This study investigates the hypothesis that second language lexical forms are often mapped to the existing semantic content of their first language translations, rather than to new semantic specifications of their own. The findings support the presence of first language semantic content in second language lexical entries.