Predictors of ear-voice span, a corpus-based study with special reference to sex

This paper reports on a study on Ear-Voice Span (EVS) carried out on corpus data drawn from the European Parliament Interpreting Corpus Ghent, where sex is included as a predictor alongside several other variables. Ear-Voice Span is considered to be an indicator of cognitive processes in simultaneous interpreting and has therefore been selected to determine whether potential cognitive sex differences trigger different EVS patterns in men and women. Differences between men and women are reported in individual studies for tasks that are crucial to interpreting (Aerts, 2003; Kimura & Seal, 2003; Loonstra et al., 2001; Maitland et al., 2004 among others). However, meta-analyses tend to show that the reported cognitive differences between the sexes are exaggerated. This study uses corpus-based research methods to analyse the EVS of male and female interpreters in the European Parliament against the background of other known predictors of EVS. The data sample consists of 180 source texts and interpretations in six language pairs. The hypothesis was not confirmed as no sex differences were found. This research project helped identify relevant predictors of EVS: delivery rate, languages and interpreter’s disfluencies