In the translation industry, translation education, and machine translation research, multiple
translations with the same original text can and do exist. In such cases, it is necessary to
recognize and share the differences between translations. This paper presents the process of
developing a scheme for describing differences between target documents (TD) that have the
same source document (SD). Our scheme includes a system for describing translation
differences and a decision tree. We first constructed a system consisting of concepts and their
definitions, which has three categories: Syntactics, Semantics, and Pragmatics. We then
created a decision tree as a procedure for applying the system, and defined several rules to
enable us to consistently apply our scheme to the translation data. This paper also describes
the application of our scheme to 50 types of TDs. The result shows that our scheme is helpful
for describing the differences between TDs with high resolution.
View full abstract