Computer tools to assess translations from English into Spanish

What is it about?

This paper is about using several computerized tools to develop ways of assessing the quality of translations from one language into another. We have used large corpora of original texts in English and Spanish to determine how these two languages employ specific linguistic resources: intensifying quantifiers (much, loads, quite a lot), nominal modifiers (premodifying adjectives and of-phrases) and past tenses. Then, we compared the results with texts translated from English into Spanish. We compiled a large translation corpus of original English texts published after 2000 and translated and published in the European variety of Spanish. The differences found between original and translated Spanish show evidence of translation universals, i.e. divergences between original and translated language use. This study has found instances of the simplification hypothesis, cases of interference, and the unique-item hypothesis. First, translated texts do not exploit the full range of options available in the language. Instead, they overuse the most frequent resources, f. e. intensifying quantifiers. In the case of nominal modifiers, the study has revealed clear cases of interference since translated texts overuse premodifying adjectives, as they are prevalent in the source language, English. Finally, the case study on tenses shows that one of the core functions of the imperfect in original Spanish is rarely used in translated Spanish.

Featured Image

Why is it important?

This paper illustrates how a careful contrastive analysis of original and translated language can provide insights into translation universals. Identifying particular linguistic resources that tend to fall into one particular category of translation universals enables us to develop tools for translation quality assessment. In addition, these results can be used in translation training and practice.

Perspectives