The effect of different fuzzy match rates in the translation memories on the effort of translators

Technological developments have rapidly made computer-assisted translation tools widespread. Translation memories, one of the computer-assisted translation tools, store previous translations aligned to the source texts. By comparing the current source text with the previous source texts and translations, translation memories suggest translations at different rates of similarity. Thus, they make translation faster and more consistent. In addition, translation memories also decrease the effort of translators. However, the assumption that translation memories decrease the effort of translators has led to a discount scheme in the translation market. According to this discount scheme, the higher the fuzzy match ratio is, the less payment is given to the translators. This discount scheme has been frequently criticized by the translators as unfair. This study focuses on the effort that translators spend while translating different types of matches in translation memories. The aim of the study is to determine whether the effort of the translators is changing with the fuzzy matching rate. A total of 14 participants translated two technical texts of approximately 190 words from English to Turkish and vice versa. The efforts of translators in translation experiments were measured. In order to operationalize the effort concept, Krings' notions of temporal, physical and cognitive effort (1986; 2001) and O'Brien's (2008) concept of subjective effort have been employed. The temporal effort was calculated based on the translation time; Levenshtein distance was employed to calculate the physical effort. In order to calculate the cognitive effort, the number of fixations and the duration of average fixation were measured. Subjective effort was measured via the editing effort of the subjects. Employing the data triangulation method, the study also employed the qualitative method of interviews with the participants. Qualitative data were analyzed by content analysis method and quantitative data were analyzed by correlation analysis via IBM SPSS Statistics program. It was found that the effort of translators at the participant level did not change according to the fuzzy match ratio. At the group level, however, as the match ratio increases, the effort decreases in both translation directions. In addition, according to the results of the interviews, it was seen that the participants did not agree on the translation memory and the demand for discounts.