Abstract
Instructors of English-to-Japanese translation know from their experience that translations by students, particularly those at the entry to intermediate levels, contain far more ko-type demonstratives (roughly equivalent to “this”) and so-type demonstratives (roughly “the” “that”
or “it”) than is typical for comparable texts written originally in Japanese. The unusually high occurrence of those demonstratives can be explained in large part by the students’ unnecessary or inappropriate use of them as translations of, among other elements, English cohesive devices that Halliday & Hasan (1976) categorized as reference items. Redundancy cannot explain all of the increased use of ko and so demonstratives, however, since professional translators also tend to produce texts marked by their heavy use. This paper attempts to determine where such extra ko and so arise from.