by the MT system. Object is to modify the English text until the Chinese translation is in the best condition. � Pre-editing L2 text written by learner: Input is an English text written by the learner. Output is the Chinese text produced by the MT system. Object is to modify the English text until the Chinese translation is in the best condition. 2. Pre-editing mother language A class session in the General Translation module instructed by the author was devoted to pre-editing MT input. The direction of translation was from Chinese to English. There were two MT systems used (from which the student should choose one) which offered free online translation facility: 1. SYSTRAN at http://babelfish.altavista.com/translate.dyn, (Babel Fish Translation) 1 and 2. Beijing Golden Bridge Translation Port Network at http://www.netat.net/. Here we will use a simple paragraph …

Content truncated to 500 words