This study discusses how different coding systems of voice cause modulation in the translation process. Previous studies (Kim 2005, Suh 2013 among others) show that more than half of Korean passives are translated into English actives and more than half of English passives are translated into Korean actives. They try to provide explanations for the voice modulation, which don't seem to be satisfactory in a sense that their accounts are not as comprehensive as this study tries to reveal in the framework of Functional Grammar (Givón 1993, 2001). This study argues that English and Korean passives code discourse-pragmatic function (such as topicality defocusing) as well as language-specific semantic functions (such as volitionality defocusing, responsibility defocusing, empathy defocusing), which are the main cause of the modulation in Korean-English translations. It also shows that typological differences between DO-language and BECOME-language (Ikegami 1991) and other language-specific grammatical systems can be other causes of the voice modulation.

• 김은일(부경대학교)