본문 바로가기
  • Home

The Significance and Limitations of Context Exploration in Parallel Corpus

Cho Joon-Hyung 1

1고려대학교

Accredited

ABSTRACT

A parallel corpus contains numerous practical examples concerning the translation; it consists of the translation texts produced by a linguistic community. The main goal of corpus-based translation studies is to extract translation equivalences from a parallel corpus. Since the 1980s, computer-assisted tools have been utilized to accomplish this goal. However, the results obtained by these tools fail to validate translation relationships because they ignore the context in which the original text was used. To overcome this limitation, researchers have to return to the context. The Keyword in Context (KWIC) and the co-occurrence analysis are the most efficient methods for determining meaning in the domain of the corpus-based studies. However, to define the context for exploration, these two methods rely mainly on formal criteria such as the sentence or the translation alignment, which corresponds to a sentence-to-sentence correspondence between the original text and its translation. Such a formal definition of context can sometimes obscure the translation results because computers are unable to recognize the human cognitive processing. Therefore, a formal study of corpus-based translation studies must consider these inevitable problems and consider more various conditions of inquiry.

Citation status

* References for papers published after 2023 are currently being built.