From Eye To Voice: Animacy Effect And Error Patterns In English-spanish Sight Translation Of Relative Sentences
Auteurs : Von Hausen Felipe . Alvarez-zuniga Ambar . Molina-medina Katalina . Machuca-veloso Fernanda . Rozas Tania . Castillo Lucía .
Résumé
In this experiment, we examine the translational characteristics of English–Spanish sight translation of relative clauses, with a particular focus on how animacy influences error production in controlled laboratory settings. We collected oral sight translations from English (L2) into Spanish (L1) produced by thirty-three participants with a B1 level or higher according to LexTALE. Participants translated 28 sentences distributed across a 2×2 experimental design manipulating animacy (animate vs. inanimate) and clause type (subject vs. object relative clauses). Their performance was captured using eye-tracking technology and audio recordings. The transcriptions of the oral productions constituted the corpus analysed in this research. We structured our analysis around three main objectives: to identify the errors produced in sight translation, to classify them using the Multidimensional Quality Metrics (MQM) framework, and to assess the translated products. We tested three hypotheses: that object relative clauses would generate more errors than subject relative clauses; that object relative clauses with animate antecedents would yield the highest error rates; and that errors would be more frequent in content-related categories than in form-related categories. Our results partially supported these hypotheses. Although object relative clauses showed higher error counts descriptively, the effect of clause type alone was not statistically significant. By contrast, the animate object condition yielded the highest error rates overall, with both inanimate conditions producing significantly fewer errors. In addition, content-related errors occurred substantially more often than form-related errors across all experimental conditions. These findings suggest that error production in sight translation is shaped not by syntactic structure alone, but by the interaction between structural complexity and semantic features, such as animacy. More broadly, the study contributes to current research on the cognitive demands of sight translation by showing how semantic and syntactic factors jointly constrain translation performance under time pressure.
Mots clés
Error Analysis ; Language Processing ; Oral Production ; Psycholinguistic Experiment ; Sentence Comprehension ; Translation Performance
